SELF-RECOMMENDING!

Predictions For 2020

At the beginning of every year, I make predictions. At the end of every year, I score them. So here are a hundred more for 2020.

Rules: all predictions are about what will be true on January 1, 2021. Some predictions about my personal life, or that refer to the personal lives of other people, have been redacted to protect their privacy. I’m using the full 0 – 100 range in making predictions this year, but they’ll be flipped and judged as 50 – 100 in the rating stage, just like in previous years. I’ve tried to avoid doing specific research or looking at prediction markets when I made these, though some of them I already knew what the markets said.

Feel free to get in a big fight over whether 50% predictions are meaningful.

CORONAVIRUS:
1. Bay Area lockdown (eg restaurants closed) will be extended beyond June 15: 60%
2. …until Election Day: 10%
3. Fewer than 100,000 US coronavirus deaths: 10%
4. Fewer than 300,000 US coronavirus deaths: 50%
5. Fewer than 3 million US coronavirus deaths: 90%
6. US has highest official death toll of any country: 80%
7. US has highest death toll as per expert guesses of real numbers: 70%
8. NYC widely considered worst-hit US city: 90%
9. China’s (official) case number goes from its current 82,000 to 100,000 by the end of the year: 70%
10. A coronavirus vaccine has been approved for general use and given to at least 10,000 people somewhere in the First World: 50%
11. Best scientific consensus ends up being that hydroxychloroquine was significantly effective: 20%
12. I personally will get coronavirus (as per my best guess if I had it; positive test not needed): 30%
13. Someone I am close to (housemate or close family member) will get coronavirus: 60%
14. General consensus is that we (April 2020 US) were overreacting: 50%
15. General consensus is that we (April 2020 US) were underreacting: 20%
16. General consensus is that summer made coronavirus significantly less dangerous: 70%
17. …and there is a catastrophic (50K+ US deaths, or more major lockdowns, after at least a month without these things) second wave in autumn: 30%
18. I personally am back to working not-at-home: 90%
19. At least half of states send every voter a mail-in ballot in 2020 presidential election: 20%
20. PredictIt is uncertain (less than 95% sure) who won the presidential election for more than 24 hours after Election Day. 20%

POLITICS:
21. Democrats nominate Biden, and he remains nominee on Election Day: 90%
22. Balance of evidence available on Election Day supports (as per my opinion) Tara Reade accusation: 90%
23. Conditional on me asking about Reade on SSC survey, average survey-taker’s credence in her accusation is greater than 50%: 70%
24. …greater than 75%: 10%
25. …greater than credence in Kavanaugh accusation asked in the same format: 40%
26. Trump is re-elected President: 50%
27. Democrats keep the House: 70%
28. Republicans keep the Senate: 50%
29. Trump approval rating higher than 43% on June 1: 30%
30. Biden polling higher than Trump on June 1: 70%
31. At least one new Supreme Court Justice: 20%
32. I vote Democrat for President: 80%
33. Boris still UK PM: 90%
34. No new state leaves EU: 90%
35. UK, EU extend “transition” trade deal: 80%
36. Kim Jong-Un alive and in power: 60%

ECON AND TECH:
37. Dow is above 25,000: 70%
38. …above 30,000: 20%
39. Bitcoin is above $5,000: 70%
40. …above $10,000: 20%
41. I have bought a Surface Book 3 laptop: 60%
42. Crew Dragon reaches orbit: 80%
43. Starship reaches orbit: 40%

SSC, ETC:
44. I do another Nootropics Survey this year: 70%
45. I do another SSC Survey this year: 90%
46. I start a Reader SSC Survey this year: 60%
47. I start a SSC Book Review Contest this year: 70%
48. I run another Adversarial Collaboration Contest this year: 10%
49. I publish [redacted]: 20%
50. I publish [redacted]: 50%
51. I publish [redacted]: 60%
52. I publish [redacted]: 80%
53. …conditional on being published, it gets at least 40,000 pageviews: 10%
54. I publish [redacted]: 60%
55. …conditional on being published, it gets at least 40,000 pageviews: 50%
56. More hits this year than last: 70%
57. Most hits ever this year: 20%
58. I finish Unsong revision this year: 40%
59. New co-blogger with more than 3 posts: 10%

FRIENDS:
60. No new long-term (1 month +) residents at group house by the end of the year: 70%
61. Koios has said his first clear comprehensible word: 50%
62. [redacted]: 40%
63. [redacted]: 60%
64. [redacted]: 80%
65. [redacted]: 80%
66. [redacted]: 95%
67. [redacted]: 10%
68. [redacted]: 95%
69. [redacted]: 80%
70. [redacted]: 80%
71. [redacted]: 50%

PROFESSIONAL
72. I’ve gotten at least one new patient to do a full wake therapy protocol: 60%
73. I have specific, set-in-motion plans to quit work / start my own business: 5%
74. I work the same schedule and locations I did before the coronavirus: 80%
75. I get a bonus for 2020: 20%

PERSONAL:
76. [redacted]: 70%
77. [redacted]: 70%
78. [redacted]: 95%
79. I travel to Alaska this year: 60%
80. [redacted]: 40%
81. [redacted]: 20%
82. I go on at least three dates with someone I haven’t met yet: 20%
83. [redacted]: 10%
84. [redacted]: 30%
85. [redacted]: 10%
86. I try one biohacking project per month x at least 5 of the last 6 months of 2020: 30%
87. I find at least one new supplement I take or expect to take regularly x 3 months: 20%
88. Not eating meat at home: 40%
89. Weight below 200: 50%
90. Weight below 190: 10%
91. [redacted]: 90%
92. [redacted]: 30%
93. [redacted]: 5%
94. I travel outside the country at least once: 10%
95. I get back into meditating seriously (at least ten minutes a day, five days a week) for at least a month: 10%
96. At least ten tweets in 2020: 80%
97. I eat at/from Sliver more than any other restaurant in Q4 2020: 50%
98. [redacted]: 30%
99. I do pushups and situps at least 3 days/week in average week of Q4 2020: 60%
100. I write the post scoring these predictions before 2/1/21: 70%

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

439 Responses to Predictions For 2020

  1. doubleunplussed says:

    So the US will be hardest hit but the response will be 50% likely to be considered an overreaction? Interesting.

    • Snickering Citadel says:

      I think he means actually numbers of dead people, not ratio.

      • oliveflowers says:

        This might be an important clarification.

        Scott?

        [ETA: wait- I’m not sure- I can’t find any prediction’s of Scott’s that “the US will be hardest hit”]

        • doubleunplussed says:

          I was happily equating most deaths to hardest hit, should’ve used the exact wording, sorry.

          Still seems strange to think it has a good chance to be judged an overreaction if it has the most deaths even in absolute terms.

          Unless this is a prediction about partisan thinking making people judge it as such despite the evidence.

          • Jerden says:

            I’m definitely hoping the response will be considered an overreaction, because that means that it’ll have worked. Unless someone you know actually dies of Covid, all the average person will remember after the pandemic is over is the massive inconvenience, and hey, nobody they knew actually died! I think this is likely in the US, based on he protests and Trumps rhetoric, but since I’m across the pond I have no idea what the average person in America thinks.

            The main problem then will be if there’s another outbreak (of Covid, or something else entirely), it’ll be worse than this one because we’ll be more reluctant to socially distance.

          • Wency says:

            I agree with Jerden. Human beings are generally bad at giving credit to precautionary measures that pre-empted something really bad from happening. We have a tendency to reward our leaders more for responding quickly and decisively to massive flooding than for building the dikes that prevent the flooding in the first place. Is there a name for this phenomenon? There ought to be.

            It’s looking like the US deaths from Coronavirus this year will ultimately be not too far from the number of deaths of a bad flu season — no more than 3x, maybe more like 2x.

            But some of the short-term economic consequences have exceeded prior problems by a factor of 10. This is what people will remember. That, and being quarantined, which is entirely unprecedented.

            The fact that the US is worst-affected is mostly irrelevant to how people think about these things. People in general are mostly just going to look at the absolutes. Was the US the country most affected by the 2017 flu season? The least? Practically no one knows or cares. The 2017 flu season just wasn’t a big deal.

            And that goes double for Americans, as we’re especially prone to not comparing ourselves to the rest of the world, for a variety of reasons.

          • ericsoderstrom says:

            The US can be hardest hit but still have overreacted if the reaction is extreme but ineffective. For example, if it turns out that the Swedish approach is more or less as effective as US response.

          • Mary says:

            It would kinda help if they weren’t shifting goalposts. If the purpose was to flatten the curve so that the hospitals aren’t overwhelmed, mission accomplished, start to open up because we have space.

            Changing the tune to “until we have a vaccination” only raises the suspicion that the first reason, and the second, are lies.

          • Edward Scizorhands says:

            There are many reasons for the lockdown. Nate Silver had this nice summary: https://twitter.com/NateSilver538/status/1252290480177655809
            (and thanks to Theodoric for helping me re-find it).

            I don’t think very many people are saying “lockdown for the general public until vaccine.” I’m sure a few people are, and a few more people might be not saying it but wishing to boil the frog. But it’s definitely a minority opinion.

            Sweden did a “controlled burn” where there was little official shutdown but lots of people voluntarily shutting down. They never overwhelmed their hospital system because the growth rate was small enough, and their initial infection number was small enough, that eventually herd immunity became a noticeable factor in slowing the spread.

            I’m not sure whether Sweden’s plan was a good idea or not. They have a bunch more deaths, but those might just be deaths pulled into the present from a few months in the future. I also don’t know how much less their economy has been hurt. And maybe waiting a few months would have given them a lot better information about treatment options.

      • Furslid says:

        I think Ratio is the way to go, because it’s the only way to grade responses. It might be easy to say that the US response was worse than country X by number of deaths, but this would be the wrong response if it had a lower death rate.

        I’d like to know how the US death rate ranks compared to similar the other developed countries, so we know how bad the response was. I’m thinking of grading the coronavirus response for each country in this group as follows. Death rate >115% of address = F. Death rate 105%-115% of average = D. Death rate 95-105% of average C. Death Rate 85-95% of average = B. Death rate <85% of average =A. This prevents us from either praising or condemning different responses unjustly.

        • EchoChaos says:

          Do you think that grading state by state in the USA is reasonable?

          They are far more autonomous than subdivisions in other countries and make a lot of their own decisions, but they don’t control their own borders.

          I could see it either way.

          • Furslid says:

            I think the grades might be less relevant, but they are still important. Some distinctions like NY vs NJ may be less relevant than others because of metropolitan areas spread over states.

          • JPNunez says:

            I don’t see why not.

            Italy’s response also varied a lot from city to city. Some areas were actually not hit that hard. And Italy is a lot less “federalized” than the USA.

            If some state was unfairly hit by a migration wave it will be reasonably notorious.

        • deciusbrutus says:

          Death rate only matters if roughly everyone gets infected, in which case you want to measure survival deaths per captia/per capita hospital capacity, with final units of Deaths/bed.

    • cevapcici says:

      It is more than likely that the response will be considered an overreaction as the response was based on an assessment of the disease’s severity, which severity will be reassessed downwards over time as a regression to the mean (for instance Public Health England downgraded the disease mid-March).

      The Oxford Centre for Evidence-Based Medicine notes ‘The current COVID outbreak seems to be following previous pandemics: initial CFRs start high and trend downwards. ‘

  2. 7. US has highest death toll as per expert guesses of real numbers: 70%

    For that to happen, India has to have less than a quarter the per capita death toll of the U.S. They have the advantage of a warmer climate, but they also have very crowded cities and, I expect, much worse medical care, so I don’t think it is likely.

    Other possible competitors, besides China, are Indonesia and Brazil.

    • Anteros says:

      Nigeria? Bangladesh? Both fairly populous countries with poor healthcare capabilities.

      One confounding factor for all these countries is that they have, possibly, fewer vulnerable people. Those with many of the co-morbidities talked about in relation to Covid-19 remain alive in America where they wouldn’t in, say India. It may that the higher proportion of obese, diabetic, hypertensive (etc) patients in America outweighs the poorer health systems in undeveloped countries.

      Not much more than speculation on my part.

      • Tarpitz says:

        I think the age profile and comorbidities issues are substantial, but so is the time at which this is being judged. I doubt anyone will have anything resembling good data for India in January next year, and perhaps not ever.

    • teageegeepea says:

      Indonesia has been doing much better than many expected. One theory is that people spend most of their time outside, where transmission is far less likely vs inside. The same might be true of India.

    • The Nybbler says:

      India has a much smaller percentage of elderly, so are likely to have a significantly lower per-capita death toll, since age is so significant. I also think they will have terrible reporting, and social desirability bias will pressure so-called experts into not calling them out (it’s bad to dunk on the Global South).

      • keaswaran says:

        I suppose it depends on whether you see describing a high death toll as “calling them out” or “urging us to help them”.

      • Robert Beckman says:

        Mild necropost, but hard to pass this up.

        India is in the Northern Hemisphere…… definitionally not the “Global South.”

    • aphyer says:

      I’m also surprised by the implied 10%-or-less probability that China has more deaths in absolute terms than the US but does not admit this in its official numbers. I would have had that substantially higher.

    • Elementaldex says:

      Having lived for a few months in India and travelled to a few of their major cities I put them at 50% chance of having the most deaths, but I would be totally unsurprised if it takes more than a year to get a good estimate on how many deaths they have. Their hygiene standards are very low, their cities are very dense, and their population is very high. They also don’t have as effective of a government as China which has many of the same risk factors.

    • IvanFyodorovich says:

      At this point I strongly suspect there is some X-factor that is reducing spread or mortality in large swaths of the developing world (including South Asia, Southeast Asia, and Africa). My top two guesses are immunity conferred by endemic respiratory viruses or differences in immune system caused by exposure to food and waterborne pathogens. The latter is a well-established phenomenon in the context of autoimmunity and allergy, though not to my knowledge in response to infectious disease.

      It’s really striking how in virtually every country with unsafe drinking water, there are no signs of mass COVID mortality. It’s hard for me to believe that’s because they all implemented sound containment policies or because of minor cultural differences. It’s not genetics, we know that because black Americans are being badly hit.

      Before someone says it, yes I’m aware that there is undertesting in these countries, but there’s no sign of an Italy/Spain/Iran/London/NYC surge of pneumonia cases.

      • ChrisA says:

        Maybe it is because they don’t have old people’s homes with a potential for one asymptomatic staff member infecting multiple vulnerable people. Most older people in India live with their family which means spread is much slower into the older population.

        • John Schilling says:

          How is that different than e.g. Italy? The usual hypothesis for why Italy had it so bad so early, after the bit with the one really dysfunctional hospital, was that younger people with mild cases would bring home the virus, infect their live-in grandparents, who would die horribly while coughing up enough SARS CoV-2 to turn all their kids and grandkids into a- or pre-symptomatic carriers.

          If you’re going to claim that old people living with their kids and grandkids is keeping India safe, then I’m interested in your theory as to what made Italy (and to some extent Spain) so bad.

    • soreff says:

      7. US has highest death toll as per expert guesses of real numbers: 70%

      >For that to happen, India has to have less than a quarter the per capita death toll of the U.S.

      Agreed.

      >Other possible competitors, besides China, are Indonesia and Brazil.

      Agreed. Perhaps also Russia?

  3. gbdub says:

    Regarding your Starship/Dragon predictions, both spacecraft have already reached orbit. I assume you mean “reach orbit with people on them”?

    • Scott Alexander says:

      That’s a good point for Dragon. Has Starship actually reached orbit?

      • ksteel says:

        It has not. No Starship has ever even had a test firing in fact. The Starhopper only reached a height of around 150 meters.

        Even if you were maximally generous and considered in-atmosphere freefall a kind of “orbit” it didn’t do that, its engines were constantly firing.

      • Lambert says:

        Starship’s not reached orbit, or even space.
        Some prototypes are being constructed and prepared to be flown around in atmosphere.
        Expect the first orbital flights of starship to be unmanned.

        Personally, I’d put the bar for sucess of crew dragon at ‘taking crew to/from the ISS’, since the Commercial Crew Program is its raison d’etre.

      • cold_potato says:

        No, Starship hasn’t reached orbit. Prototypes have been built (“Starhopper”, Mk1-4, SN1-4) and have made test flights up to an altitude of 150m. But Starship is only the upper stage, depending on the lower stage (“Super Heavy”) to put it in orbit, and Super Heavy still exists only on paper, as far as I can tell.

        I’ve seen an article in Ars Technica stating that SpaceX has an aspirational goal to get Starship in orbit before the end of 2020. But I strongly suspect that this is a misunderstanding, and the actual goal is to get Starship to space, not orbit (to an altitude >100km, but not going sideways at 8km/s) which is much easier, and within the capabilities of Starship on its own, without Super Heavy.

        I would put the probability of Starship reaching orbit by the end of 2020 at 2%, mostly associated with the possibility that I’ve misunderstood something fundamental.

        Your estimated probability of 80% for Crew Dragon reaching orbit by the end of 2020 seems reasonable to me, provided it’s interpreted to mean “Crew Dragon reaches orbit while actually carrying crew”. (There was an uncrewed test flight of Crew Dragon in March 2019.)

      • gbdub says:

        Mea culpa. I mixed up Starship and Starliner (the Boeing crewed capsule). Starliner has flown – although the flight was a partial failure, it did reach orbit.

        The chance of StarSHIP reaching orbit this year is zero, which may be why I mixed them up – your numbers make more sense for StarLINER.

        • Lambert says:

          For starliner, the question isn’t whether it makes orbit, it’s whether it gets back.

          During the abort test, only two of the three parachutes deployed properly. And they discovered and fixed a potentially lethal bug in orbit while they were dealing with the clock bug.

          • gbdub says:

            And of course the Crew Dragon capsule that went to space blew up on the pad due to a valve failure in a thruster test (which would have been a 100% lethal failure) and SpaceX has had issues with parachutes as well.

      • gbdub says:

        Actually the whole thing might be badly formed… “Starship” CANNOT reach orbit, it’s a suborbital vehicle. It needs the “super heavy” booster to get orbital, and I don’t think even Elon is claiming that will be ready this year.

        • cold_potato says:

          It’s still well-formed, I think. Starship can reach orbit with the assistance of Super Heavy, in the same way Crew Dragon can reach orbit with the assistance of Falcon 9.

          But I think you’ve identified an important point of confusion: Scott may well have meant to say Starliner rather than Starship.

          • gbdub says:

            It’s well formed in the sense that it can eventually happen. It’s a bad prediction in that any plausible test flight of Starship this year will certainly be suborbital.

  4. gbdub says:

    I have to say I am disappointed in your Biden / Reade predictions. On the one hand, you’re very sure that the evidence will point to Biden being guilty of a serious sexual assault. On the other hand, you’re very sure you will vote for him anyway.

    You are a resident of California. There is no plausible world in which your vote for Biden will be decisive (if CA is in play, Trump is winning in a relative landslide). There is no good reason for you to vote for someone you believe to be a sex criminal, even if you believe that Trump is legitimately worse. Just leave it blank, or write yourself in, or whatever.

    • Scott Alexander says:

      I’m in California, so nothing I do matters at all, whether that’s endorsing Biden or disendorsing him. I should probably vote for president just because I’ll be filling out the ballot anyway to vote for downballot offices. It seems dishonest not to vote for Biden if I actually want him to win. I’m considering voting for Justin Amash, the new Libertarian candidate, but I don’t like most of his policies. If my vote sends some sort of meaningless signal, I’d rather signal Justin Amash not to oppose the Paris Agreement than signal Biden not to assault people. This seems especially true since nominating an anti-fixing-climate-change-person was a choice that the Libertarians can avert, whereas nominating a sexual assaulter seems to have been an accident on the Democrats’ part.

      • sty_silver says:

        Couldn’t you trade your vote for someone who will vote Biden in a swing state? I believe that’s what EY did last time (for Clinton), partially so that he could have a good answer for “what did you do to fight Trump?”

      • Bugmaster says:

        This sounds like the classic coordination failure / Prisoner’s Dilemma problem.

        If everyone agreed to vote for the best candidate, then neither Biden nor (arguably) Trump would be elected. However, most people would vote for literally anyone or anything just to make sure the opposition doesn’t win. Therefore, voting for anyone but the official party candidate is pointless… and thus the cycle continues, since there is no way for people to signal things like, “we’d rather prefer someone who is not a sexual assaulter, thanks”.

        I used vote in national elections, but I don’t anymore — it just seems like a total waste of time. I don’t see any point in sending those “meaningless signals” that are not being received by anyone who matters.

        • EchoChaos says:

          That’s what primaries are for.

          • deciusbrutus says:

            What fraction of people were eligible to vote in a primary election with more than one presidential candidate?

        • Edward Scizorhands says:

          Voting is speech. Probably the ultimate speech. Everyone decides for themselves how they want to do it.

          • Bugmaster says:

            In most places in the country, voting for the President is speech that no one will ever hear.

            (voting in local elections is still meaningful, though)

        • @Bugmaster
          At a certain point it would be better to skip all the drama and fighting, and have the government switch between a right leaning policy sheet and a left leaning policy sheet every 8 years (since two term Presidents are more common anyway), and then we get a bunch of pre-selected technocrats, filtered from a random pool of STEM and economics PhDs to act in place of the President. They will have ideological leanings of their own, though they will be filtered to be relatively centrist and pragmatic. Their leanings will give the government an overall flavor and direction within the constraints of relative blandness.

          Yes, new groups of technocrats will act to undo the work of the previous group after 8 years, but that’s sort of the point, as that already occurs now, only we won’t be wasting the public’s time, and having them getting riled up into ideological battles that lead to terrorism, when the outcome is more or less Democrats following Republicans following Democrats forever, with a few wiggles along the way.

          The people instead elect a head of state whose job is nothing more than relaying the technocrat’s decisions to the public and other heads of state, people selecting him neither on ideology (since the outcome is almost exactly the same anyway), nor competence (since competence has nothing to do with a popularity contest), but instead on the basis of how much they like their personality, looks, criminal record etc. Just like today, only completely disconnected from the important stuff.

      • elspeth artemis says:

        i’m curious to know what you would think of johannon ben zion.

        • Scott Alexander says:

          I looked him up very briefly on your recommendation. It looks like he wants to give every citizen a $52,000 UBI, funded mostly by renting out land in national parks, which he says will earn $17 trillion per year. The entire US GDP is only $20 trillion per year, so unless he thinks renting out national parks will double GDP, I don’t know what his deal is. I conclude he is unserious and his ideas make no sense.

      • gbdub says:

        To be clear, I don’t fault you for supporting the Biden batch of likely policies over the equivalent policies of Trump and Amash.

        But the President of the United States is not just a policy paper, he’s a person chosen to lead the most powerful nation in the world. Some things ought to disqualify you for that role, if we are to pretend we have any values at all other than naked partisanship.

        And if ANYTHING can disqualify you for that role, using your position as a Senator to sexually assault a subordinate and keep her shut up about it for the better part of 3 decades ought to be on that list.

        The Democrats may not have been aware of Reade per se, but they knew they were nominating a career party man who by sheer longevity almost certainly had some skeletons in his closet. Not to mention that he had a long history of foot-in-mouth disease and seems to be rapidly losing his mental faculties.

        A system that leaves us a choice between Trump and Biden is a very broken one, and voting for either is, in a small way, endorsing that system. You can’t really justify it on consequentialist grounds (again, California). So I would urge you to consider morally objecting to the broken, hyper polarized mess by writing in “none of the above” if you can’t find a third party to your liking. It’s certainly what I’ll be doing.

        • keaswaran says:

          I’m interested in this point: “The Democrats may not have been aware of Reade per se, but they knew they were nominating a career party man who by sheer longevity almost certainly had some skeletons in his closet.”

          Is this a general argument against voting for anyone like Bernie Sanders or Hillary Clinton or John Kerry or Amy Klobuchar and in favor of people like Pete Buttigieg and Donald Trump who are relatively new to politics?

          • gbdub says:

            I don’t think it is general argument against running career politicians. But if you’re a party with a strong progressive wing, nominating somebody who was a main streamer 30 years ago with no real progressive cred is dicey. Add in that he’s a white man known for gaffes and it gets worse. Add in that one of the things that has shifted most strongly in the last several years is “what powerful dudes can expect to get away with re: sex and subordinates”…

            The DNC May not have known about Tara Reade specifically, but it’s the sort of thing they should have been working very very hard to vet Joe for.

            In other words, I agree with Scott that this was an “accident” by the Dems, I’m just saying there is a certain degree of negligence that enabled that accident – somebody like Biden has a much bigger risk of this sort of thing coming to light than somebody like Mayor Pete (or Barack Obama for that matter).

          • Jaskologist says:

            It’s certainly an inversion of the old conventional wisdom, which was that a career politician was the safe choice, since they’ve already had multiple opponents dig through their dirty laundry. There shouldn’t be any surprises left, relative to some young unknown who may have avoided scrutiny until recently.

          • Clutzy says:

            Longtime career politicians have not been the safe choice for a long time. There have really only been 2 DC insiders to win elections in the postwar era: Johnson (as an incumbent) and Bush I. Neither could win a 2nd term.

          • Michael_druggan says:

            ” Add in that he’s a white man known for gaffes and it gets worse”

            What’s with the blatant anti White racism?

          • Aapje says:

            @Michael_druggan

            That seems to be a recognition that anti-white racism is substantial in the Democrat party, not anti-white racism itself.

          • gbdub says:

            I don’t think it should be controversial to recognize that identity politics play a significant role in today’s Democratic Party. As such, running a straight old white guy risks turning off some of your most enthusiastic voters, who are going to be less inclined to overlook any faults in a candidate that checks zero underrepresented identity boxes.

        • and keep her shut up about it for the better part of 3 decades

          Perhaps I missed it, but I haven’t seen any evidence that Biden or his staff pressured her to keep her mouth shut. Her statements seem to imply that she didn’t want to go public.

          • gbdub says:

            She told several people contemporaneously to the event and sounds like she was probing for what could be done, and even made some sort of complaint that was ignored.

            “Not wanting to go public” sounds more like, at best, “was discouraged that she’d get anywhere going public and her claim would not be supported”

            Which is not quite as bad as “actively quashed her complaints / paid hush up money” but, especially if you’re a supposed supporter of #metoo the implied pressure against her is almost as bad.

          • Edward Scizorhands says:

            30 years ago lots of people were implicitly discouraged from coming forward. Because what would happen if you came out against a powerful figure? Ha ha, that’s right.

        • IvanFyodorovich says:

          I’ll bite. Kennedy, LBJ, and Clinton probably did comparably bad stuff all toward subordinates. They were all pretty good presidents. Carter was a saint and a flop. I just don’t think this stuff predicts president-quality.

          Re: the certainty of skeletons in his closet, that’s kind of an odd statement. Either this is exceptionally bad behaviour that shouldn’t be expected of anyone, or you are indirectly claiming that most powerful men of a certain age have done comparably bad things.

          • ChrisA says:

            OK but then you have to apply that same logic to anyone on the other side of the political fence. Also just how bad is bad enough not to vote for. I must admit sexual assault crosses my threshold. But what is yours?

          • Doctor Mist says:

            I must admit sexual assault crosses my threshold.

            Mine, too. But that position has to be tempered. First, what if one candidate is a sexual assaulter but the other is a child murderer? And second, we are none of us omniscient: Does a 50% chance that the candidate is a sexual assaulter cross your threshold?

          • that the candidate is a sexual assaulter

            How about “that the candidate once committed a sexual assault a long time ago”?

            In the case of Kavanaugh, the claim was that he attempted sexual assault as a drunk teenager. In the case of Biden, it’s that he committed one, pretty clearly in the mistaken belief that what he was doing was desired, almost twenty years ago.

            I wouldn’t take either of those, even if the claim were true, as a strong reason to refuse to appoint Kavanaugh to the court or vote against Biden.

      • Jaskologist says:

        whereas nominating a sexual assaulter seems to have been an accident on the Democrats’ part.

        If it turns out that she did make a complaint to the Senate which was ignored, this was not an accident, but rather a systemic suppression of sexual harassment claims.

        And I think that part very likely, and very bipartisan. Biden’s time in the Senate has included the “waitress sandwich”* by 2 fellow Democratic Senators, and Dennis Hastert’s (Republican Speaker of the House) molestation of boys, Bill Clinton’s… everything. There are surely many more that are not off the top of my head.

        * And if you want to get a grip on just how systemic such silence is, go look up “waitress sandwich” and make note of the political alignment of the top results, and by extension, the political alignment of those who choose silence.

      • Tatterdemalion says:

        I’m in California, so nothing I do matters at all, whether that’s endorsing Biden or disendorsing him.

        I don’t think this follows. You’re in California, so who you vote for for president doesn’t matter a hill of beans, but you have a large, widely-distributed readership, so I suspect that your endorsement or disendorsement will sway votes in swing states too.

      • deciusbrutus says:

        The most (by such a margin that it might as well be the ‘only’) important elections this year are for state legislatures, since the 2010 census will bring new congressional districts, which will be determined by the state legislature, and the current court rulings are that gerrymandering is perfectly legal so long as it’s purely partisan and not racist.

    • John Schilling says:

      I could use about 90% fewer “you’re a hypocrite if you vote for Biden after hearing about Tara Reade!” takes from people who had already decided to vote for not-Biden before anyone had ever heard of Tara Reade. And 80% less back-and-forth discussion of the takes that even are worth making.

      • gbdub says:

        I’m not calling Scott a hypocrite. I’m saying that I am disappointed he would vote for someone he believes to have committed a serious sexual assault (and not only that, one that involved abusing his position as an elected official over a subordinate).

        If he honestly believed Trump, or anyone else, was a serious sex offender and was 80% sure he’d vote for them anyway, I would also be disappointed.

        I don’t think voting for Biden, if you believe he is guilty of assaulting Tara Reade, is inherently hypocritical, I think it is inherently wrong.

        And, incidentally, I had not “decided to vote for not-Biden”. I am in a sorta-kinda swing state, and really don’t want to vote for Trump. I also really didn’t want to vote for Bernie Sanders, and pre-Reade-looking-very-plausible definitely supported Biden over Bernie. How the Dems and Biden handle this issue really will have an impact on whether my vote is “Biden” or “Libertarian / none of the above” this November.

    • grendelkhan says:

      Much as the candidates both oppose chattel slavery and thus, if you care about chattel slavery either way, it shouldn’t factor into who you vote for, both candidates are old, white, full of malapropisms and have been credibly accused of sexual assault. So if those things matter to you in either direction, they can’t factor into your decision here. Go to iSideWith, fill out the form, and vote for the high score like an adult.

      Also, it would be excellent if we could get rid of the idea that voting carries some kind of moral stain or glow. You’re imparting one literal bit of information to the political process; that’s all. I’m with Michael Sweeney on this one.

      • gbdub says:

        You are assuming a binary choice. Yeah, end of the day one of them is going to win. But you can choose to spend your bit of information signaling “I approve of neither of these”. Voting “present” is a valid choice.

        • grendelkhan says:

          If they get equivalent scores on ISideWith, sure, you can do that. But if one is clearly superior to the other, you’re back to making category errors about what voting is for.

        • deciusbrutus says:

          Only if you claim that you can’t influence the victor.

      • So if those things matter to you in either direction, they can’t factor into your decision here.

        Sure they can.

        Prior to the Reade story, there was some reason to believe that Trump was guilty of sexual misdeeds, no reason to believe that Biden was. For some voters, that was a negative for Trump. Add in Reade and that negative disappears, which could rationally make someone who had been mildly inclined to vote for Biden change his mind.

        • grendelkhan says:

          I think we’re saying the same thing in two different ways–the contribution from those issues is now zero, which might indeed make for a net change in your opinion.

          Is it weird that I’m this excited about the prospect of more people voting solely on policy issues?

          • Robert Beckman says:

            I don’t think the contribution from the sexual assault factor is now zero, for those who have looked at the underlying facts rather than how they were often framed in the media at the time.

            For instance, assuming all plausible allegations are true (you may disagree with my assessment of plausible, or know of examples that aren’t coming to mind):

            Joe Biden, a powerful man: committed sexual assault (and rape under some state statutes, I don’t know the relevant at the time) 25 years ago

            Donal Trump, a powerful man: blew the whistle on the systematic ability of powerful men to commit sexual assault 20 years ago

            A rational person could easily see this one strongly weighing in Trumps favor. Of course, a Bayesian could also conclude that since he’s in the very category he says commits sexual assault and is so forthright about it that the odds he did so, perhaps many time, are higher than the one plausible claim against Biden, and so this weighs against Trump.

            Please recall that Trump did not say he had indeed “grabbed women by their pussy” (as widely falsely reported), but only that powerful men could readily do so (and implicitly get away with it), which is essentially Reade’s very claim.

      • MoebiusStreet says:

        Is “iSideWith” even relevant in deciding on a Presidential candidate?

        For a congressperson it’s one thing, because their job is to determine policy. So it’s pretty important to understand their leanings.

        But the President’s job isn’t to create policy (except in foreign relations). Their primary job is to act as a manager, being the executive to enact the policies that have already been decided by Congress. So why are the candidate’s personal predilections important, at least relative to their qualities as an executive?

        I’ll go even a step further, and question it for congresspeople as well. I find that there’s a lot of space between the platform that a candidate espouses publicly – especially while campaigning – versus that revealed by their actions. To the degree that iSideWith is basing their recommendations on that public policy sheet, which I assume is most or all of their data, they’re going to be off by the same amount that the candidate is obfuscating their real positions.

        • anonymousskimmer says:

          So why are the candidate’s personal predilections important

          SCOTUS
          Executive orders
          Veto potential (partisan vetos have almost no chance of being overridden)

    • I can’t speak for Scott, obviously, but I don’t think the accusation, if true, is a reason not to vote for Biden, or even to think he is a wicked man.

      There is a large difference between someone who attempts a sexual act and continues after the other party has made it entirely obvious that it is unwelcome and someone who attempts a sexual act in the belief that it is welcome and stops as soon as it is clear that it isn’t. The latter isn’t evidence that the perpetrator is a bad person but that he is socially incompetent, doing a bad job of reading the other person’s signals. The normal pattern of courtship as I understood it when I was of a relevant age, which is about when the incident occurred, was that you found out how far the girl was willing to go by attempting the stage after what she and you were currently doing and seeing whether she went along.

      What Biden is accused of doing and Trump boasted of doing is pushing that approach unreasonably far, but that doesn’t change the essential distinction.

      • theredsheep says:

        When you’ve gotten to the point where you assess initial interest by putting your hand on or in a woman’s vagina, you’ve skipped about eight steps. If Biden sincerely thought it was reasonable to approach a subordinate with whom he had had a wholly professional relationship up to that point, kiss her, then immediately progress to digital penetration–like that’s just how courtship manners worked in the nineties–then the sexual-assault charge is more of a catastrophically-terrible-judgment charge, which arguably makes him a still worse candidate for present.

        • then the sexual-assault charge is more of a catastrophically-terrible-judgment charge

          Yes. That was my point.

          which arguably makes him a still worse candidate for present [president].

          Arguably yes.

      • thisheavenlyconjugation says:

        Is a man who rapes a baby also not necessarily wicked if he has genuinely held albeit very misguided ideas about consent?

    • zardoz says:

      If you believe the mainstream media (NYT / WaPo / etc), Trump is ineffective and you should not vote for Trump.

      However, given a choice between two evil rapists, it is better to pick the ineffective one rather than the effective one. Therefore you must vote for Trump. On the other hand, if you believe the mainstream media is lying, then you should do the opposite of what they suggest, which also means voting for Trump.

      Therefore, it is 100% mathematically proven that you must vote for Trump. Q.E.D.

      You can’t argue with mathematics.

  5. willard337 says:

    What is the value in predicting the US has the largest death toll? The US is the 3rd largest country in the world. The first (China) will not report accurately to deceive. The second (India) will not report accurately because accuracy costs and they don’t have the funds to properly track.

    The figure to look at is deaths per capita. In that regard, the US is handily beating the EU. The EU deaths per capita is 50% higher than the US. We have states (TX, FL, CA) that are handily beating even Germany. Those 3 combined are larger than Germany.

    • doubleunplussed says:

      Though most of Europe has declining new case numbers whereas the US is seemingly in a long plateau – plenty of death still to come in the US.

    • Robert Jones says:

      To 28 April, EU has 226 deaths per million people vs 178 for the US, which is 27% higher. If the UK was in the EU, the EU average would rise to 239 deaths per million, but that’s still only 34% higher than the US.

      Incidentally, I find it remarkable how a map of Europe by Covid19 deaths divides so clearly along the line of the Iron Curtain.

      • Deej says:

        Could be travel and/or more cosmopolitan nature of western Europe?

        Similar to how cities like London and New York have tended to be hit worse and earlier.

        • Robert Jones says:

          I’m sure that’s right, but the Warsaw Pact existed for 36 years and ended 29 years ago. I find it surprising that the Warsaw Pact countries haven’t become more integrated with the rest of Europe, particularly those which are now EU member states.

          • Anteros says:

            How about differences in measuring, attributing, and testing?

          • viVI_IViv says:

            Well, in the last 29 years, how many Brits do you think immigrated to Poland? How many Italians immigrated to Romania? How many Chinese people travel to Paris as opposed to Sofia?

          • matkoniecz says:

            I’m sure that’s right, but the Warsaw Pact existed for 36 years and ended 29 years ago.

            Partitions of Poland lasted for 123 years, ended 102 years ago and still have clearly visible effects. See https://cdn.static-economist.com/sites/default/files/20181124_WOC902_0.png

            (result of Googling “partition of Poland election map”)

          • MilesM says:

            I visited Poland last summer, and even so long after the borders opened up it can be striking how homogeneous it is compared to Western Europe or the US.

            Only the busiest parts of Warsaw really had a noticeable population of obvious foreign nationals (and it was the only city where I saw significant numbers of non-white people) but it was still just a fraction of what you’d expect in Western capitals.

            Gdansk, a major tourist destination, had some people speaking English and German, but was virtually 100% white.

            The small city I was born in might as well not have changed at all since the time I lived there, nearly 30 years ago. (although there actually have been some changes, like a recent influx of Ukrainian nationals – but they’d essentially be invisible when using my “sampling method” of walking around looking at people)

        • DarkTigger says:

          Well I believe Putin’s Russia only marginally more than I believe China. Same goes for Belorussia.

          And than: Looking at this graph (https://de.statista.com/statistik/daten/studie/249029/umfrage/urbanisierung-in-den-eu-laendern/ sorry German) the former Warschau Pact states seem to be less urbanized. Which would explain a slower spread of the Virus.
          Germany is also less urbanized than many other West European countries, which is the reason why we have many small localized hospitals, which offer a buffer against the break down of the health care system.

          • zqed says:

            The hardest-hit Eastern EU country, Slovenia, had 44 deaths / million pop. In Western Europe, only Finland and Norway had fewer deaths / million pop. and 90% of WEU currently have more than 100 deaths / million pop. That’s a very clear difference. The difference in urbanization is much much smaller, about 10% higher for WEU.

            If urban population explained a large chunk of the observed difference in deaths between EEU and WEU, we’d still expect somewhere in the range of 100 deaths per million population in the EEU. Instead, we have the following (deaths taken from Worldometers today, urban population data taken from Wikipedia) for the EU27 countries:


            EEU (joined 2004-):
            mean deaths (95% CI) = (10 to 23)
            rho(death,urban_pop) = 0.003317551

            WEU (joined pre-2004):
            mean deaths (95% CI) = (139 to 353)
            rho(death,urban_pop) = 0.251358229

      • keaswaran says:

        “I find it remarkable how a map of Europe by Covid19 deaths divides so clearly along the line of the Iron Curtain.”

        Do you have any such map available? I thought the big divide was between Germany/Scandinavia and Italy/Spain/France/UK.

        • Robert Jones says:

          Map at the bottom of the page here (also with EU/US comparison graph at the top of the page).

          • keaswaran says:

            Interesting – when I slide this to March 30, I get a clear divide with Germany and Scandinavia on the low side and the rest of western europe on the high side. When I slide it to April 17, you get something close to the iron curtain (but with Estonia and Slovenia on the west), and at April 26 it starts looking like it’s something about the Balto-Slavic languages (since Hungary, Romania, Turkey and Estonia are all on the western side). Today it starts to show Czechia and Serbia moving towards the high side.

            There does seem to be a strong correlation with the former Iron Curtain border, but there also seem to be linguistic barriers (Romance>Germanic>Uralic>Balto-Slavic) and others as well.

      • Pandemic Shmandemic says:

        Less testing, probably no post-mortem testing of previously undiagnosed people and an older population that grew up knowing that anything that can be plausibly weathered at home is worth avoiding a hospital visit for.

      • A1987dM says:

        Older population in Western Europe?

      • yodelyak says:

        Some regional differences go back a *very* long way. See e.g. the SSC post on Albion’s Seed, and also here: https://marginalrevolution.com/marginalrevolution/2019/11/did-the-medieval-church-make-us-weird.html

      • cevapcici says:

        Could be BCG use.

  6. eyeballfrog says:

    61. Koios has said his first clear comprehensible word: 50%

    Who’s Koios?

    • Scott Alexander says:

      A baby I live with.

      • Vorkon says:

        Are they named after the character from Brent Weeks’ Lightbringer series, by any chance? That’s the only place I’d ever heard that name, and assumed it was just made up whole cloth, like so many fantasy names.

        • Deiseach says:

          Are they named after the character from Brent Weeks’ Lightbringer series, by any chance?

          Greek mythology, one of the Titans, and the spelling threw me because I’m more used to seeing it spelled Coeus.

          • Lambert says:

            Predictions for 2053:
            Titanomachy: 70%
            Olympian Victory: 30%
            Kronos condemned to Tartarus: 10%

          • Robert Beckman says:

            Better prediction for 2053:

            Tiberius restores the Empire: 60%
            Imperium expands beyond Earth-Moon: 30%
            Colonist Ship launched: 5%
            Imperator Tiberius overthrown by Empress Temperance (includes joint imperium): 20%
            Imperium lost by both of the Twin Emperors (via plebiscite, conquest, or revolt): 2%
            Imperium converted to a monarchy: 2%

            Obviously some of these are dependent on others happening, and I intend them as conditionals: IF my son has restored the Empire by 2053, THEN there’s a 20% chance his twin sister has at least seized his power as her own, if not dethroned him.

            These are, of course, essentially predictions on how well I raise my own children, and hence are highly susceptible to the planning fallacy, but we had a few examples in the last century on how to do it, and my children don’t suffer from racial bigotry (themselves recently deriving from every continent save Antarctica, including Oceania), and after the coming war it’ll be easy enough to regroup the Americas against the spreading Chinese Hegemony, as a single entity spanning from Moscow to Australia would be a bit terrifying if you’re not the one in control (and if you are, you’re not the sort of person who would be terrified of it anyway).

            🙂

    • spinystellate says:

      61. Koios has said his first clear comprehensible word: 50%

      Koios, interrupt us!

  7. Ketil says:

    25. …greater than credence in Kavanaugh accusation asked in the same format: 40%

    Kavanaugh was accused out of nowhere by hard-nosed political opponents with no contemporaneous corroboration, accusations against Biden is confirmed by at least three people who claim to have heard it first hand, and who mostly(?) are solid Democrats planning to vote for Biden even so. Also, Reade was allegedly assaulted at work during the daytime, while Ford claims to have been assaulted at some drunken college party – with implications for reliability of accounts. I don’t see how you could consider the Kavanaugh accusations more credible, except by having extreme priors (a.k.a. prejudice) against Republicans. Anybody care to strongman this?

    I can accept differing views on how much unproven and decades old allegations should count, but anybody who thinks Kavanaugh should be disqualified on the grounds of the allegations yet votes for Biden is a hypocrite, cares less about women than about political gains, and can probably not be trusted to tell the truth.

    • CaptainBooshi says:

      Just so you know, I think the disconnect seems to come from some of your information just being not true:
      -Neither Ford nor Ramirez were “hard-nosed political opponents”, they were private citizens with basically no connections to politics at all
      -It’s literally on the record that Ford was talking in 2012 to a couple’s therapist about being assaulted by boys who become high-ranking people in Washington, the notes from back then were made public
      -A bunch of people heard about the Ramirez incident from literally the next day, I’m just going to quote the Atlantic article that came out last September, found here:

      At least five people have a strong recollection of hearing about the alleged incident with Ramirez long before Kavanaugh was a federal judge. Their Yale classmates Kenneth Appold and Richard Oh recall hearing about it immediately after it happened. Ramirez’s mother, Mary Ann LeBlanc, recalls being told about it by her daughter—without specifics—during Ramirez’s college years. Michael Wetstone, a graduate-school classmate of Appold in religious studies, recalls being told about the incident by Appold within a few years of when it allegedly happened. A fifth person—an unidentified friend of Ramirez’s who said in a recent affidavit that she’d heard about the incident in the 1990s—remembers being told about it within a decade of its alleged occurrence. And two other people from Kavanaugh and Ramirez’s Yale class, Chad Ludington and James Roche, vaguely remember hearing about something happening to Ramirez during freshman year.

      I don’t think there’s any point in debating the actual event here, or the evidence for/against, but your specific claims about the credibility of the accusations just seem outright incorrect.

      • viVI_IViv says:

        Neither Ford nor Ramirez were “hard-nosed political opponents”, they were private citizens with basically no connections to politics at all

        Neither were part of his staff, for sure. In fact, there is no evidence that they had ever interacted, the closest connection that can be proven is that they attended the same schools at the same time, with thousands other students.
        Kavanaugh, at the time of the of the alleged facts, was a student just like to his accusers, he had no authority over them so there is no reason for them not to speak up for decades and then speak up when he was nominated for an effectively political office.

        Moreover, Ford completely scrubbed her social media before making the accusation (I wonder what she had there), and she first made the accusation mentioning Kavanaugh in a private letter she sent to Dem senator Dianne Feinstein (ref).

        -It’s literally on the record that Ford was talking in 2012 to a couple’s therapist about being assaulted by boys who become high-ranking people in Washington, the notes from back then were made public

        These notes did not mention Kavanaugh and mentioned four boys unlike the later accusation.

        A bunch of people heard about the Ramirez incident from literally the next day

        Again, no mention of Kavanaugh.

      • Ketil says:

        Neither Ford nor Ramirez were “hard-nosed political opponents”, they were private citizens with basically no connections to politics at all

        I’m talking about how Ford was obviously weaponized and used tactically by leading Dems in the case against Kavanaugh.

        -It’s literally on the record that Ford was talking in 2012 to a couple’s therapist about being assaulted by boys who become high-ranking people in Washington, the notes from back then were made public

        This is still not contemporaneous, and not particularly specific. If you know how easily memories can be changed and the horrible track record of psychotherapy, I don’t see how this is the same as a staff member who tells her closest confidantes at the time it (allegedly) happened.

        You are not addressing at all drunken students in college parties vs sober staff member at work. This makes it more likely that something happened, but much less likely that the accounts are accurate – including the names of people involved.

        The fact that there are multiple accusers and events strengthen the case against Kavanaugh, and the supportive testimonies from women on his staff weakens it (arguably in contrast to Biden’s “touchyness”, but I don’t count that for much)

        • Aapje says:

          The fact that there are multiple accusers and events strengthen the case against Kavanaugh

          The Swetnick allegation began as an accusation of gang rape and spiking drinks, but upon a little probing by the media, was immediately downgraded to an accusation that Kavanaugh stood in a group and handed drinks to girls. At that point, she didn’t actually claim to have witnessed any crime by Kavanaugh. She speculates that they were part of a large scale rape conspiracy that she believes existed, but to which no witnesses have been found (even though this would have involved many drugged and raped women & many rapists). Note that Swetnick was also found to have falsely accused colleagues at work of sexual misbehavior.

          Neither the two other allegations involve rape and the weakest part of both allegations is whether Kavanaugh was the perpetrator. Ramirez reached out to potential witnesses, where she herself said to some of them that she was unsure that Kavanaugh was the perpetrator.

          All the witnesses spoke out against Ford’s accusation, including her own friend, who, according to Ford was at the party where Ford was allegedly assaulted, but who recalls no such thing or ever meeting Kavanaugh.

          The single accusation by Reade looks more convincing to me than the three accusations against Kavanaugh combined. Reade seems to have told witnesses shortly after the alleged incident and was specific then that her (ex-)boss was the perpetrator, who presumably could be no one but Biden. The accusation is also far more serious (if we dismiss Swetnick’s rape gang conspiracy). It is an actual accusation of rape, where Ford and Ramirez alleged assault.

          • Edward Scizorhands says:

            Let’s figure it out by looking through Biden’s yearbook.

          • Aapje says:

            I looked into Reade a bit more and she seems to have escalated her accusations over time, going from accusing Biden of the things we have video of him doing, to putting his hands under her skirt, to penetration.

            So that makes it a lot less convincing.

          • JRM says:

            Aapje’s followup is important. There’s also this.

            This particular fawning piece on Putin, combined with the changed stories, has to serve as a credibility downgrade, no?

            It’s all probabalistic. I think it’s possible to find the claims against Kavanaugh more credible than these claims without being a partisan hack.

            For a really good, long analysis, see: here.

          • gbdub says:

            Why is the fact that Tara Reade used multiple names on social media relevant to the credibility of her assault claim? Why is her flip flopping on Russia relevant? Why is the fact that she thinks Julian Assange is a “hero” relevant? Why is the fact that her pro Russia article flipped the gender in the Russian translation relevant?

            Seriously your “very good, long” article reads like the rankings of a conspiracy theorist. The inclusion of a bunch of dubiously relevant excerpts from her twitter account and the wink wink implication that she’s some sort of Russian plant seriously detract from the parts that are more relevant, namely whether her story re: Biden assaulting her changed in a meaningful way.

          • m.alex.matt says:

            gbdub: This is really the only thing that matters.

            Sexual assault survivors change their story from time to time. Memory formation in traumatic events can be sketchy, so it can be hard to come up with a coherent narrative around a very real memory in the moment. Trying to re-live the memory to harvest the actual details is also difficult.

            But sexual assault survivors do not go back and deceitfully edit prior publications on the matter to make it seem like the current narrative they’re telling was always the narrative.

            Sexual assault survivors do not carefully time the release of their allegations for maximum political effect and taunt the target of their allegations over the matter.

            Sexual assault survivors do not need to remind one of their corroborators after years of friendship and have one of their other corroborators change the story they told reporters via text message days later.

            This whole thing is what the kids call a ‘ratfuck’.

          • Edward Scizorhands says:

            The Krassensteins are at the top of the list of Resistance Grifters. Literally, right here at the top: https://www.motherjones.com/politics/2019/07/a-quick-guide-to-whos-making-money-off-trump-paranoia/

            Of course they are going to make it about Russia. Ignore them.

          • Byrel Mitchell says:

            gbdub: This is really the only thing that matters.

            Sexual assault survivors change their story from time to time. Memory formation in traumatic events can be sketchy, so it can be hard to come up with a coherent narrative around a very real memory in the moment. Trying to re-live the memory to harvest the actual details is also difficult.

            But sexual assault survivors do not go back and deceitfully edit prior publications on the matter to make it seem like the current narrative they’re telling was always the narrative.

            Thanks for providing that link. I don’t actually think your description (or that of the author of that comparison) is a fair description of the changes. From my perspective, none of those changes make the article particularly more compatible with the sexual assault accusation.

            They all look like simple editorial changes to punch up the emotional impact a bit, with very little difference in questions of fact.

          • gbdub says:

            Byrel Mitchell beat me too it. Editing an already published article without notice is not a good look. But the actual changes noted in the linked article are weak sauce – the most impactful by the author’s own standard are adding “only” to “this is not ONLY a story about sexual misconduct”, added a note indicating that not everything Joe Biden had done to her had come out publicly yet, and changed, “I resigned, well really I was forced to resign” to “I was fired, well really I was forced to resign”

            I don’t think any of those are “changing her story.” They certainly don’t directly contradict her more recent accusations. The first two are consistent with someone who was not originally willing to come forward with a specific allegation of sexual assault, but now is. That’s pretty much anyone who makes an accusation of an old event, and calling that decisive would require dismissing basically the entire collection of #metoo allegations offhand. The third change is really just a rhetorical flourish, in both cases the implication the reader is supposed to take away is that her departure was only nominally voluntary.

            If that’s “the only thing that matters”, I am unconvinced. It certainly doesn’t increase Tara Reade’s credibility that she made those changes. On the other hand, if that’s the kind of thing being trotted around as a smoking gun of “she changed her story, not credible!” then that article does make me much more skeptical that the people attempting to discredit her are doing so in good faith.

          • Edward Scizorhands says:

            Yeah, my Bayesian estimates that Reade is deliberately lying went down when I see the claim that the best argument is . . . minor changes to an old essay.

            I don’t think Reade’s accusation (or Ford’s) should be enough to sink someone. But I think some people have painted themselves into a corner where the only path out is to destroy the woman making an accusation, and that path is something those same people would have called a shameful strategy just a few years ago.

      • Vorkon says:

        -It’s literally on the record that Ford was talking in 2012 to a couple’s therapist about being assaulted by boys who become high-ranking people in Washington, the notes from back then were made public

        It’s been quite a while since I read this, since this issue has been out of the public consciousness for quite some time, but my understanding is that while the records of that therapy session do include Ford claiming she had been assaulted at some party, they do not include the allegations that the people who did the assaulting went on to high ranking political careers.

        The context of the session was couples therapy, in that she was explaining that she believed the reason her marriage was not going well could be traced back to her having a hard time trusting men due to an unpleasant experience with some boys at a college party, but who those boys were was not pertinent to the discussion at hand.

        There were reports from her husband and friends that later on (sometimes years later than the therapy session in question) she claimed that the boys who assaulted her went on to political careers, but as far as I know there is no record of that from the therapy session.

    • Purplehermann says:

      1. Scott professed a belief about others’ beliefs, though I do think it’s more likely that the commenters will think the more likely accusation is more likely.

      2. Voting for someone despite their personal evils can make sense as a lesser of two evils, BK being made a SJ was a double evil for lefties. Biden is a lesser of two evils.
      In a utilitarian view, the correct action seems obvious

      • viVI_IViv says:

        2. Voting for someone despite their personal evils can make sense as a lesser of two evils, BK being made a SJ was a double evil for lefties. Biden is a lesser of two evils.
        In a utilitarian view, the correct action seems obvious

        If the standards that the lefties use is that their are going to support their own no matter what, but they will capitalize accusations against the right for political gain, then it’s fair to call them hypocrites.

        In a utilitarian view, the correct action seems obvious

        And this is a good reason why act utilitarianism doesn’t play well with policy: in order to perhaps gain a short-term advantage you get a reputation of being untrustworthy. See for instance the “transplant problem“: taking the utilitarian choice might work once or a few times, but eventually people will become suspicious and stop trusting you.

        • leadbelly says:

          If the standards that the lefties use is that their are going to support their own no matter what, but they will capitalize accusations against the right for political gain, then it’s fair to call them hypocrites.

          Exactly. Even if you think senile, “literally nothing will change if I’m elected”, will-lose-to-Trump Biden is the best candidate, any future Democrat accusations of sexual misconduct against Republicans will be dismissed as political wrangling, and justifiably so.

        • Purplehermann says:

          The question here is about voters as far as I can tell.

          In general, the standard breing “we will always look at the situation and do the best thing ” isn’t necessarily hypocritical, and can lead to different decisions in similar cases if there are other factors too. President Trump is a factor.

          If Biden were a murderer I think the lefties would turn on him, or if the R candidate was an upstanding human that was fairly center likewise.

          As for trust lost, voters probably won’t lose any from utilitarian voting.

          Take into account that many d voters probably agree with the party on this, and the politicians probably won’t lose out much. See other POTUSes who were not super decent human beings.

        • gbdub says:

          One thing is for sure, if the Reade accusations don’t completely fall apart and Biden remains the candidate, I am going to be very annoyed if I ever see any more left leaning media think pieces about “how can Christians possibly support immoral Trump?” Look in the mirror guys.

          • keaswaran says:

            I don’t understand your claim. Does anyone think that Trump is actually less of a sexual assaulter than Biden?

            The claim isn’t that Trump is too un-Christian to be supported regardless of the opponent. It’s that he’s more un-Christian than his opponents.

          • Aapje says:

            This depends on how you define sexual assault, which is a can of worms.

            A lot of deception on this topic happens by invoking images of rape and then actually giving statistics which are dominated by a lighter category of assaults, including someone putting their hands on people’s knees.

            I think that Biden is much more likely to perpetrate unwanted touching than Trump, given that he can’t even stop himself during public events.

          • gbdub says:

            My point is that Dems would be doing exactly what Christians who vote for Trump are doing – deciding that their candidate’s personal moral failings are less important than the fact that the other guy is much worse (in their view) on policy. You can’t say “I don’t understand how Christians could do that” if you do the same exercise of internal justification.

            This is in reference to a lot of “think” pieces tut-tutting along the lines of “how can Christians possibly vote for a known adulterer / generally not very Christian guy”

          • keaswaran says:

            Fair enough – saying “why would Christians vote for such an adulterer” is pretty silly. But at least some people are making the point specifically comparatively – why would people who justify their votes against various Democrats on the basis of un-Christian-ness then somehow go on to vote for Trump over Clinton (or even Trump over Cruz!)

          • Conrad Honcho says:

            Does anyone think that Trump is actually less of a sexual assaulter than Biden?

            Yes, I do.

            Trump has a history as a shameless womanizer, and what the women who came out before the election accused him of were attempts to initiate/escalate sexual encounters with kissing or groping that they turned down, and then he stopped. The Access Hollywood tape described consensual sexual encounters.

            The only other accusations were Ivanka Trump’s which she recanted as an exaggeration during a messy divorce, and E. Jean Carrol whose accusation is…confusing, calling it “not sexual,” and conveniently timed to sell a book. I’m not sure it even counts as an accusation since she’s extremely vague about the actual events.

            While I think Trump is crude and crass and a womanizer, I do not think he’s sexual assaulted anyone. Given the vagueness of Carrol’s allegation and the recantation of Ivanka’s, to my knowledge there are no women in 2020 accusing Trump of forced penetration. I have no idea if it’s true or not, but the allegation against Biden is more severe than any existent allegation against Trump.

          • Vorkon says:

            I think you mean Ivana. If Ivanka had accused him of sexual assault, that would be a WHOOOOOOLE different can of worms…

          • Eigengrau says:

            @ConradHoncho

            Man. “The Access Hollywood tape described consensual sexual encounters” is an absolute hell of a reach. The quote:

            I better use some Tic Tacs just in case I start kissing [actress Arianne Zucker]. You know I’m automatically attracted to beautiful—I just start kissing them. It’s like a magnet. Just kiss. I don’t even wait. And when you’re a star, they let you do it. You can do anything. Grab ’em by the pussy. You can do anything.

            I assume you are hanging a galaxy of good faith on “they let you do it”? At this point you might as well be defending Dennis Reynold’s “The Implication”.

            On the Ivana Trump accusation: she testified in a sworn deposition to a very specific, detailed act of rape committed by her husband. Later, as part of the divorce settlement, a gag order was placed on her barring her from discussing the marriage or divorce without Donald’s approval. The recantation came after the gag order. Finally, her recantation merely said that the act wasn’t rape in a “literal or legal sense”. She never to my knowledge recanted any of the actual details of the incident. This “literal or legal” wording is also the same language used by Trump’s ex-lawyer Michael Cohen when asked about the incident (he also claimed falsely that spousal rape is legally not rape).

            As for E. Jean Carrol, her “not sexual” comment was regarding her common sex-positive feminist belief that consent is an inherent element of sex and therefore rape does not count as sex. This is obvious within the context:

            The word “rape” carries so many sexual connotations. This was not sexual. It just, it hurt[..]
            I think most people think of rape as being sexy.

            To dismiss Carrol’s allegation for its vagueness is both factually off the mark and a double standard. She recalls a significant amount of detail regarding the assault, its exact location, their conversation leading up to it, and what the two of them were wearing at the time. There are also plenty of details which she is very open about not recalling, such as whether it was in fall of 1995 or spring of 1996. Such lapses in memory are expected in recalling something that happened more than 20 years ago, and are consistent with other high-profile rape accusations that happened long in the past, including Tara Reade’s. Finally, Carrol told two friends about the incident immediately after it happened, and they have both corroborated her account. By all measures the E. Jean Carrol accusation is as severe an allegation as the one against Biden.

        • keaswaran says:

          “it’s fair to call them hypocrites.”

          Sure. Everyone’s a hypocrite. I’m happy to embrace that label.

          Debate on the actual issues. Is it bad or irrelevant to support someone who sexually assaults people? Should we support candidates on the right or on the left? Should policy trump personal behavior? Does it matter whether there is an open nomination process that could choose anyone from a pool of dozens or hundred, or whether we are in the final round of an election with only two viable candidates left?

          The claims I would make are that it’s prima facie bad to support people who are guilty of sexual assault, that liberal policies are better than right-wing policies, that policy trumps personal behavior, and that in a larger pool there are likely many more options available that aren’t guilty of sexual assault, unlike the upcoming general election. I don’t actually think this set of views involves any hypocrisy.

          But if one thinks it’s hypocrisy to say that Kavanaugh’s alleged sexual assault is relevant and Biden’s alleged sexual assault is not, then one should say either that both sexual assaults are relevant, or neither is, or else give some very detailed explanation for why the reverse pairing is not itself equal hypocrisy.

          • rumham says:

            How about “The Biden allegations are above Kavanaugh’s on the credibility standard set by opponents during the Kavanaugh hearings. It is hypocritical to dismiss them if you thought the Kavanaugh allegations were credible.”

          • keaswaran says:

            Sure – as long as you don’t claim that voting for someone in a two-serious-candidate choice amounts to “dismissing” allegations the same way that voting to confirm does, when this is a nominee for a seat where any number of other people could easily be nominated (and a whole list of them even was drawn up just weeks earlier).

          • rumham says:

            @keaswaran

            Unless you’re calling on the Democratic party to publicly urge Biden to withdraw before the convention, yes. Any number of people could easily be nominated. The Democratic party has rules in place for this already.

          • keaswaran says:

            “Any number of people could easily be nominated. The Democratic party has rules in place for this already.”

            I still think this is very different from a Supreme Court nominee. When a Supreme Court nominee withdraws, the nomination of the next nominee works exactly the same way the first one did. But when a Presidential nominee withdraws, the nomination of the next nominee is extremely different – sure, it’s still the delegates at the convention voting, but we are now imagining a situation where a majority of delegates are unpledged. Letting the choice of nominee come down to the choice of unpledged delegates is exactly the divide that has been tearing apart the Democratic party for the past 5 years or so. It certainly could be done, but there’s no good mechanism for how delegates should make their decision, since the delegates were never empowered to make a decision – just to vote for the candidate their constituents supported.

        • Sandpaper26 says:

          Not to “no true Scotsman” this, but no leftist I know is planning on voting for Biden. We’re pretty united in the Howie Hawkins camp. It’s only Liberals who think they’re on the left (because of how far-right American politics is) who are considering Biden.

          • viVI_IViv says:

            Only if your definition of leftist is tankie.

          • Sandpaper26 says:

            @viVI_IViv

            Hawkins is the Green Party candidate. Unless you consider the entire Green Party (as well as the Socialist Party with which he is also affiliated) to be “tankies,” then you are wrong. If you do consider both of those groups to be “tankie” organizations, then I think you have a skewed perspective of Leftist ideologies. “Tankie” doesn’t just mean non-Anarchist anti-capitalist.

          • EchoChaos says:

            @viVI_IViv

            I am the furthest thing from a lefty, but @Sandpaper26 is backed up by polls on this one.

            https://www.usatoday.com/story/news/politics/elections/2020/04/29/bernie-sanders-supporters-not-yet-board-voting-joe-biden/3047485001/

            Biden is getting nowhere near the support from Bernie supporters that Clinton was.

            Now, there could be multiple factors about this that don’t actually hurt Biden (fewer Bernie supporters, they will return during the campaign, he gets more moderates than Clinton), but @Sandpaper26’s assertion seems valid to me.

      • Ketil says:

        2. Voting for someone despite their personal evils can make sense as a lesser of two evils, BK being made a S[upreme court] J[udge] was a double evil for lefties.

        Sure. But you can’t simultaneously claim that BK is unacceptable because of the allegation and Biden is not, unless a) the accusations against BK is much more credible, or b) the alleged event is much more heinous.

        I am hard pressed to see how one could reasonably arrive at a) and certainly not at b). A person of authority assaulting a subordinate in broad daylight is a completely different kettle of fish than some college frat boy placing his junk in some drunken sorority girl’s hand to gross her out.

        I would probably vote for Biden, clinging to the hope that the allegation isn’t true. But I wouldn’t simultaneously clamor for Kavanaughs dismissal on much weaker grounds.

        • thisheavenlyconjugation says:

          A person of authority assaulting a subordinate in broad daylight is a completely different kettle of fish than some college frat boy placing his junk in some drunken sorority girl’s hand to gross her out.

          Please elaborate. I’m especially curious as to the relevance of “drunken”.

          • Deiseach says:

            I’m especially curious as to the relevance of “drunken”.

            I think the relevance of it here is that everyone at this party where the incident was alleged to have happened was drunk, and the accuser (though that’s rather a strong term to name her) admitted she took six days and ringing up her friends to refresh their memories before she could be sure “yeah it was Brett who flashed me, not some other guy, and it definitely was a guy pulling his dick out and not a dildo or something”.

            In short, the circumstances were fuzzy, the credibility level was low, due to the lack of sobriety on the parts of all and the effect of that on memories and inhibitions.

            “Stupid shit drunk late teens/early twenties get up to because they’re drunk and think it’s funny” is different to “actual grown adult who is sober and does something more serious in terms of assault”.

        • keaswaran says:

          “you can’t simultaneously claim that BK is unacceptable because of the allegation and Biden is not, unless a) the accusations against BK is much more credible, or b) the alleged event is much more heinous.”

          Third option – BK is unacceptable because there are dozens or hundreds of qualified jurists with similar judicial profiles and political acceptability that could be nominated; while Biden is acceptable at this point because there are only two candidates, and Trump has a far longer history of more serious sexual assault (or at least, has a very uncontroversially accepted history of sexual assault, even if one wants to debate whether his are worse than Biden’s).

          Someone who said that Kavanaugh was unacceptable as the nominee while Biden was acceptable during the primary are guilty of some sort of hypocrisy, but at this point there’s no easy way to replace Biden the way there was an easy way to replace Kavanaugh at the point that everyone was calling for his replacement.

          • Aapje says:

            The problem with your ‘no other reasonable choice’ reasoning is that it is highly subjective. Trump was the only somewhat anti-globalist candidate, so based on that, you’d have to admit that Trump voters had no other reasonable choice, even in the primaries, right?

            Yet I’ve not seen many people on the left accept that argument, so aren’t we then back to motivated reasoning?

            Also, the pool of candidates for positions is going to become much worse if you punish people over really poor accusations, like Ford’s. To you, it may seem that you try to set a strong norm against sexual assault, but to me, it looks like a (sexist) witch hunt and destruction of the rule of law. Standing up to that is good in itself.

            Great injustice is commonly happening in colleges because of Democrat with hunts with Title IX, where it seems that many injustices happen & against men, blacks and foreigners in extreme disproportion, to no apparent concern of the ‘powerful’ left.

            but at this point there’s no easy way to replace Biden the way there was an easy way to replace Kavanaugh

            There was no easy way to replace Al Franken, but he was pressured into resigning. Why don’t you advocate trying to do that? Don’t you accept a reduction of the chance for a Democratic appointment (to President)?

            Trying to replace Kavanaugh had the risk of losing the majority and then being blocked from replacing him. So trying to replace him reduced the chance of a Republican appointment (to the SC).

            Are you really being objective here, or do you consider harm to Republican prospects a lot more acceptable than harm to Democrat prospects?

          • but at this point there’s no easy way to replace Biden

            Sure there is. Biden withdraws, and the Democratic convention chooses someone else.

            What there isn’t is an easy way to replace Biden if he is unwilling to be replaced.

            My own view is that the Reade case is interesting only as evidence of dishonesty by much of the left and the media, not as a serious reason for people to change their votes. The point of electing someone president isn’t to reward him for living a good life, it’s to choose the person who will do the best job of being president.

            Similarly for Kavanaugh. How he behaved as a drunk high school student or drunk college student tells us very little about how good a judge he will be.

          • Edward Scizorhands says:

            BK is unacceptable because there are dozens or hundreds of qualified jurists with similar judicial profiles and political acceptability that could be nominated

            Then why didn’t Dianne Feinstein, who had knowledge of Ford’s accusations well before the public and most of the Senate, do something?

            I’m rather sympathetic to “this accusation isn’t serious enough for any punishment but serious enough that we could have skipped over them for [big job].” But that requires Dianne Feinstein to have acted on her information early. Instead of waiting until the voting was imminent in order to maximize damage. Like trying to replace Biden now, which we all know would be a complete clusterfuck. (Remember, some people are rooting for that clusterfuck, from different sides.)

          • keaswaran says:

            Re Aapje – I’m not sure what “no other reasonable choice” criterion you’re imagining here, since I didn’t use that phrase.

            My point was just that there’s a very easy process where a President withdraws a Supreme Court nomination and then comes back a few days later with someone from the list they had drawn up a few weeks earlier. (See Harriet Miers and Samuel Alito.) Whereas there’s not a very easy process whereby a party de-certifies their nominee and comes up with a new one. So someone in a position to vote on the Supreme Court nominee has much more reason to either vote “no” or complain about the nominee even if they are otherwise ideologically aligned.

            You bring up the case of Trump himself rather than Kavanaugh. For a Republican primary voter motivated seriously by anti-globalism, you seem quite right – one might vote for him out of sincere anti-globalism despite sexual assault confessions that would have caused one to vote against Brett Kavanaugh. If Kavanaugh had been a unique jurist in the pool of Supreme Court nominees drawn up by Trump, then *maybe* some senator could have justified their lack of opposition to him on these sorts of grounds. But I never heard anyone make that case.

            I won’t debate the claim about various allegations being “poor”.

            > There was no easy way to replace Al Franken, but he was pressured into resigning.

            Wasn’t he replaced by Tina Smith? There was an easy process to replace a Senator – get the Governor to appoint a replacement. And in this case, the governor was a member of the same political party as Franken. There is no one empowered to do the same for a major party nominee (and thanks to the efforts of the Sanders campaign to prevent individuals from having power in the national party, there is unlikely to be such a process any time in the near future. Despite what David Friedman thinks, the history of conventions choosing candidates other than those selected by the electoral process isn’t exactly a smooth one – 1968 is the only example we have in recent decades.)

            And Edward Scizorhands – you won’t find me defending Dianne Feinstein on basically anything, except for some basic appreciation of her ruthless brand of politics being inspired by finding the dead bodies of her predecessor Harry Moscone and Harvey Milk when she got her big break.

          • Deiseach says:

            while Biden is acceptable at this point because there are only two candidates,

            Excuse me, but that line of reasoning can go to Hell. Would you seriously put forward “Yes, Tom Smith has been accused of deliberate murder, but guys we can’t drop him, there’s only two candidates in this race”? I can see “There’s no evidence beyond some flakey dame making wild accusations” as valid for “this does not mean we should drop him” but never, ever, no matter who it is or what it is for, go down the line of “even if this is a credible accusation/he done it, we can’t drop him because…”.

            If there was only one candidate in the race and he/she/they were credibly accused and backed up by evidence of being a rapist, murderer, long-time embezzler or other serious crime, then they should not be the nominee for anything, be that dog catcher or President.

            We need some fucking standard of ethics in public life that applies impartially to everybody regardless of political party.

          • Edward Scizorhands says:

            In order for one side to retreat, retreat needs to remain on the table as an option.

            Retreating on Harriett Miers was simple. They put the name forth, debated it, and then decided “no.” No one had put serious effort in. There was lots of time for more candidates. It wasn’t going to ruin either Bush or Miers.

            Retreating now on Biden would be a 100-megaton clusterfuck. Some commenters here talk about how it would be legal for Biden to withdraw. And, y’know, he could probably be pressured to withdraw if various Big Names decided to do it. But this would still be a complete disaster for the Democrats and we all know it.

            Retreat on Biden 6 months ago? Easy.

            Same with Kavanaugh. Retreating on Kavanaugh early in the process? No problem. A quiet word that there was an accusation, so just turn back and pick someone else would have done that.

            But to put him through all the legal vetting, confirmation hearings, look like he’s going to win the imminent vote — and then declare him an attempted rapist? It left no room for retreat. Pulling him back out would have been a complete disaster for the Republicans and we all know it. Would the Democrats leave a rapist sitting on the Federal appellate court?

          • nkurz says:

            @DavidFriedman:
            > What there isn’t is an easy way to replace Biden if he is unwilling to be replaced.

            What do you think the difficulty would be? Can they not just unilaterally say “Because of new information that has come to light, we have instead chosen to nominate Candidate X”, for any information and any Candidate X of their choosing? My understanding (for which I’d happily accept correction) is that the DNC can change its rules at any time, and that the legally the results of the primary elections are just advisory.

            As you probably know, there was a lawsuit a couple years ago (Wilding v DNC Services) which accused the DNC of favoring Clinton over Sanders despite claiming neutrality in their bylaws. The DNC’s lawyer used the colorful hypothetical in court that it would be perfectly legal for the DNC to “go into back rooms like they used to and smoke cigars and pick the candidate that way”. While as a hypothetical the court didn’t directly issue a ruling on this, the case was dismissed in manner consistent with the lawyer’s statement being legally correct.

            If you are not familiar with it (the coverage in the mainstream media was spotty), here’s a (biased) summary of the case with some of the choice quotes, and links to the actual filings: https://www.dailykos.com/stories/2017/5/31/1662843/-DNC-lawyers-say-it-can-pick-candidates-in-smoke-filled-back-room.

          • viVI_IViv says:

            Third option – BK is unacceptable because there are dozens or hundreds of qualified jurists with similar judicial profiles and political acceptability that could be nominated;

            Replacing Kavanaugh on the basis of a dubious accusation would have resulted in a moral hazard: effectively giving any woman with support of the leftist media and political establishment veto power over who gets nominated as SC justice.

          • Jaskologist says:

            Same with Kavanaugh. Retreating on Kavanaugh early in the process? No problem. A quiet word that there was an accusation, so just turn back and pick someone else would have done that.

            To expand on this a little more, there was a process put in place for precisely this purpose. The Senate Judiciary Committee does its own vetting and takes all complaints to review and enable a nominee to quietly withdraw if red flags appear. This was added in response to the Clarence Thomas hearings, whose nomination was also nearly derailed by an eleventh-hour accusation of sexual harassment.

            Diane Feinstein is a senior member of that committee. She withheld the accusations against Kavanaugh from everyone else there until the eleventh hour. By avoiding the process that had been put in place for exactly this sort of thing, she ensured that Republicans would see it as nothing but a dirty trick and stalling tactic, and a reused one at that.

          • keaswaran says:

            > > What there isn’t is an easy way to replace Biden if he is unwilling to be replaced.

            > What do you think the difficulty would be? Can they not just unilaterally say “Because of new information that has come to light, we have instead chosen to nominate Candidate X”, for any information and any Candidate X of their choosing? My understanding (for which I’d happily accept correction) is that the DNC can change its rules at any time, and that the legally the results of the primary elections are just advisory.

            You seem to think that there is a relatively unitary consciousness known as “the DNC” that can make decisions. There is a board (chaired by Tom Perez) that has the ability to decide some rules. But my understanding is that the rules of how the nominee is selected are formally written by the delegates at the previous convention four years ago. The current process was a hard-fought compromise between Clinton and Sanders delegates in 2016. I don’t believe that Tom Perez, or even the whole executive board, is legally allowed to just redefine how the nomination process works, let alone overrule the various state laws requiring delegates sent from a state to vote for the candidate they are pledged for (in this case, mostly Biden).

            I suppose it’s possible that they could just suspend the rules, and have the delegates pick someone in a smoky back room, as long as that person is Sanders. There’s been enough years of Sanders supporters accusing every aspect of the Democratic process as being undemocratic in this way, that they can’t possibly credibly agree to any candidate other than their own being chosen in this way. It helps that Sanders was the distant second choice of this election season, but there would still be a very distinct stain of illegitimacy on his candidacy, in a way that guilt of a crime wouldn’t match.

        • anonymousskimmer says:

          Judicial branch is not equivalent to the Executive branch. Personal flaws vary in their importance to the decisions made by the two branches.

      • Vorkon says:

        I’d just like to add some support to your point 1, which I think is important to remember here.

        Per previous surveys, Scott knows the majority of his readership leans left. As such, it’s perfectly rational for him to assume that a slight majority of his readership will give less credence to the accusations against Biden than against Kavanagh, regardless of whether he does so himself. And even then, he’s far from certain we will.

    • Dan says:

      No. Your argument only makes sense for why someone shouldn’t have voted for Biden in the primary. At this point, the choice is between Biden being elected and Trump being elected. Even if you disapprove of sexual predators becoming president, there is no longer any 2021 in which there is not a plausibly-accused sexual predator in the White House, and so you are forced to vote based on other criteria.

      OTOH, if Kavanaugh had been rejected, the WH would almost certainly have found a Plan B candidate who hadn’t been accused of such things, and then that person would have been appointed instead of K.

      • JG28 says:

        This would be true if you could only vote for one or the other. There are other people not credibly accused of sexual assault running for president. You’re choosing to vote for someone who is.

        • thisheavenlyconjugation says:

          The other people aren’t running credibly though.

          • gbdub says:

            Unless you’re in a swing state, only one candidate is “credibly running” by that standard. You can refuse to participate.

          • thisheavenlyconjugation says:

            @gbdub
            Yes, I agree. You shouldn’t justify voting for Biden in a non-swing state by saying it’s to stop Trump. Personally given current evidence I wouldn’t vote for Biden (or Trump) in a swing state either, but Dan’s justification is applicable there.

        • Edward Scizorhands says:

          If someone thinks that their vote is a binary choice on whether Donald Trump continues to be President, that is a completely valid choice.

      • Edward Scizorhands says:

        if Kavanaugh had been rejected, the WH would almost certainly have found a Plan B candidate who hadn’t been accused of such things

        Not if the White House believed that the response to the next candidate would have been to sit on a bunch of bad information for the next candidate, too, and wait until the voting process was about to begin, spring it, and say “well, we can’t be sure, so try again.” Just one more iteration would have run past election day 2018, and everyone recognizes a delay game.

        Dianne Feinstein had Ford’s accusation extremely early in the process. She could have said to other people “we have information that we think could disqualify him; we can’t prove it, but the optics suck. Pull back this one and put forth someone else.”

        Instead Feinstein sat on the information in order to cause maximum political damage. You can say “hey, all’s fair in love and war,” but then you can’t expect the other side to simply calmly bring out the next candidate to be executed.

      • Deiseach says:

        Suppose Biden dropped dead right now of health problems associated with contracting the coronavirus.

        Are you honestly, seriously going to tell me “Oops, too bad, Dems can’t find another candidate, looks like Trump will win by default”? Of course they’d find another candidate (seems to me one B. Sanders is a name I’ve heard mentioned).

        If the Democratic side of this mess is really going to argue “but even if he really is a rapist/sexual assaulter, we can’t find anyone else to run at this late date!” then to hell with the entire rotten system, I’m with the extremists, burn it down and start over again because this stinks to high heaven and is contrary to any principle of law or justice or righteous behaviour.

        Now, I don’t know Biden did what he is alleged to have done. Until that’s established, he’s still on course to be the nominee and he has every right to do that. But the party that had loud mouths yelling for an FBI investigation over Kavanaugh had damn well better be looking for a proper investigation into the accusations in this case, too. Otherwise it’s all come down to “accusation of rape is only a useful tool against political rivals and rape is only serious insofar as that is what makes it a useful tool of accusation. Otherwise, rape is no big deal and we’d still go with our guy”.

        And the people who were screaming in the media about “what kind of signal does appointing Kavanaugh send to women and rape survivors?” had better be every bit as willing to ask “what kind of signal does appointing Biden send to women and rape survivors?”

        • Luke Somers says:

          It is much, much easier to change your nominee when the nominee does not object (either alive or dead) than when when they do object. Biden is presumably not going to fail to object, and forcing the issue would cause a huge mess.

          This is a very unfortnate effect of our electoral system, where the spoiler effect is overwhelmingly strong, so parties can’t simply run two or more candidates in case one turns out badly. Other systems that would not suffer this defect to nearly the same extent (or at all) would be STAR, Score, any Condorcet system, or 3-2-1. Instant Runoff (IRV, a.k.a. Ranked Choice Voting) would help some.

    • To my mind, the biggest difference is that any of millions of women could have made Ford’s accusation with equal plausibility, and it only takes one who is willing to do it for political purposes or deluded. The number who could have made Reade’s accusation with equal plausibility is much smaller.

      • yodelyak says:

        Short question: What should my priors be that Russia will pressure the 1 – 3 women in the best position to embarrass Biden, if Russia discovers that it has a significant and likely deniable way to pressure them?

        Longer thoughts: Yeah, the fact Reade actually worked with Biden, however briefly, is a big difference that makes Reade’s allegations seem much more credible. One difference that cuts the other way though, is that Biden has worked in public-facing politics for 5 decades, so the list of people who could accuse him who’ve ever worked with him is quite huge, and he is now the candidate for President who could undercut Trump’s hold on power, meaning the powers of foreign governments might be directed at undermining him, and, of course, at this point they’ve had plenty of time to research every possible attack angle, and will have a dossier that includes every person–especially women, because that’s Uncle Joe’s sore point in the metoo era!–who’ve ever worked for Biden and who might be dissatisfied or succumb to pressure. Say you’re a Russian working for the “screw America’s elections” bureau, whatever that’s called these days, and you learn of a former Biden staffer, who complained to family about an excess affectionate touching, and who happens to speak Russian. A tricky situation to evaluate what’s a priori likely/unlikely…

        Although, Ford did actually grow up near and go to parties at the same places Kavanaugh went to parties (afaik, it’s not independently shown that she was at one where Kavanaugh was also present). And the rest of Kavanaugh’s persona in college, including allegations of alcoholism and frathouse-wide sex abuse, was something we could have looked into pretty definitively (at least insofar as asking three people who actually lived with Kavanaugh) and the fact the GOP didn’t do that seemed to me to prove that Kavanaugh was connected to “one of those frats” which is pretty damning, in my opinion, all on its own. What I would have liked is for a little effort to be taken to confirm/refute the wilder accusations about what Kavanaugh was like in college, and then to confirm him or not after that. Naturally that is not what we got–breathless over-reporting of every possible rumor, plus hyperpartisan mud-slinging and etc…. ah well.

        edited to remove mention of Reade having family in Russia. That may be true, but I haven’t been able to confirm the first place I saw that (which was itself just a facebook comment, not a sourced piece of journalism).

        I sorta regret trying to wade into this topic at all, it turns out I have nowhere near enough time to think about this with sufficient care to do it justice.

        • viVI_IViv says:

          Short question: What should my priors be that Russia will pressure the 1 – 3 women in the best position to embarrass Biden, if Russia discovers that it has a significant and likely deniable way to pressure them?

          Oh for f**k sake, still at this Russian conspiracy thing? Very well then, what should my priors be that China will pressure the 1 – 3 women in the best position to embarrass Trump, and failing to prevent Trump’s election directly, embarass a SCOTUS justice nominated by Trump?

        • Clutzy says:

          Short question: What should my priors be that Russia will pressure the 1 – 3 women in the best position to embarrass Biden, if Russia discovers that it has a significant and likely deniable way to pressure them?

          Your priors should be that the Russians would prefer a Biden presidency.

          • John Schilling says:

            Your priors should be that the Russians would prefer a Biden presidency.

            Citation needed.

          • Clutzy says:

            Look at his FP record. Its quite clear he is a Sino and Russio phile. The Trump-Russia theory has been thoroughly debunked and was always not a great theory because it would have required Clinton to do a huge pivot from he old Clintony ways and a huge pivot from Obama’s FP. Joe is even more Obama-y which is even more Russophillic.

            What sort of actual citation do you want? Some sort of current affairs magazine article detailing how the Stelle Dossier was almost certainly a Russian counterintelligence operation?

      • IvanFyodorovich says:

        As someone who grew up in that area, the odds of a Holton Arms girl meeting a slightly older Georgetown Prep boy are pretty high. There are thousands or tens of thousands of people who could have made the allegation with similar plausibility (anyone who went to Yale over a four year stretch at any point) and frankly I think the case against Kavanaugh was pretty weak, but it’s not like she was just some totally random person.

      • xq says:

        Ford was two degrees away from Kavanaugh at a fairly high standard (she went out for a few months with a friend Kavanaugh spent a lot of time with). I don’t see how millions of people were in that position. Biden has been in politics for 50 years and has had a lot of female staff. Not clear to me that the numbers are that different.

        • viVI_IViv says:

          Ford was two degrees away from Kavanaugh at a fairly high standard (she went out for a few months with a friend Kavanaugh spent a lot of time with). I don’t see how millions of people were in that position.

          Maybe not millions but at least tens of thousands.

          • xq says:

            I don’t think so. Social networks of high school students are generally not that big. “Squi” seems to have been in a pretty close friend group with Kavanaugh; if we say there were ~10 people in that group and consider all the girls they went out with in high school I don’t see how that gets you to tens of thousands.

            Now there’s a much larger class we could consider: the set of girls who could plausibly have been at a house party with Kavanaugh is quite large. But then, we could also greatly expand the class of women who could accuse Biden beyond his staffers. Politicians meet with lots of people! I actually think that if you consider all the women who politicians regularly interact with in private, much larger than the number of girls a student at an all-boys school interacts with, Friedman’s argument starts to favor Biden.

          • The Nybbler says:

            Being two degrees away from Kavanaugh wasn’t the standard, anyway. Being in a private school in the DC area at the right time was; knowing Squi wasn’t necessary for the allegations.

          • thisheavenlyconjugation says:

            What is the standard then?

          • viVI_IViv says:

            The accusation is that Kavanaugh assaulted her at a house party, any woman that could have attended a house party where Kavanaugh was present could plausibly make such claim, and that population is likely in the tens of thousands.

            I don’t see the relevance of this “Squi” guy and this kind of reasoning reminds me of conspiracy theorists that try to fish for “anomalies”.

          • thisheavenlyconjugation says:

            Why “at a house party”, not “at that specific party” or “at any social gathering”?

            As I see it, what we are considering is how many people could make an equally or more plausible accusation, as approximated by closeness to the accused party. The fact (as asserted by xq) that Ford was closer to Kavanaugh than a random person at a party decreases the pool.

          • The Nybbler says:

            As I see it, what we are considering is how many people could make an equally or more plausible accusation, as approximated by closeness to the accused party. The fact (as asserted by xq) that Ford was closer to Kavanaugh than a random person at a party decreases the pool.

            First, I don’t think “having dated a friend” is closer than “actually met”, the later not having been established for Kavanaugh. Second, the accusation did not require Ford having dated Squi; in fact, that came out fairly late in the hearings. The bar for the accusation was “could have been at a party with Kavanaugh”.

        • I don’t see how millions of people were in that position.

          You have to add in all of the women who were in as plausible a position to have had such an experience. That includes all women who attended Yale, all of Kavanaugh’s staff through his career, all women who ever dated a friend of Kavanaugh through college, all women who attended parties Kavanaugh was at through his college career.

          I hadn’t remembered, or hadn’t seen, the link through Kavanaugh’s friend, which narrows it a little, but that’s still considerably farther than being a staff member of Biden’s. I was thinking in terms of all women who lived close enough to Kavanaugh that they could conceivably have attended a high school party he was at, plus the equivalent for college.

          • xq says:

            But why do you add all these additional groups for Kavanaugh but consider only the specific category of “staff member” for Biden, when many women who are not on Biden’s staff could have made similar claims of sexual assault? What is the general principle here?

            In terms of lifetime exposure to potential false accusations of sexual assault they seem pretty similar.

          • I don’t think there are a lot of people in as good a position to make a believable claim as someone who actually worked for him.

            You don’t agree that “is known to have worked on his staff” is a closer connection, statistically speaking, than “could have attended a high school party that he attended”?

          • xq says:

            I don’t think there are a lot of people in as good a position to make a believable claim as someone who actually worked for him.

            Maybe, but lots of people worked for him. In addition to his time in the senate he ran three presidential campaigns, two VP campaigns and eight years as VP.

            At the same time, I think “girl my close friend is seeing” is in fact a pretty close connection; people often do socialize with their friend’s romantic partners. How many women had that degree of social connection to Kavanaugh vs. low-level-staffer connection to Biden? That depends on a bunch of details I don’t know.

            But it seems like we’re discussing minor points now that shouldn’t much affect our assessment of the cases. If Ford was just some random high school student who happened to live in the same area as Kavanaugh at the time, the argument you used for dismissing her accusation would be correct. But that’s not the case. So there just isn’t some multiple orders of magnitude difference between the situations like implied by your original post. That means we have to actually look into the details of the accusations to figure out which is more probable.

    • yodelyak says:

      Hm.

      Steelman, based on links here and here. Not really my own opinion–I didn’t like Kavanaugh, who was graceless, angry, entitled, and said boorish things in year books and had a terrible reputation in college–but I didn’t take a position on whether Ford’s memory was correct or not. At that amount of time, very hard to really know. Still would have preferred Trump discovered the idea of being graceful and asked Kavanaugh to step aside and nominated someone equally satisfying to partisans, and not the subject of a woman’s sincere belief (true or not) that he’s an abuser. I still don’t know what to think re: Reade, I’m clearly experiencing some dissonance as different things I think are true (Biden sincerely believes himself a good person, is deeply motivated/guided by that narrative, women almost never make up or exaggerate sexual assault stories, even years later… these two things don’t seem to line up.)

      W/r/t/ Christine Ford, it’s very hard to assign a high positive # to the odds that she is exactly right about everything–I don’t think very many people correctly remember everything from a traumatic event when they were 15 exactly right decades later. The fact she didn’t come forward at any other point is no real cost to her story, since anyone who’s really dug into how much goes unreported knows there’s a lot of crap that never gets reported. I think there has to be a category of person who, faced with allegations like this, would sail to a nomination anyway, because memory is notoriously imperfect. That person is honest and has a reputation for detailed work and chivalry and sobriety. Brett Kavanaugh’s highschool and college persona appeared not to match that, and his seeming willingness to lie under oath, and inability to control his temper in what has to be one of the most important interviews in his life, where remaining graceful under pressure was the number one thing he needed to do… not ideal, and to many people confirmed the worst, or at least demanded some serious effort to actually establish Kavanaugh as a sober and effective person *now*. If I’d have been in the Senate, I think I would have voted against his appointment, and pointed out that Trump could find another equally partisan person who didn’t have Kavanaugh’s problems, or just asked for a little more information, like some quiet conversations with people Kavanaugh had worked with in the last couple decades, to ensure he’s a sober boss who treats employees of all genders well. Once Rs tried to force the process to not allow any further vetting of a candidate who had (in my mind, not sure if you read the same things) been credibly accused of being an unrepentant misogynist and unrecovered alcoholic, and who had become defensive enough that I suspected at *best* we’d get a deeply divided country and a permanently embittered justice, a la Clarence Thomas…. yeah, I think I’d have voted not to confirm, and at the very least I’d have asked if there was some way to get Trump to nominate someone else instead, without even needing to vote.

      W/r/t/ Tara Reade, the fact she was hired by Biden and people close to him is a hugely damaging thing for Biden. If she’s not trustworthy, then why did he hire her? If she is trustworthy, why is she lying now? But, simply put, people do change. Russia has many times made it abundantly clear they are willing to spend millions in a pretty open attempt to sow chaos in our political process, and to elect Trump. They have also assassinated multiple people for political reasons, even outside Russia, and are certainly capable of making extreme threats. Is it really so unlikely that Tara Reade left Biden’s office at the time reported, and for the reasons originally reported, and she has Russian relatives in Russia, and although she was sharing and praising articles about Biden’s work to stop sexual assault in 2017, she then suddenly started posting weird love letters to Russia and specifically to Vladimir Putin, with bits like, “President Putin has an alluring combination of strength with gentleness. His sensuous image projects his love for life, the embodiment of grace while facing adversity. It is evident that he loves his country, his people and his job…. President Putin’s obvious reverence for women, children and animals, and his ability with sports is intoxicating to American women.” The woman has two people who corroborate that she originally told them of an event with Biden, but it’s not that solidly established (to me at least) that either is unequivocal and confident that she described a rape, not a finger along their neck–and that is a huge difference! Then she pulls/deletes all the pro-Russia stuff (thanks waybackmachine!) a week or two before interviewing to accuse Biden. Add to that some extra inconsistencies, like that Biden has multiple staffers that Reade says should be able to back up her story as something she complained to them about at the time, who deny it emphatically (and not with mealy “I don’t remember’ but rather with “if anything even remotely like that happened, there’s no chance I’d forget any allegation anything like that.”) Plus she says she filed a report, but no report exists now. Overall, seems like she might be somewhat less credible, more because there’s a significant chance of something extremely sinister here.

      • The Nybbler says:

        I’m not going to relitigate the entire Kavanaugh thing, but

        Still would have preferred Trump discovered the idea of being graceful and asked Kavanaugh to step aside and nominated someone equally satisfying to partisans, and not the subject of a woman’s sincere belief (true or not) that he’s an abuser.

        No. This is pure conflict, and allowing your nominations to be held up because the opposition can find one woman to present an apparently sincere claim that he’s an abuser is akin to paying the Danegeld.

        • yodelyak says:

          I’m sympathetic to the idea that you can’t pay the Danegeld, and I guess I’d say that the problem was with Kavanaugh’s failed response to the not-quite-sufficiently-backed charges. Even if the memory was unsound, Kavanaugh’s response felt disqualifying–my belief at the time was that Kavanaugh was entitled, appeared to have been drinking, and was angry and rude when he was being considered for one of the highest honors anyone can get.

          I recognize we shouldn’t relitigate the whole thing, and maybe I just ended up with the wrong facts at the time. In my memory, even if Kavanaugh was innocent, he perjured himself at his confirmation hearings and was connected to the absolute worst kinds of frat-brother excess in college, and was angry and entitled with challenged on those points.

          FWIW, I’m extremely unsympathetic to “pure conflict” as a theory. Trump could at least have sounded regretful about the fact that a better consensus-building candidate wasn’t available. Or that intuitions about Ford’s credibility varied so starkly, while remaining committed to his choice. Anything other than “everyone is a liar and this is a pure zero-sum game so nothing is so duplicitous that it’s off the table”–that’s a shitty thing to say or think, always. (In my view. )

          • The Nybbler says:

            Yeah, Kavanaugh was angry at being accused of attempted rape, and showed it (presumably deliberately). If he had been controlled, the headlines would have been about his psychopathic lack of emotion. It’s an obvious double bind.

            FWIW, I’m extremely unsympathetic to “pure conflict” as a theory.

            Yet it fits the facts. Not sure what else would make you refer to a charge with so little evidence for it (and even some evidence against it) as “not-quite-sufficiently-backed”. Nor to complain that Kavanaugh perjured himself without any specifics. Nor to claim he was “connected to the absolute worst kinds of frat-brother excess in college”, a statement also without any evidence. Or to complain that when accused, a candidate for the Supreme Court “was angry and entitled with challenged on those points.” If a candidate for the Supreme Court feels entitled to challenge accusations made against him…. that’s a point in his favor. The justice system, after all, is not designed for people to simply quietly accept accusations made against them.

          • Chevalier Mal Fet says:

            I was always confused that the fact that Kavanaugh was angry at being accused of rape was used as a point against him. “Angry and entitled” feels like it heavily depends on the observer – someone sympathetic to Kavanaugh could easily see “decent and honorable man righteously outraged”.

            I think it’s important to distinguish between subjective evaluations of Kavanaugh (Was he too intemperate for the SC? Or justifiably upset at being falsely accused of horrible things in front of his wife, daughter, and everyone he’s ever known? Was he acting entitled, or angry at mistreatment?) from actual facts.

      • SmilingJack says:

        “just asked for a little more information, like some quiet conversations with people Kavanaugh had worked with in the last couple decades, to ensure he’s a sober boss who treats employees of all genders well.”

        This definitely did happen. The vetting process is extensive. After the Blasey-Ford accusations came out, the Kavanaugh team produced a letter signed by a 60+ of women who were willing to speak to his character. One woman who supported him was a law professor whose daughter was clerking or had clerked for Kavanaugh, and had sent multiple female clerks to him in the past.

        “Once Rs tried to force the process to not allow any further vetting of a candidate who had (in my mind, not sure if you read the same things) been credibly accused of being an unrepentant misogynist”

        I think “unrepentant misogynist” is likely untrue for the above reasons, and because Kavanaugh made a point of hiring an all-women team of clerks when he begain his supreme-court tenure.

        I also think this interpretation of the Republican response is extremely uncharitable at best. The Republicans did request some additional FBI investigation, and it was a mistake for them to drag their feet on that point, but this allegation was released at the maximally disruptive time in the process. I don’t think Feinstein deliberately deployed it in that matter, but it sure looked like she could have. A lot of the Democratic requests motions looked like attempts to stall the process until after the mid-term elections when they hoped to flip the Senate. Giving in to that stuff probably would have felt like, as you wrote, paying the danegeld.

  8. Anteros says:

    I generally lean towards the ‘50% predictions are sort of meaningless..’, but the 300k Corona deaths at 50/50 seems quite different. If the result is 300k we can say it was a perfect prediction, and the further away from 300k in either direction, the poorer the prediction.

    On a slightly different note, I think the 90% probability of <3 million deaths is much too unlikely (a probability..) If we end up with 25% infection rates – similar to Spanish Flu – then the fatality rate would have to be 4% which is vastly higher than any plausible current estimates. Even <1 million deaths I would put at more than 99%.

    To the nearest whole number, I would estimate <3 million deaths at 100%.

    ETA To put it another way, is it realistic to say there’s a 10% chance that 3 million Americans will die in the next 8 months from Coronavirus? An average of 12,000 a day right the way through to January?

    • Robert Jones says:

      I think 90% is a bit low, but not by as much as you suggest. I would say more like 95%. There’s no reason why the infection rate should be bounded by the infection rate for Spanish Flu: it could be 80% and in that scenario the fatality rate would certainly increase from lack of medical capacity.

      Current US fatality rate is around 2,000 a day, so your figure of 12,000 doesn’t seem like an inconceivable increase (although obviously that average would involve some much higher peak).

    • Purplehermann says:

      Would you bet 5 million dollars to 1 thousand

      • Anteros says:

        No, but I’d take a 200 to one bet – in accordance with my >99.5% chance [To the nearest whole number, 100%]

        • faul_sname says:

          On the flip side, I would offer the other side of that bet at 50 : 1 (i.e. I think there’s at least a 2% chance we exceed a death toll of 3M in the US). I would possibly go as low as 25:1 or 20:1 (“possibly” mostly because that’s a pretty high ratio and such a bet is actually “this happens _and_ the other person actually pays out their side – for a small bet I’m not too concerned but a large bet is a different story).

          Split the difference at 50:1?

          Terms of bet: if there are >= 3M official covid19 deaths in the US on or before January 1, 2022, I win the bet (in this case, I’d just want the money donated to the charity of my choice at the time). If there are <3M official covid19 deaths by Jan 1, 2022, you win the bet, and either I send you the money or donate it to a charity of your choice, as you prefer (don't feel obligated to choose a charity).

          I'm fine with betting any amount up to USD 1,000 of my money against USD 50,000 of yours, where my side of the bet is that we will hit or exceed 3M US deaths (though I'd expect you probably would want to take a smaller bet than that — I'd suggest about 1% of that unless you really really know what you're doing).

          Edit: this offer is open to anyone who wants to take those odds, up to my limit. Email me at my username with a dot instead of an underscore at gmail if you actually want to take the other side of the bet.

          Edit 2: Also in my estimation, this is a really bad bet and probably nobody should take it, because you're betting that there's no significantly worse resurgence in fall / winter that hits most of the population AND there's no other mutation which makes the virus more deadly along with failure to contain the virus long term (which seems pretty much inevitable) AND there's no resurgence of the virus as normal that overwhelms hospitals AND there's no significantly-delayed covid mortality where people seem to get better but then get reinfected with a different strain which causes much worse symtoms the second time around, etc etc. Each of these possibilities seems pretty unlikely, but there are a lot of them. The NYC numbers are already pushing 1% fatality rate for infected, so it doesn't take all that much of a bump to get to 3M US-wide.

  9. Sniffnoy says:

    Feel free to get in a big fight over whether 50% predictions are meaningful.

    OK, this is getting stupid. 50% predictions are clearly meaningful; the only reason they seem meaningless in this context is because all you do every year is chart your calibration. For calibration purposes they’re meaningless. For every other purpose, all the other ways that people judge prediction, they’re perfectly meaningful. Scott, you ever thought about judging your predictions on something other than calibration? Like the Brier score or log score?

    • Peter says:

      A single Brier score, on its own, isn’t very meaningful. Scott could get a good one by picking a bunch of obvious things to make predictions about. On the other hand, if there was an invitiation for people to play along at home, and compare scores, that would make them meaningful.

    • J Mann says:

      If somebody has the patience, would they be willing to explain this issue one more time? If Scott predicts 100 events as 50% likely and 50 of them happen, isn’t that some evidence that his 50% predictions are useful?

      Alternatively, why is 50% more meaningless than 49% or 51% Are those relatively less meaningful than 25% and 75%?

      • gbdub says:

        The problem, as I understand it, is basically that a 50% prediction does not have a non-arbitrary “direction” – so how do you adjust your calibration based on the results?

        • BenChaney says:

          Scott generally gets >50% of his 50% predictions correct. That means that he has a cognitive bias that affects the way he phrases the predictions. It also means that he is underconfident on (some of) those predictions.

      • keaswaran says:

        The way that Scott is scoring this is by calibration. The way he calculates calibration is that for every prediction of P at X%, he assumes there is also a prediction of not-P at (100-X)%. He then counts all the propositions that he predicted at X% (including the implicit negations of propositions he predicted at (100-X)%) and divides the number that are true by this number, and sees how close the result is to X%. Because of this duplication of propositions and their negations, he is always automatically perfectly calibrated for the 50% predictions.

        If he were counting 20% predictions in a separate bin from the 80% predictions (this would be very natural if all his predictions were choices of which horse will win a race, or which candidate will win a multi-candidate primary, or which nation will be highest in some score, or something else with a clear set of multiple choices), then these 50% predictions wouldn’t naturally pair with their negations.
        But the way he’s scoring things, they end up automatically calibrated.

        • drunkfish says:

          That isn’t a correct summary of how he scores things. He doesn’t create a second prediction for every prediction, he just takes all the 50 negated predictions. The 50% predictions don’t get duplicated. If he originally predicted a 10, 30, 50, 70, 90, when he scores it his scores would show a 50, two 70s, and two 90s.

          • drunkfish says:

            not sure what happened there but “takes all the 50 negated predictions” should read “takes all the sub-50 predictions and negates them”

        • muskwalker says:

          The way that Scott is scoring this is by calibration. The way he calculates calibration is that for every prediction of P at X%, he assumes there is also a prediction of not-P at (100-X)%. He then counts all the propositions that he predicted at X% (including the implicit negations of propositions he predicted at (100-X)%) and divides the number that are true by this number, and sees how close the result is to X%. Because of this duplication of propositions and their negations, he is always automatically perfectly calibrated for the 50% predictions.

          As drunkfish mentioned, this isn’t the math Scott uses (he’s never reported perfect calibration on his 50%s anyway—for 2019 he reported getting 64% right). But beyond that, the method described would give you “50%” for any set of predictions, not just ones at 50%. For example take 2019’s 70% predictions:

          Of 17 predictions at 70%, I got 5 wrong and 12 right, for an average of 71%

          If for every P prediction at 70% you also assumed an implicit ¬P prediction at 30%, then you’d have 17 true predictions (12 of P and 5 of ¬P) out of 34 total predictions (17 of P and 17 of ¬P), which 17/34 is also 50%. Notice it will be 50% no matter how many he actually got right or wrong, so this method doesn’t look like it would help for the purposes of calibration.

          • keaswaran says:

            But the 70% predictions and 30% predictions would be in different bins – there would not be a single bin that is 50% true, the way there is when the two bins are both at 50%.

      • drunkfish says:

        Scott scores his predictions by what fraction of his x% predictions come out correct. The issue with 50% predictions is that IF Scott were malicious and just wanted to seem like a good predictor, he could take all of his 50% predictions, flip a coin, and if that coin came up heads he’d reverse the prediction. Then even though each prediction looks like a prediction about the event, it’s really just a prediction about the coin flip, and he’s basically guaranteed to look well calibrated.

        Example:

        I predict with ~100% chance that: No major meteor will strike Earth tomorrow, the temperature in Los Angeles will be above 10 F, Trump will still be president at the end of the day, and the San Andreas fault will not produce a magnitude 10+ earthquake. Now, if I want to be malicious, I can flip a coin (I’m actually flipping it and negating predictions for tails), and then make the following predictions:

        50% chance that no major meteor will strike earth
        50% chance that the temperature in LA will be below 10 F
        50% chance that trump will not be president
        50% chance that the San Andreas will produce a magnitude 10+ tomorrow

        Now despite all of those predictions being stupid, I’m going to have a 75% success rate. If I had a bunch more predictions, my calibration would look perfect even though I faked it.

        Note that this level of maliciousness can be done for non-50% predictions, just use a random number generator instead of a coin. That said, it’s easy for someone reading this to detect that I didn’t actually believe it would be well below freezing in LA. If you make a bunch of sincerely 50% predictions, and use a coin to reverse a random half of them, you come out looking perfectly calibrated despite not needing to be well calibrated at all.

        • keaswaran says:

          This is one reason why many people think calibration is really a bad method to evaluate probability functions, compared to some sort of accuracy score like Brier or log.

          • drunkfish says:

            That’s fair, but my impression Scott’s main goal is to test himself, not prove something, and so the issue with bad actors manipulating the score shouldn’t apply. It’s only if you’re testing people you can’t trust that this really becomes an issue.

        • J Mann says:

          OK, but since Scott’s goal is to test his own ability to calibrate probabilities, I’m not sure that the possibility of him cheating is that important, unless we think he could be subconsciously or unintentionally cheating.

    • craftman says:

      Hmmm…I swam down to the comments to argue that 50% was useful precisely because of calibration. If his 50% predictions come out 90% true, Scott has some adjustments to make about how he views the world or his personal habits/behaviors. I haven’t been involved in the past, but I’m guessing the argument that they are useless is that the way Scott words the 50% prediction directly affects whether it is correct or not (50% A happens vs. 50% A doesn’t happen). I say as a singular prediction taken on its own, yes, that is true. But looking at how well the 50% predictions turn out overall gives Scott useful information. Maybe he’s putting 50% on too many “sure thing” predictions. Also, if that’s the main critique, then 70% predictions could just be 30% prediction worded the other way, so all predictions are useless?

      Maybe we just end the whole argument by having him pick 51% as somewhat of a signal boost that “yes I worded the prediction this way because I do slightly believe that A will happen vs. NOT happen”.

      ETA: Very aware that I’m adding nothing new here and I’m sure that much smarter people than me have hashed out these same points ad nauseum in previous comment wars.

      • Gerry Quinn says:

        Though if his 50% predictions come out 90% true, he could fix that easily without changing his worldview in the slightest, by randomly switching half of them to the equivalent negative prediction.

        • BenChaney says:

          He could do this but he clearly doesn’t. Why does the fact that he could matter?

          • Gerry Quinn says:

            A 50% prediction *implies* the equivalent 50% negative prediction. If the 50% ones are coming out at 90%, there is something skew-wise in Scott’s assessment of likelihood, but in some unconscious fashion he actually knows the better assessment, and orders the two alternatives so as to put the most likely one first.

            (Or alternatively, his predictions are all correlated, and in a given year the ones predicting something he likes to put in the ‘positive’ side mostly go one way or the other.)

            But as far as I know, this is all counterfactual, and about 50% of his 50% predictions are in fact coming true, as should be the case.

          • BenChaney says:

            Generally his 50% predictions are the worst (at being close to 50%). This is implies he has a cognitive bias related to making 50% predictions. Given that he is trying to measure how good he is at predicting, the 50% predictions seem highly valuable.

    • Freddie deBoer says:

      But a 50% prediction of a binary result is an equally strong prediction of the other result and thus can’t be used to inform behavior which is the point of predictions.

      • Randy M says:

        Depends on the subject. If my brakes have a 50% chance of failing tomorrow, I’m probably going to have the car towed the mechanics today. 50% chance of rain is enough to find the umbrella, but not cancel plans yet.

      • Sniffnoy says:

        On the contrary, it absolutely informs behavior. To give the most direct example, if you assign event E a probability of 1/2, and I offer you a bet which pays you +1 utility (relative to present) on E and -2 utility on ~E, you won’t take that bet; and if I offer you a bet which pay you +1 utility on ~E and -2 utility on E, you also won’t take that bet. This pretty well distinguishes it from a probability of, say, 3/4 or 1/4.

        It only can’t inform behavior if you believe that the way to make decisions is to act as if the more likely side of an alternative is the true side. Which is not how correct decision making under uncertainty works. If a probability of 50% isn’t informing your decision making, you’re making your decisions wrong.

        • Freddie deBoer says:

          For me, if I ask someone “should I do 50/50 thing,” and they say “it’s 50/50,” I’m not going to consider myself meaningfully educated about what I should actually do.

          • Aapje says:

            Such an answer suggests that you asked a dumb question. Is is reasonable to expect them to know your preferences better than yourself? Or do you actually want more information about the potential outcomes?

            The proper way to make choices based on probabilities is to weigh the costs/benefits of the outcomes versus their probability.

            If getting soaked has a utils cost of 10 if it happens, while lugging around an umbrella has a utils costs of 8 if it stays dry, then it is smart to carry an umbrella when there is a 50% chance of rain (as 10/2 > 8/2).

            Sometimes other people can reasonably be expected to know your own preferences better than you, but very often, it is an unreasonable thing to ask.

            And if your actual issue is that you don’t understand the consequences of one or both of the possible outcomes, then you should ask people to help you figure that out, rather than demand that they both do that and judge how you would feel about the outcomes that you don’t even understand yourself.

          • Randy M says:

            Certainly not. But that’s not the form that most of Scott’s predictions take. They take the form of “Is this thing likely to happen?” and based on that–and the costs and benefits thereof–you can decide if it’s worth preparing for.

            If you end up with 50/50 chance that your plans will help on net in any way, sure, you’re right that you don’t know what to do.

          • Sniffnoy says:

            Freddie deBoer:

            It sounds like the “50/50” there isn’t describing probabilities, exactly? Or maybe you’re not using probabilities right?

            Actually it sounds like there are a few issues here. The first is that — taking everything as probabilities — what you’ve just described is a statement that conveys no new information, not a meaningless one. (Like, oh, geez, what’s the information gain of a distribution P relative to that same distribution P? 0, duh.) If I tell you that 2+2=4, that conveys no new information, but it’s still quite definitely meaningful.

            But note that it only conveys no new information because of your starting assumption that you already judged it as 50/50! If you started out juding it 75/25, or 100/0, or whatever, then being told, actually it’s 50/50, would definitely carry information! So, that’s a quick demonstration that it’s a meaningful statement. You just happen to be receiving it only in situations where it tells you nothing new.

            But back to the other issue here — it sounds like you’re mixing together your probabilities and utilities here, or your “50/50” numbers aren’t really probabilities, exactly, or something like that.

            Right, probabilities are, I’m choosing between action A and action B, and you live either in world E or world ~E, and each action yields a certain result depending on which world you live in, and you want to choose the one that gives you the best expected value (taken over all possible worlds you might live in). Note that the probabilities here are assigned to the possible worlds, E or ~E, not the actions A and B.

            Whereas what you’re describing is putting probabilities directly on “should I do A” and “should I do B”. That’s… not really what probabilities are for. Like — let’s think for a moment about what that actually means. If you’re uncertain whether you should do A or you should do B, presumably it’s because you think that there is information unknown to you that would affect the outcomes of those actions. So, either you live in world E or you live in world ~E, and you can put probabilities on that; say E has probability p and ~E has probability 1-p.

            Now maybe in world E, action A would produce a better result, and in world ~E, action B would produce a better result. So it’s tempting to say that you should do A with probability p, and that you should do B with probability 1-p. And it sounds like you’re basically doing something like that? But that’s not actually proper probablistic decision making! Which, y’know, works as described above. Doing this is a misuse of probabilities. Probabilities are assigned to states of the world, not to actions.

            Of course the other possibility here is that the “50/50” you talk about just aren’t really probabilities at all? Like I said, assigning probabilities to actions just doesn’t really make a lot of sense. It sounds like what you’re describing are maybe not really probabilities but some other measure of whether you should make a particular decision? Like you’re saying that options A and B seem equally strong, but that’s not actually a statement that certain probabilities are 50/50. Indeed, if you’re picking between a high-upside bet and a low-upside bet, it could actually be expressing that the underlying probabilities are quite lopsided!

            So I’m just not sure you’re really talking about probabilities in the first place, even if you’re calling them that. If you are then it sounds like you’re misusing them. But either way the “no information” conclusion is only a result of the assumption that you started out 50/50 — such statements aren’t no-information in general — and either way a no-information statement isn’t a meaningless one!

      • J Mann says:

        There are a lot of situations where knowing that something is 50% likely would be useful.

        The simplest is betting. We play a game where I roll a 6 sided die and pay you $1 if the die comes up 1, you pay me $1 for any other result. You learn that 3 sides of the die have a 1 on them, and conclude it’s a fair game.

        Similarly, if you know that your furnace has a 50% chance of failing in the next year, you have a pretty good idea how much you should be saving or paying for insurance. If you’re the insurance company, knowing that probability lets you know what price you can profitably offer for insurance.

        If it’s 50% likely to rain today, you’ll probably pack your umbrella.

        Now it’s true that it doesn’t really help you a ton to know that it’s 50% instead of 49% (unless you’re the insurance adjustor!) but it helps you a lot to know it’s 50% instead of having no idea.

        • Edward Scizorhands says:

          When you put it this way, even 30% likelihoods of events are pretty useful.

          • keaswaran says:

            Exactly! Depending on how significant the consequences of relevant action are, *any* probability *can* be useful.

            If for some reason I know that this particular flight has a 1% probability of crashing, that could be very useful for me, to choose to take the next flight instead.

  10. Deej says:

    10% chance of another country leaving the EU by end 2020?

    Surely vanishingly small given the practicalities of leaving – look how long it took Britain from the Tories being elected with manifesto promise of a referendum to the UK leaving. Even if Scott just means voting to leave, moves to have a referendum would need to start fairly soon which seems particularly unlikely in current the crisis.

    • Robert Jones says:

      An exit might happen in a less orderly way. It’s possible that Italy will crash out after defaulting on its debt or that Hungary will be expelled for ceasing to be a democracy. I agree it’s probably not as high as 10%.

      • Lambert says:

        Does the Treaty of Lisbon contain any mechanisms to leave the EU other than Article 50?
        Could they kick out Hungary if they wanted to?

      • Aapje says:

        @Robert Jones

        Hungary can’t be expelled. Article 7 only allows for the rights of membership to be suspended, not for a member to be expelled.

        A country can only leave the EU voluntarily, not be forced. Although…it may be possible to vex a country so much that it wants to leave, of course. Then again, the Europhiles are dogmatically in favor of expanding the EU and having as many European countries in it as possible, so trying to force out Hungary goes against their ideology.

        • keaswaran says:

          Europhiles are also dogmatically in favor of electoral checks and balances. So leaving Hungary in without any consequences also goes against their ideology.

      • Deej says:

        Yeah, I guess if a country’s government just turns round and says we’ve left, puts border checks up and stops paying then it doesn’t really matter what rules say.

        This seems massively unlikely to me though. I want to say less than 1%… surely below five at least.

        I’d love to hear Scott’s reasoning on this one.

        • Lambert says:

          IANAL, but I think you can invoke Art. 50 effective immediately.

          The problem is then that the EU’s border checks will also be up.
          The tarrifs and red tape combined will break all your supply chains and your economy will tank.

          • keaswaran says:

            Although if the borders are shut right now anyway, this might be exactly the moment that some hardliners decide to take, for precisely the reason that during the period of lockdown is the least costly time for the break to be made.

    • Ketil says:

      10% chance of another country leaving the EU by end 2020?

      I think Scott rounds to nearest 10. So consider it a <15% chance (85-100%).

  11. silver_swift says:

    Oh boy, I really hope [redacted] gets published. Sounds like a really fun read.

  12. chaosmage says:

    Crew Dragon already reached orbit in early March. I take it you made at least some of these predictions a while ago and only got around to publishing them now.

  13. Dan says:

    Is the Kim Jong Un prediction a response to the rumors that he is already dead or did you write it before that (or were unaware of those rumors)?

  14. Robert L says:

    I hope this is not bannably impertinent, but this would be more interesting if it were a lot shorter. I see no point in publishing the [redacted] stuff at all, nor things within your control like doing surveys or leaving the country.

    I am also not clear about the percentages. Does “Shutdown till election day – 10%” mean that you are just stating the odds (uninteresting) or that you think those are the hypothetical odds but you are going all in and betting at long odds that this will happen, contrary to popular opinion? Because backing a horse at 10/1, and winning, is interesting whereas just stating the odds at which you think a bookmaker would lay the horse is less so.

    • Anteros says:

      If your comment was bannably impertinent, SSC would have very few commenters left.

      I think one of the reasons Scott makes so many predictions is that it makes the calibration process more meaningful.

    • keaswaran says:

      I think the point about the redacted stuff is that it marks a public commitment to some private list he has somewhere. Making this list public helps him privately avoid cheating himself by deciding that something really was or really wasn’t on his list of predictions.

    • drunkfish says:

      I really doubt this list is meant to be interesting (although I certainly find it interesting). His assessment at the end of the year is more of a real post, but I’m pretty sure this list is just meant to be a public record, not something to entertain you.

      Your statement about what’s interesting and what isn’t also completely misses the point. This very much is about figuring out the odds that a (perfect) bookmaker would make, because the goal is for Scott to figure out his own bookmaking ability (and how to improve that ability).

      The idea that stating the odds is less interesting than taking a bet is just not true. You might be more entertained by that, but really those are the same thing if your goal is to be well calibrated. Edit: See this comment https://slatestarcodex.com/2020/04/29/predictions-for-2020/#comment-889928, all of these predictions are bets waiting to happen, that’s why they’re so interesting.

      • zzzzort says:

        The idea that stating the odds is less interesting than taking a bet is just not true.

        Well, there is a difference in the risk preference of the individual making the bet. Robert seems to have a large preference for other people’s risk.

        • Robert L says:

          Well – I find “I won at 100/1” more interesting than “I won at evens.” In this case we have 200 bets, many of them odds on, and I am not sure what overall outcome would count as interesting.

          • drunkfish says:

            The outcome that would count as interesting is if 50% of his 50% predictions come true, 60% of his 60% predictions come true, 70% of his 70% predictions come true, and so on. That’s why he has to make so many predictions, so that he/we can see whether he’s good at setting odds.

    • Enkidum says:

      You need enough predictions at each of the % levels for the values to be meaningful compared. So the more, the better. Assuming Scott is honest about what the [redacted] predictions’ outcomes are (and I do), then the electrons this post takes up are not wasted.

    • awalrus says:

      I see no point in publishing the [redacted] stuff at all

      Yeah, it seems to defeat the point of making predictions public. If you don’t want to give information away at the time but still want a public record, it’s better to share a cryptographic hash of the prediction.

  15. Tarpitz says:

    I think you’re underconfident on Johnson remaining in office: he’s got a huge majority, is in the first year of this Parliament, has tremendous personal approval ratings and no plausible challenger for the party leadership, is relatively young and has already had Coronavirus. Something truly outlandish would have to happen to overturn all this inside 7 months.

    As for extending the transition deal, I think you’re just dead wrong. The British government are going to be very clean to be free and clear as quickly as possible to avoid any possibility of being on the hook financially for however the Southern European debt crisis ends up playing out (lots of possibilities, none of them good). I’d score this one more like 75% the other way.

    • Anteros says:

      I agree with you about Boris. If bookies were offering ten to one against Boris losing his job by January, I think they’d have precisely no takers.

    • Alkatyn says:

      Re Brexit, its an issue of state capacity rather than political will. UK government was reassigning huge numbers of civil servants and other officials to deal with Brexits affects on internal UK policy and negotiating a lot of extremely complex trade deals. All that’s had to go on hold because of the pandemic and all excess capacity has been redirected to fighting it. Its likely to just be legislatively impossible to get it done in time.

      Right now the government has an obvious political and negotiating incentive to say the deadline is still in place. But when it reaches crunch the easiest option will be to publish some kind of rough outline, declare victory with “a few minor details to resolve” and extend again. That placates the pro brexit base without the massive economic and political mess that leaving without trade deals etc in place would bring. Which would be politically suicidal when the UK will already be dealing with the pandemic related depression

      • ec429 says:

        That placates the pro brexit base

        It might have, had May’s antics not already thoroughly poisoned the well for anything that smells of fudge and delay. Really, I don’t think we’ll stand for that again.

        (It’s also worth noting that — unlike in the previous Parliament — more than half of Tory MPs were Leavers at the time of the referendum, according to a Remainer think tank.)

    • thisheavenlyconjugation says:

      Agree with underconfidence, but I think there is still time for the government to catastrophically mismanage Covid-19 (as opposed to merely badly mismanaging it as the currently seem to be doing). I’d put his chances of leaving office a fair bit higher than the average Prime Minister in his position (but they’re still low, like 1% at most).

      For the transition period, I am skeptical. With no pandemic, I doubt negotiations would be finished more than a couple of months before 2021, and I expect there have been at least a couple of months disruption to them due to the pandemic.

      • Tarpitz says:

        My model is more that no meaningful negotiation will take place under any circumstances until a few months before the deadline, whenever that deadline might be. I think that’s Johnson’s model too, and it implies that agreeing to a delay can only be an admission of weakness leading to a worse deal.

    • Scott Alexander says:

      I made a high-confidence prediction for May last year along these lines (she’d just won an election, she’d just won a no-confidence vote and it was illegal for them to have a new one too soon, she’d already not resigned many times before when she could have resigned), I got burned, now I might be overly cautious.

      • Tarpitz says:

        May’s net approval was around -20, and the election had gone about as badly as it possibly could have while leaving her in office. There was also a very obvious alternative waiting for the glass to be broken in case of emergency. Johnson’s net approval seems to be in the high thirties (positive) and he won a huge majority. It’s not easy to find good comparables for his position because it’s so rare for a prime minister to have it this good: Thatcher was never this popular, even in the immediate aftermath of the Falklands War. Blair was, for a good chunk of his first term and a short period after 9/11.

        I can imagine a world where Coronavirus or Brexit eventually brings Johnson down, but it will take years, just as Iraq did for Blair. It would take some massive, currently unknown scandal to force his resignation this year, and he’s been such a high profile and controversial figure for so long that it’s hard to imagine it wouldn’t have come out already if it was out there.

  16. Jon S says:

    18. I personally am back to working not-at-home: 90%

    Does this mean you’re working not-at-home at the end of the year, or at any point in the remainder of 2020?

  17. Glen Raphael says:

    39. Bitcoin is above $5,000: 70%
    40. …above $10,000: 20%

    Am I correct in guessing you wrote down bet #40 a while ago and didn’t check the current price/trend just before posting? Or do you still stand by it? If so, wanna bet?

    (BTC jumped by $1000 in the last 24 hours. Given that it’s around $9k now, $10k doesn’t seem like much of a stretch.)

    • Loriot says:

      The prediction is for Jan 2021, not tomorrow.

      • Glen Raphael says:

        The prediction is for Jan 2021, not tomorrow.

        Right, but the sudden spectacular jump in price this week affects the odds for Jan 2021. Given where it is now, I would put the odds of BTC being over 10k on 1/1/2021 at around 50%. Whereas given where it was a week or two ago, I probably wouldn’t have.

    • Scott Alexander says:

      Yes, you’re correct.

  18. JPNunez says:

    How many of these predictions were made after the market crash caused by the corona virus?

    April 29 is a high value for “the start of 2020”.

    • craftman says:

      He specifically mentioned in his “2019 Prediction Results” post that he was lucky to have procrastinated on making the 2020 predictions. So yes I believe they were made (or re-calibrated) recently. Obviously some of them have knowledge of the lockdowns which didn’t start until March.

  19. An Fírinne says:

    You’re putting too much faith in Biden. The fact that he’s a rapist (adding allegedly here in case Scott deletes this post) is going to hurt him as is the fact that he basically has dementia and seemingly can’t string a sentence together.

    • Loriot says:

      It might hurt him more if it weren’t for Trump’s senility and alleged sexual assaults.

      Though really, partisanship trumps pretty much all else.

      • An Fírinne says:

        I think it will encourage those who were already lukewarm about him to either stay at home or vote third party.

        • craftman says:

          I’ve seen lots of friends on social media who posted frequently about Bernie, Warren, et al very quickly fall in line with the Biden nomination. Even one who currently works for a non-profit that helps women get on ballots and get elected. Right in line with whoever is wearing the blue tie. I’m not good enough friends with her to probe exactly why…

          • keaswaran says:

            I think it’s easy to explain why.

            At this point there is no other credible alternative to Biden than Trump. Trump is clearly worse on the sexual assault front, even if one ignores policy preferences. I don’t see what more needs to be explained.

            An explanation would be needed if she is now saying that one should have supported Biden back in February, or January, or last year. An explanation would also be needed if she were to now be supporting Trump, or saying that the question of whether a Democrat or a Republican is in office is irrelevant to concerns about women in politics.

          • craftman says:

            I see what you’re saying. But that assumes you’re just playing the “lesser of two evils” game in politics, which I don’t (or try not to). Is it too much to ask that someone have a deontological approach to voting and – at worst – abstain from voting for either?

            I don’t agree with 100% of everything that a candidate I vote for believes, but if my career is dedicated to women’s issues and empowering women, that seems like a pretty low threshold to clear for NOT checking the Joe Biden box even if you think it will lead to a Trump electoral win.

          • Chevalier Mal Fet says:

            Trump is clearly worse on the sexual assault front, even if one ignores policy preferences. I don’t see what more needs to be explained.

            Could you unpack this a bit for me? Speaking as someone who did not vote for Trump in 2016 and will not vote for him in 2020, I don’t see how he’s “clearly” worse than Biden on the sexual assault front. To my knowledge, Trump has never been accused of sexual assault at all – being a cad, yes. Crude, certainly. A rapist, though?

            I’d just like to know what incidents you had in mind when you said Trump was clearly worse, since I’m clearly missing something obvious.

          • Enkidum says:

            Is it too much to ask that someone have a deontological approach to voting and – at worst – abstain from voting for either?

            Yes it is too much to ask, because the election is a power struggle between groups whose ideas do not have equal moral worth but do have real consequences, not a philosophical debate.

          • Thomas Jorgensen says:

            Chevalier… really? You are in one heck of a filter bubble. https://en.wikipedia.org/wiki/Donald_Trump_sexual_misconduct_allegations

      • JPNunez says:

        Trump may be senile but it is not as far gone as Biden.

        Biden looks really lost at times. Some times he manages to pull himself together, but dunno if he can do it on demand.

  20. Deiseach says:

    23. Conditional on me asking about Reade on SSC survey, average survey-taker’s credence in her accusation is greater than 50%: 70%
    24. …greater than 75%: 10%
    25. …greater than credence in Kavanaugh accusation asked in the same format: 40%

    This interests me because I haven’t seen much discussion of whether people think the accusation is credible. Lots of “this demonstrates the hypocrisy of the Democrats/the left because compare it with how they treated Kavanaugh” and “Reade’s accusation is better based and more credible than Blasey Ford’s one” but not much, if any, “yeah I think it happened the way she said it did”.

    I’m in the same boat as I was for Blasey Ford: it’s plausible, sure. It could have happened. Did it happen? I don’t know. Is Ford/Reade lying (because Dem shill/Russian asset)? I don’t know.

    Now, if the same thing happens as with Kavanaugh and suddenly a raft of accusations, each one dumber than the last, comes flooding in the wake of Reade’s accusation, then I’m going to give Biden the benefit of the doubt. But at the moment, it’s only Reade, and it’s not one of the crazier ones (gang rape drug parties, for instance) so yeah: could have happened, no idea if it did happen.

    • Aapje says:

      It doesn’t seem like there is sufficient evidence for a conviction, but it does seem credible in the sense that she told witnesses that her boss penetrated her back then. In contrast, the Kavanaugh accusations seem more like case studies in recovered memories, where the accusers didn’t accuse Kavanaugh at the time (and not even until recently), but became more and more convinced that he was the perpetrator, at the same time that Kavanaugh became a far more prominent Republican.

      • Randy M says:

        Reade’s story is slightly more likely to be true but it’s still he said (assuming Biden gets around to flatly denying it)/she said (even if she is consistently saying the same thing, that’s one single source).
        That’s not enough to have a strong opinion on. Error bars way too high. And due process requires more (admittedly neither case was a legal trial).
        I’d advise against a daughter interning with either. Not sure it should influence a vote at this point.

        Note that my opinion might be different if I knew any of the involved parties well.

      • Aapje says:

        Note that I changed my opinion on this after looking into it more, see my other comment.

    • craftman says:

      This interests me because I haven’t seen much discussion of whether people think the accusation is credible.

      Isn’t the first non-Fox/Brietbart news source that brings this up going to be #believewomen’ed out of existence? It’s definitely a tough story to hit “Publish” on if you’re a journalist with a promising or established career.

      • BPC says:

        Not so sure about that. Eg:
        https://reason.com/2020/04/12/tara-reade-joe-biden-new-york-times-sexual-assault/
        At the very least, the Grey Lady has not exactly been credulous, and the language they used was… honestly, kind of stomach-churning for me. Hell, you want #MeToo? TimesUp refused to help her because Biden was running for president. https://theintercept.com/2020/03/24/joe-biden-metoo-times-up/
        I have been spending entirely too much time on Twitter but a lot of very vocal #MeToo voices from the liberal corner (people who work with various campaigns, for example) have been either silent or dismissive of this case, in a way that sickens me, as someone who doesn’t just pretend to care about this stuff.

        • Edward Scizorhands says:

          It some ways the silence is better than the dismissals.

        • Deiseach says:

          I have been spending entirely too much time on Twitter but a lot of very vocal #MeToo voices from the liberal corner (people who work with various campaigns, for example) have been either silent or dismissive of this case, in a way that sickens me, as someone who doesn’t just pretend to care about this stuff.

          It’s all too reminiscent of the Monica Lewinski affair; prominent women (feminists or otherwise) making remarks about strapping on the Presidential kneepads to give him blowjobs so long as he was keeping abortion legal.

          So long as Biden keeps the tax dollars flowing to Planned Parenthood, he can grab ’em by the pussy and they’ll let him. If he’s found wanting in unflagging support for abortion, that will be what will sink him, not any rape/sexual assault accusations.

      • Edward Scizorhands says:

        USA Today has an opinion piece about reasons to be skeptical of Reade’s story.

        https://www.usatoday.com/story/opinion/2020/04/29/joe-biden-sexual-assault-allegation-tara-reade-column/3046962001/

        I think it even leaves out the fact that at one point Reade said “it wasn’t him.”

        • blumenko says:

          When did she say that? I seem to remember something similar, but can’t find it.

          • Edward Scizorhands says:

            It wasn’t quite as I remember. But she seemed to be more annoyed with Biden’s staff than with Biden, which is weird if Biden assaulted her.

            https://twitter.com/robbysoave/status/1249707393904193537

          • It’s not that weird if she assumed that the assault was a mistake, that he really believed she was romantically attracted to him and would welcome what he did. That makes him socially very clumsy, but not evil. If the staff then tried to shut her up, force her out, pretend it hadn’t happened, she could have seen that as deliberate bad action.

    • mtl1882 says:

      The Larry King phone call indicates something was up. That’s unlikely to be a red herring. But this could mean:

      1) She is telling the truth, and current circumstances led her to speak up now even though she previously kept quiet. This makes sense to me, given the emphasis on #MeToo and the behavior of the Democratic Party establishment. Also, there were probably a lot of calls for someone to let anyone know about problems in Biden’s past, especially when he looked less secure in the nomination. My understanding is that she was understandably disturbed but believed he had misread her interest (not that this would excuse the way he approached her), and that she didn’t view it as a crime. So it’s not weird that she didn’t speak up before (of course, victims of sex crimes also are often reluctant to come forward) or might have spoken positively of him at times.
      2) She had another sort of difficulty with Biden, which is what her mother was referring to. However, the idea that she had tried to deal with and found her only recourse was going to the press, which she didn’t want to do because it would damage him, seems more consistent with what she alleged than some sort of more routine complaint like being overworked or not getting credit for work. It’s impossible to know exactly what that call was about, or what her mom’s exact impression was.
      3) She has had a vendetta against Biden for decades and was telling her mom and other people bad things about him, but didn’t feel it was smart to go public until he was totally in the spotlight in a contentious election and the issue of inappropriate behavior with women became a central discussion.

      I find the scenario she describes plausible enough, and don’t see any major red flags, but certainly I can’t say it’s anything close to proven. I don’t think her credibility has been effectively undermined in any major way, nor do is there anything that doesn’t add up. But I have no idea what happened.

      • m.alex.matt says:

        Also, there were probably a lot of calls for someone to let anyone know about problems in Biden’s past, especially when he looked less secure in the nomination.

        Or, she was a major Bernie fan and Biden was shaping up to be his main competitor in February when she first started shopping around for a competitor.

        I’ll admit, I’m baffled that everybody is treating the whole story as ‘credible’. I was deeply incredulous about the Kavanaugh allegations and I’m deeply incredulous of these, so I feel like I have some consistency on the matter of, “Not easily buying into allegations about decades old misconduct with little or no physical, verifiable, or otherwise objective proof of anything”.

        Seeing it shape up in real time it’s fairly obvious to me that Reade herself did this for (crazy) political reasons and the Bernie-or-Bust side of social media pushed it as hard as possible into some sort of mainstream consciousness. Now it has taken on a life of its own, so congrats to them, Bernie is still not getting the nomination.

        • mtl1882 says:

          The Bernie side definitely pushed it hard for political reasons, but they would do that for any accusation because it was so fun to point out the Democratic establishment’s ridiculous hypocrisy. I wasn’t aware that she was a Bernie supporter–that does make me more suspicious of her motivations.

        • Cliff says:

          Well she doesn’t have a time machine, so we know she didn’t make it up after Bernie v. Biden became a thing. I do think it would be weird for someone to make up this story and repeat it to various people over the years with no apparent underlying motive, but not impossible.

      • benf says:

        “2) She had another sort of difficulty with Biden, which is what her mother was referring to.”

        Her story from the 90s until this year was that some of Biden’s staffers were more or less bullying her, making demeaning comments about her looks, etc. Perfectly reasonable and totally plausible but didn’t get pretty much any traction because, c’mon.

        Now her story is he ALSO totally forcibly fingerbanged her and she didn’t feel comfortable talking about it until he was firmly in the lead for the Democratic nomination because…reasons.

        There are two stories: one I believe, one that’s pretty obviously horseshit. And note, when headlines say “Witness backs up Reade allegations” they ALWAYS mean the first set. Literally nobody has backed up the second in a meaningful way.

        • Cliff says:

          …a former neighbor, Lynda LaCasse, as saying that Reade told her, in 1995 or 1996, that she had been assaulted by Biden: “I remember her saying, here was this person that she was working for and she idolized him. And he kind of put her up against a wall. And he put his hand up her skirt and he put his fingers inside her…

    • BBA says:

      As best I can tell, “credible” just means it’s not accusing someone you like.

      I try to counter this natural cognitive bias by treating allegations against people I like as more credible and allegations against people I dislike as less credible, but that just makes me even more miserable than my baseline, so maybe I should just stop it.

    • shakeddown says:

      All things considered the Reade allegation seems completely implausible. I have a hard time seeing how anyone could give it high probability unless their decision algorithm is “automatically believe all accusations no matter how ridiculous”. That Scott’s even giving it above 10% – let alone 90% – is flabbergasting.

      • thisheavenlyconjugation says:

        Do you consider the accusations against Kavanaugh, Trump and Bill Clinton equally implausible? If not, what are the relevant differences?

        • keaswaran says:

          I think it’s odd to include Trump in there when the main allegations come from his own mouth in the Access Hollywood tape, while the others all denied the accusations.

          • thisheavenlyconjugation says:

            I am loathe to defend Trump, but the tape doesn’t feature any specific allegations.

        • Edward Scizorhands says:

          Biden and Kavanaugh only have one credible accuser. It’s rare (not impossible) for a sexual abuser to do it once and stop; typically they do it repeatedly.

          Clinton and Trump (and Cosby, to go outside of politics) each had multiple credible accusers.

          • mtl1882 says:

            Yeah, this is relevant, but the type of assault alleged matters. There’s a difference, at least in my mind, between having been highly inappropriate in a culture of entitlement and being like Cosby, who can only be described as a predator. Cosby wouldn’t have taken consent if it was offered. He only wanted nonconsenting women, and was quite calculating. It’s unbelievably disturbing and he can’t be in a position of trust ever again.

            I don’t really get exactly what Kavanaugh was accused of, but I can believe a group of drunk teenage boys might have sexually assaulted a girl without all of them going on to be serial rapists. It seems unlikely to me from the record that Kavanaugh was predatory. He may have come from an entitlement culture, and could may have been sexually aggressive at times in adulthood, but nothing that led to public complaints, and his personal life seems stable enough from what I can see.

            Biden is a very touchy social person who probably felt entitlement around women, based on his age and position. I can easily see him misreading a situation and being inappropriately aggressive, but then backing off when he realizes he misread the situation. But it is surprising to me that, if this is the case, there were not more such claims. Possibly because he is so well liked and has had so many personal issues, no one has felt like doing so, especially if they felt he thought they were interested and was just really recklessly aggressive in making a move. As far as I know, there haven’t been infidelity allegations, either, which always surprised me, because of that hypersocial, charming nature and position of power. What Reade claims to have experienced would be disturbing and humiliating for anyone in her position and is definitely wrong. But she said he immediately backed off, and there aren’t more claims, so I doubt he has a history of really predatory behavior. He probably did a lot of inappropriate flirting that made younger staff uncomfortable.

            I have no doubt Clinton was sexually aggressive with women, but I don’t know the details of every case. Clearly he was sexually involved with other women while married and felt entitled to do so. What stands out to me is more the vilification of and lying about Monica Lewinski–there was something ugly and sinister in that. I know part of it was just the craziness and vitriol of the political frenzy and trying to fight back (I still don’t quite get how it all became so public, and the way politicians on both sides handled it was both predictable and sleazy). But as a result, I’d feel more uncomfortable voting for Clinton than for Biden, as a moral issue. I believe the sex with ML was consensual and I don’t really care that much about the infidelity itself, but there seems to be a pattern of entitled callousness that doesn’t mix well with power. On a practical level, however, I’d definitely take Clinton right now over Biden, because we need someone who can lead and communicate more effectively than Biden appears able to. Although Clinton may have deteriorated quite a bit as well at this point.

            The unprecedented mania for taking down Trump by any means possible means that I can’t evaluate any of those claims. There was just a highly unusual incentive to make and corroborate claims against Trump and get highly favorable coverage of them. We know Trump says a lot of things to be outrageous and likes to portray himself as the ultimate winner. It is easy for me to believe Trump was unfaithful to his wife, that he was flirty and made inappropriate comments to some women, and that was sexually aggressive with women he was convinced “really wanted it,” because he was such an alpha guy. (It’s also easy for me to believe that this might have just been bragging to appear as such an alpha guy.) I don’t think he’s a crazy womanizer–he seems to go for one woman at a time. When he cheated on Ivana, it seem things had already kind of fallen apart, and he seems to have confined himself to one mistress. I get the sense his germophobia extends to fears of STDs. Supposedly, when Melania was pregnant, he took up with Stormy Daniels–not sure I buy that one. I’m sure he makes lots of cringey comments about women’s looks, etc., on a regular basis (granted, he does this to men also). But I’m inclined to think he doesn’t have a history of predatory behavior, just boorishness.

            From what I’ve read of American presidential history, they appear to have a surprisingly good record for marital fidelity. Maybe it just didn’t make it into the history books, but even that indicates exceptional discretion for such powerful men. There were obvious exceptions, like JFK and Clinton.

      • bmcfluff says:

        @shakeddown It looks like you’re missing a level of indirection. Scott’s exact prediction was “Balance of evidence available on Election Day supports (as per my opinion) Tara Reade accusation: 90%”. My understanding of that sentence is that Scott is not assigning 90% probability to the statement “Tara Reade accusation is true”, but rather to the statement “On Election Day I will assign >50% probability to the statement ‘Tara Reade accusation is true’.”. Which presumably means he’d also assign >50% to the object-level claim right now, but not necessarily 90%. Though of course that’s still a lot higher than the 10% you mention.

      • mtl1882 says:

        I think the Larry King tape alone takes it to 10%. But when I say credible, I don’t mean high probability or even more likely than not. I mean that it seems to make sense in the overall context, such that I can’t easily reject it as fabricated. This is rarer than one might think—a lot of national profile false accusations have glaring issues and are suspicious from day one, even if they gain traction. The only two coming to mind are Fish tank cleaner lady and Jussie Smollett, but I’m sure others can think of examples. This recent article is a really good explanation of how little things can be red flags.

        What Marta obsessed over was that “Rebecca James” had referred to her as Marta Cabrero. In Spain, everyone has two last names. Hers are Tecedor and Cabrero. The first last name is the primary one, so people in her department would call her Dr. Tecedor, though most of the time, per her preference, everyone just calls her Marta.

        Note what they considered a “credible” claim. Not everyone uses it this way, but credible really often just means a claim that isn’t absurd or malicious on its face. It doesn’t necessarily mean any actual evidence was provided. It’s confusing the because press also uses the term in reference to well-substantiated reporting.

        I say credible because we know she did work with Biden, and talk to her mom about some kind of issues there, but they ultimately did not go public with those issues, which would make sense to me under the circumstances. It would also make sense for her to not feel compelled to go public until now, when it has been headline news for two years, both Biden and sexual misconduct allegations, including Biden often denouncing sexual misconduct and Democrats strongly encouraging women to speak up. There would have been nothing but a hassle in going public earlier. What she describes happening is seems like an incident that could have happened, in terms of behavior and circumstances, and it doesn’t read like a hit piece on Biden. Her characterization of the incident is straightforward and fairly nuanced; it doesn’t contain dramatic or moral language. In other words, it doesn’t sound scripted.

        But I wouldn’t give it anywhere near a 90%. The media and political environment right now is such that, for a lot of reasons, fake claims can thrive, and not just ones about sexual assault. This election is really bitter and absurd and a lot of people are just really cynical or desperate for attention or a cause. The behavior of the Democratic establishment related to Bernie Sanders and some other things probably made a lot of Democrats/progressives feel betrayed, which could explain why Reade made her comments when she did, whether true or false.

        I don’t know much about the Bernie Sanders connection, but that lowers my assessment of her credibility a little bit. However, it’s also possible that the allegations were true and she finally decided to come forward because she wanted to save Sanders. I also read she was drowning in student debt and medical bills after attending law school recently, so it could be desperation for money or a desire to escape from the frustration of her life by becoming part of the story. She seems like an intelligent person who wanted to go into government, and it must be disappointing for it not to work out.

  21. Nikitis says:

    It seems you’re letting your bias affect your judgment and vastly underestimating Trump’s chance of winning, just like you did 2016. I can’t plausibly see him doing any worse than he did then, and Biden’s ONLY positive is that he’s not Trump. And even then there’s not that much difference between two rich, old white guys who provably don’t respect women and probably plan to benefit their equally rich friends at the expense of the common people.

    I predict an 80% chance of Trump winning, and 40% that it’s going to be by a larger margin than in 2016. And I’m not even sure if these are conservative estimates.

    • keaswaran says:

      “I can’t plausibly see him doing any worse than he did then”.

      Do you really think it’s completely implausible that 8 months of coronavirus anxieties influence the electorate strongly in some unexpected way? I think there’s no clear reason to think it would hurt Trump rather than help him, but do you really think it’s so clear what the reaction would be that it’s implausible that it would hurt him?

      Furthermore, if the expectation is that Trump does exactly as well as he did last time, then that sounds like 50% is exactly the right prediction. Last time was well within the margin of error of things like weather gradients across swing states changing the outcome.

      You need to be predicting that the electoral response to coronavirus is a clear boost for Trump and/or that Biden is clearly a worse candidate than Clinton in order to think that the right prediction is greater than 50% for Trump.

      (Also, when you say “it’s going to be by a larger margin than in 2016”, what margin are you talking about? Presumably not popular vote, because he didn’t win that. And if it’s electoral vote, then you need to be predicting that there are some states Clinton won that Trump is likely to win this time, which may be true, but it would be helpful to have specified some.)

      • Nikitis says:

        Yes, I find it completely implausible that coronavirus will cause a significant amount of people to change their minds regarding their party affiliation. It will simply cause them to further entrench in support of their side. Even if Trump fails badly, the WHO, the experts and the states are sufficient scapegoats for his supporters. And if he somehow succeeds wildly, those same organizations and individuals will take the credit for his enemies. The epidemic is gonna be a wash, support-wise, with people shouting at each other that Trump is a monster vs Trump is a savior, convincing approximately no one in either direction.

        I thought it was obvious, but yes I am predicting Biden is a worse candidate than Clinton, who was already a bad candidate. He has all of her disadvantages, none of her advantages, and is somehow even less charismatic than her.

        The margin I’m talking about is the only margin that matters in the presidential elections, which is the electors. I’m not even gonna try predicting specific state results, because that way lies madness and nitpicking.

        • John Schilling says:

          I am predicting Biden is a worse candidate than Clinton, who was already a bad candidate. He has all of her disadvantages,

          Biden is not viscerally hated by nearly as many people as Hillary Clinton was. He is not regarded as hopelessly mired in corruption by nearly as many people as Hillary Clinton was. You may believe he should be held in that same lack of esteem, but he really isn’t outside – except in the bubbles that are always going to vote for the candidate with the (R) after their name.

          Biden is short on absolute advantages, but he avoids most of Trump’s and Clinton’s big disadvantages.

          • Nikitis says:

            The people who actually hated Clinton were never likely to vote for any Democratic candidate. It doesn’t matter if Biden is more acceptable to them, if he doesn’t mobilise the people who might actually vote for a Democrat. And to them, he doesn’t really offer much. He’s just as much of an insider as Clinton was, if not more so. And he’s an old white man, like his opponent. This matters for the party that has made identity politics a pillar of its policies.

            But who he is doesn’t matter as much as what he offers. Obama offered Change and Hope. Trump offered to “Drain the Swamp” and “MAGA”. After the defeat of 2016, the Democratic party as a whole settled on the explanation that their message was fine, it was just Clinton that was terrible. Biden’s largely running on the same message: Obama’s years were just [i]swell[/i], let’s just keep doing that. This didn’t work for Clinton, and it won’t work for him, as there a lot of people who were not exactly happy with Obama’s term. Trump may not be offering the right solution to these people, but at least he acknowledges there is a problem. I really hope that after their coming defeat the Democrats finally realise that betting on enforcing the status quo is not gonna cut it anymore.

          • John Schilling says:

            The people who actually hated Clinton were never likely to vote for any Democratic candidate.

            I do not believe that is true of e.g. the white working-class voters who defected to Trump. Their support for (often union) labor concerns had made them fairly reliable supporters for generic Democrats, and I don’t think it was at all 100% “yay Trump!” that made them turn away from their traditional party in 2016.

          • Nikitis says:

            I really, really doubt the people you’re referring to actually hated Clinton. And of course they weren’t suddenly stuck with how awesome Trump was. But for the first time in a long time, they could believe that the Republican candidate would address their, as you said, labor-related concerns better than the Democratic candidate.

            This is the point I was trying to make when I talked about the message. A generic Democrat won’t cut it anymore. Not against Trump, and probably not against whoever’s the Republican candidate in ’24. The Democratic Party didn’t realise this in ’16, and they even doubled down after their defeat. I really hope 8 years of Trump will be enough to teach that lesson, but I’m not exactly optimistic in that regard.

        • bullseye says:

          Biden’s personality is much more appealing than Clinton’s. He seems like a nice guy; she’s smug and condescending.

          I agree that the pandemic isn’t going to make partisans switch sides, but it might well influence swing voters. People always give the President more credit or blame than he deserves for the economy, so why not for the COVID-19? If things are still bad in November, Trump’s going to lose, regardless of how much it’s actually his fault and whether Biden could have done better.

          • Nikitis says:

            Trump won the last elections. A pleasant personality doesn’t count nearly as much as you think.

            And I don’t think the pandemic is gonna influence the election that much. Obama didn’t lose automatically in 2012 when the world was hit by the worst financial crisis in recent memory. Why would this case be different?

          • bullseye says:

            Trump isn’t pleasant. But he’s unpleasant in a way that a lot of people like. Clinton comes across as an elitist prick who looks down on voters; neither Trump nor Biden have that issue.

    • thisheavenlyconjugation says:

      You should bet on it then, since the odds are around 50/50. Also note that Biden is currently leading in the polls.

      • EchoChaos says:

        But interestingly, losing in the betting markets.

        Which means either the betting markets are irrational, or they have a strong reason to disbelieve the polls.

        • thisheavenlyconjugation says:

          The betting markets are definitely irrational, but I think the obvious explanation in this case is that a lead in the polls for Biden (i.e. in the popular vote) doesn’t equate to better than 50% odds of winning a majority of electoral votes.

        • shakeddown says:

          Betting markets pretty consistently underperform polling averages on politics. So yes, they are irrational.

          • Dan L says:

            The spicy way to phrase it is “conventional wisdom deviates from polling in the direction opposite of truth”, but ultimately I think that framing confuses which is more responsive to reality. Rephrase as “most people will be more informed by incorporating polling averages in their beliefs” and it’s fairly straightforward.

      • Nikitis says:

        PredictIt isn’t available in my country. Probably because gambling like that without a license is not exactly legal here.

        • thisheavenlyconjugation says:

          Betfair? It’s better than PredictIt anyway.

          • Nikitis says:

            Not available either. And I thought it was exclusively a sports site?

          • thisheavenlyconjugation says:

            They’ve got politics as well, and the overhead is a lot lower. Other than that, the only one I can think of is Augur but that would probably only make sense if you already own cryptocurrency and it looks like they’re (temporarily) shutting down anyway.

          • Nikitis says:

            Oh well. I expect we’re too small a market to be worth the jump through legislative hoops. It’d be nice, but I’m not that torn up about it.

    • m.alex.matt says:

      two rich, old white guys who provably don’t respect women and probably plan to benefit their equally rich friends at the expense of the common people.

      Yeah, the co-sponser of the Violence Against Women act provably doesn’t respect women.

      • Nikitis says:

        Biden LITERALLY “grabbed a woman by the pussy”. Remember the outrage when Trump merely said those words? A bit hypocritical, if you ask me.

        • Biden is claimed to have done so, but we don’t know if it is true and he denies it.

          Trump implied he did so, but we don’t know if that is true either.

          • Nikitis says:

            Which brings us back to my original point, that they are equally bad in that regard.

        • keaswaran says:

          This seems completely backwards. Trump grabbed many women by the pussy multiple times. Biden did it once. (If you want to start doubting these claims, that’s reasonable, but Trump brags about it while Biden tries to cover it up, which at least seems to make clear their attitude about such violations.)

    • Scott Alexander says:

      I’m mostly going off the fact that Biden beats Trump handily in every general election poll and in polls of most swing states.

      If you actually believe 80%, I’ll bet you at 3:1 odds. My $100 vs. your $300. Up for it?

      • alchemy29 says:

        Sorry to spoil your free money bet but this is what PredictIt is for. Think other people are hopelessly irrational? Easy money.

      • Bugmaster says:

        I’d be up for it, but I’ve never placed a semi-anonymous Internet bet — how do they work, mechanically ?

      • Nikitis says:

        Polls have proven notoriously unreliable in recent years though. I’m not sure why you still give them such credence.

        And I’m afraid I’ll have to decline your generous offer. You are asking me to gamble with literally half my paycheck on terms that are extremely favorable to you and with an amount of money that I assume you’ll not especially miss. I can counteroffer a symbolic bet with 2:1 odds, splitting the difference between our predictions. My $2 for your $1. Mostly for the bragging rights.

        And if Trump wins, I would very much like you to not dismiss it as getting unlucky in a coin toss, but actually stop and think about what you missed. Hopefully step out of your bubble a bit.

        A little less smugness would probably be good too, but I won’t hold my breath.

        • Bugmaster says:

          I believe that you are right (and I offered to take up Scott’s offer in my comment above), but I’ve got to defend him a little. I don’t think he’s (just) being smug; he’s just following the math where it leads. He really does believe that he’s offering you a good deal; the problem is that money is worth more to you than it is to him, and the naive expected value equation does not account for that. That is kind of a weakness in those types of calculations, in general.

          • Nikitis says:

            I do read some smugness in his tone, and in the fact that the 75% odds he offered are A LOT closer to my prediction than to his. It reeks of a desire to put the arrogant upstart in his place, and make some money on the side.

            I could be wrong, tone is pretty hard to read in the internet. I apologise in advance if that’s the case.

          • Cliff says:

            Talk is cheap. Many people say things they do not believe or have no confidence in, and in an aggressive manner to boot.

          • Taleuntum says:

            @Nikitis

            Scott’s opinions about betting on predictions: https://slatestarcodex.com/predictions-bets/

            Note, if Scott’s time wasn’t presumably more valuable than yours, ie the bet were between equal partners, a fair odds would be 1.85714:1 (or 13:7). Ie, a bet with such odds would be worth equal amount of expected money to the bettors. Of course, any odds less than 4:1 is valuable to you (ie, you expect to profit on it), but I can understand if you are risk-averse at such high amounts of money.

            I recommend reading https://www.lesswrong.com/posts/aiz4FCKTgFBtKiWsE/even-odds for a strategy where the bettors are incentivized to report their honest predicted probabilities.

        • My $2 for your $1.

          My father’s rule was never to bet more than nickel. But there has been some inflation since then.

        • keaswaran says:

          “Polls have proven notoriously unreliable in recent years though.”

          Polls have been notoriously doubted by the public opinion in recent years, but the 2016 polls were basically no different than they had been in the past few years. It just so happened that this time the error in a few swing states went exactly the same way and happened to cross the plurality mark, while in other years there weren’t enough swing states for the standard size of error to make a difference to the final result.

          https://fivethirtyeight.com/features/the-state-of-the-polls-2019/

        • michaelkeenan0 says:

          I’d like to avoid coming across as smug to people, but worry that I don’t know all the ways smugness can come across. If it had been me, I might have written the same thing Scott did. Could you rewrite his comment in a way that wouldn’t have felt smug?

    • benf says:

      Actually, Biden’s main positive is that he’s not HILLARY. You really cannot underestimate the impact that having such an unpopular nominee had on the 2016 race. Trump won Wisconsin with fewer votes that Romney, who lost Wisconsin by seven points.

  22. craftman says:

    52. I publish [redacted]: 80%
    53. …conditional on being published, it gets at least 40,000 pageviews: 10%
    54. I publish [redacted]: 60%
    55. …conditional on being published, it gets at least 40,000 pageviews: 50%

    I love what this implies. The thing I am more likely to publish is not interesting to the general public at all, while the thing I am less likely to publish will probably be spread around the internet like crazy.

  23. BPC says:

    14. General consensus is that we (April 2020 US) were overreacting: 50%
    15. General consensus is that we (April 2020 US) were underreacting: 20%

    This does not gel with the understanding I have of the virus.

    I find it extremely unlikely that we were anything resembling overreacting this month. We are still having public gatherings with no real crackdown; states are preparing to reopen everything despite the fact that many still aren’t even at the peak of the first wave yet. The idea that it’s significantly less likely that we were underreacting than overreacting just does not compute for me.

    • Pandemic Shmandemic says:

      I think this is all about how one gauges “General Consensus” – it is probably one of those things people will be arguing about for years if not decades and will ultimately reduce to how lives should be weighed against economic interests which in turn is MP-complete.

      *Morals and Politics.

    • salvorhardin says:

      Where are we still having public *indoor* gatherings with no real crackdown? Outdoor yes, but there is little evidence that outdoor gatherings are a major transmission vector.

    • drunkfish says:

      Unemployment is at like 20%. I don’t think I’d bet at even odds that we’re overreacting, but I think it’s *very* plausible that we are. If it takes multiple years for the economy to recover (not obvious by any meanS), the economic costs (which have corresponding human life costs) are enormous.

      • BPC says:

        Unemployment is at 20%. So what happens when we send people back to work and a bunch more people start dying?

        https://www.ft.com/content/e593e7d4-b82a-4bf9-8497-426eee43bcbc

        The IGM Economic Experts Panel’s latest survey of top US macroeconomists asked for their view of the statement “Abandoning severe lockdowns at a time when the likelihood of a resurgence in infections remains high will lead to greater total economic damage than sustaining the lockdowns to eliminate the resurgence risk”. Eighty per cent of the panel agreed, the rest were uncertain or did not respond. Not a single expert disagreed.

        Remember, they’re only being asked about the economics of it, and they’re saying with near-unanimity that loosening the lockdown will make things worse in the long term. And it makes perfect sense why – how good could the economy possibly be when the second wave of a deadly pandemic is raging through the nation?

        • The Nybbler says:

          It’s an obviously loaded question and the answer is nearly meaningless as a result.

        • John Schilling says:

          <blockquote.when the likelihood of a resurgence in infections remains high…

          …is an unquantified hedge and a punt to a domain outside their expertise. That it would be economically disastrous to end the lockdowns at a time when the likelihood of a “resurgence” was “high” enough to result in economic disaster, is a worthless tautology. Put numbers on “resurgence” and “high”, and you’ve got something useful to say. Then tell us whether those numbers reflect reality, or get out of the way for someone who can.

        • drunkfish says:

          Economists surveyed during the beginnings of recessions don’t generally realize there’s a recession. Economics is not a terribly predictive science, even if that question weren’t loaded.

          I support the lockdowns, but it’s still entirely plausible that they’re an overreaction. The main way, aside from the economy not rebounding, that I see it turning out to be an overreaction, is if it takes so long to get a vaccine that ~everyone gets sick anyway. Then all we did is draw it out, without saving all that many lives, so we’ve traded much longer economic damage for a similar death rate.

  24. SEE says:

    You’re substantially underconfident on Crew Dragon, given that it had a successful unmanned orbital test already and has a manned launch scheduled in under a month (May 27th). And you’re substantially overconfident on Starship, given it just passed its pressure test for the first time.

  25. Alex319 says:

    Prediction 12 (Scott personally getting coronavirus) seems very high given his other predictions. If 30% of the population gets the virus, and IFR is about 1%, that would mean about 1 million deaths which is a lot higher than he expects the number of deaths to be. And I would guess that Scott has a much lower probability of getting coronavirus than the average member of the population (he’s probably taking a lot more precautions than most people, and he can work from home)

    • Randy M says:

      I thought that sounded high, too. I wonder if he’s working from home or going to see patients in person?

    • keaswaran says:

      I think most people are assuming that IFR is actually quite a bit below 1%, given that CFR is somewhere in the 0.5-2.5% range, and most infections remain undetected.

      • doubleunplussed says:

        Low CFRs in some countries might be evidence of low IFRs in those countries, but from what I can see these countries have one of the following attributes: a) young population, b) low death counts such that randomness plays a big role, or c) Are the kind of place I wouldn’t trust to report deaths accurately.

        So I think that CFRs of 0.5 % that exceed IFRs by a factor of 10 or more are not plausible except in places with a very young and healthy population.

        I see people converging on an IFR of about 0.5 – 1% for the US as a result of the NY antibody studies. And if you think New Yorkers are healthier or younger than other parts of the country, you might want to look at the upper end of that range to extrapolate to the rest of the country. If I had to bet, I would put the 1 sigma range of the US’s IFR at 0.8 – 1.2%.

  26. Pandemic Shmandemic says:

    US has highest official death toll of any country

    What is the point of addressing the US death toll in absolute rather than per-capita numbers ?

    • thisheavenlyconjugation says:

      It’s interesting to predict. The chance of the US having the highest toll per capita is uninterestingly low, even if you exclude very small countries.

    • Kindly says:

      I mean, a prediction’s a prediction. If Scott’s 50% estimate of 300,000 US deaths happens to be exactly right, it’s a lot like predicting “no country except the US will reach 300,000 official deaths”.

      Compared to that prediction, it’s just hedging one’s bets against the death toll being much higher than expected everywhere, including the US.

      Lots of people do actually compare the absolute US death toll to other countries. This makes the statement an interesting one whether or not it’s a useful comparison. And to try to guess how likely it is to happen, we don’t need it to be a useful comparison.

  27. salvorhardin says:

    “Bay Area lockdown still in force after June 15” seems very sensitive to the definition of lockdown. If your definition is literally just “dine-in restaurants closed = lockdown” then I agree, but do you think for example that with p = 0.6 we will still be in Stage 1 and not Stage 2 of the staged roadmap Newsom outlined?

    • tgb says:

      Also, the “before election day” prediction is ambiguous about a possible second lockdown. If the lockdown lifts but comes back after a fall surge, would that count or not? I think there’s a >10% chance of this happening, so Scott should probably clarify.

  28. Doctor Mist says:

    Feel free to get in a big fight over whether 50% predictions are meaningful.

    I’ve been trying to figure out whether predictions under 50% are meaningful. I would have saif that a prediction like

    Fewer than 100,000 US coronavirus deaths: 10%

    is not actually a prediction at all. If the death toll is 75,000, you don’t intend to point at this and say, “See, I predicted it; I didn’t give it a high probability, but I did predict it”. Right?

    In other words, is this prediction equivalent to

    More than 100,000 US coronavirus deaths: 90%

    or to

    Fewer than 100,000 US coronavirus deaths: I predict it but with very low confidence.

    • SEE says:

      As his post itself says:

      I’m using the full 0 – 100 range in making predictions this year, but they’ll be flipped and judged as 50 – 100 in the rating stage, just like in previous years.

      So, it’s the same prediction as

      More than 100,000 US coronavirus deaths: 90%

  29. meh says:

    i think my comment was eaten.

    is 20. at all conditioned on prevalence of mail in voting?

  30. Atlas says:

    Steve Sailer commented on the scoring of 2019 predictions that there was, naturally enough, no prediction about the likelihood of a novel coronavirus pandemic. That made me wonder: Is it worth having a “something ‘unexpected’ and very important happens” prediction? It seems like every ~10 odd years for the past few decades, there’s “unexpectedly” been a big change—the collapse of the Soviet bloc 1989-1991, 9/11 in 2001, the financial crisis in 2008-9 and the Arab Spring in 2011, and now the coronavirus in 2020. So maybe there’s a yearly ~10% chance of a big unexpected change happening? Does it make sense from the outside view to think about big “known unknown unknowns” that might happen?

    • Edward Scizorhands says:

      Maybe Scott needs predictions for the upcoming decade in addition to the upcoming year.

      Of course, who knows if we’ll still be hanging around this blog in 10 years.

    • keaswaran says:

      It definitely makes sense to consider these things! But it’s likely going to be hard to write down a prediction that gives an unambiguous answer on Dec. 31. Probably on Dec. 31, 2001 we would agree that “big unexpected event” happened. Would we have agreed on Dec. 31, 2011? Or on Dec. 31, 2004 (just a week after the tsunami, which was the deadliest natural disaster in history)? Probably even on Dec. 31, 1989 there would have been some debate, because some people would have said the collapse of the Iron Curtain was on everyone’s radar already, and the fact that Tiananmen didn’t lead to anything was evidence that nothing beyond the predicted collapse was happening.

      But I think people making other predictions always have to take these sorts of things into account. You can’t give more than 90% confidence on a presidential election outcome in January for this sort of reason.

  31. David Condon says:

    Let me see if I can do any better:

    CORONAVIRUS:
    “1. Bay Area lockdown (eg restaurants closed) will be extended beyond June 15: 60%”
    I don’t know enough about this to say.
    “2. …until Election Day: 10%”
    This seems too high. I’ll go with 5%
    “3. Fewer than 100,000 US coronavirus deaths: 10%”
    Way too low, I’ll say more like 40%
    “4. Fewer than 300,000 US coronavirus deaths: 50%”
    90%
    “5. Fewer than 3 million US coronavirus deaths: 90%”
    99%
    “6. US has highest official death toll of any country: 80%”
    Agree with this
    “7. US has highest death toll as per expert guesses of real numbers: 70%”
    I’m not sure who the expert guessers are supposed to be
    “8. NYC widely considered worst-hit US city: 90%”
    95%
    “9. China’s (official) case number goes from its current 82,000 to 100,000 by the end of the year: 70%”
    50%
    “10. A coronavirus vaccine has been approved for general use and given to at least 10,000 people somewhere in the First World: 50%”
    25%
    “11. Best scientific consensus ends up being that hydroxychloroquine was significantly effective: 20%”
    Agree
    “12. I personally will get coronavirus (as per my best guess if I had it; positive test not needed): 30%”
    5%
    “13. Someone I am close to (housemate or close family member) will get coronavirus: 60%”
    Don’t know how many you are close to, but if 12 is 30%, then this is clearly too low, anyways, 20%
    “14. General consensus is that we (April 2020 US) were overreacting: 50%
    15. General consensus is that we (April 2020 US) were underreacting: 20%”
    Not sure what you mean by “we” or who determines the general consensus. In an election year, Democrats will say Trump was underreacting, and Republicans will say he was reacting appropriately. So my prediction…
    14.5 A general consensus will not yet be achieved on whether Trump was over or underreacting. 70%
    “16. General consensus is that summer made coronavirus significantly less dangerous: 70%”
    Agree
    “17. …and there is a catastrophic (50K+ US deaths, or more major lockdowns, after at least a month without these things) second wave in autumn: 30%”
    10%
    “18. I personally am back to working not-at-home: 90%”
    Can’t comment
    “19. At least half of states send every voter a mail-in ballot in 2020 presidential election: 20%”
    Agree
    “20. PredictIt is uncertain (less than 95% sure) who won the presidential election for more than 24 hours after Election Day. 20%”
    Agree

    POLITICS:
    “21. Democrats nominate Biden, and he remains nominee on Election Day: 90%”
    99%
    “22. Balance of evidence available on Election Day supports (as per my opinion) Tara Reade accusation: 90%”
    can’t comment on your opinion
    “23. Conditional on me asking about Reade on SSC survey, average survey-taker’s credence in her accusation is greater than 50%: 70%
    24. …greater than 75%: 10%
    25. …greater than credence in Kavanaugh accusation asked in the same format: 40%”
    dunno
    “26. Trump is re-elected President: 50%”
    30%
    “27. Democrats keep the House: 70%”
    “28. Republicans keep the Senate: 50%”
    70%
    “29. Trump approval rating higher than 43% on June 1: 30%”
    50%
    “30. Biden polling higher than Trump on June 1: 70%”
    90%
    “31. At least one new Supreme Court Justice: 20%”
    10%
    “32. I vote Democrat for President: 80%”
    can’t comment
    “33. Boris still UK PM: 90%
    34. No new state leaves EU: 90%
    35. UK, EU extend “transition” trade deal: 80%
    36. Kim Jong-Un alive and in power: 60%”
    don’t follow foreign politics that closely

    • EchoChaos says:

      Only a 10% chance of one of nine people, several over 80 or having major risk factors, dying in the middle of a massive pandemic that is hitting DC hard?

      I think 20% is perhaps even low for a new SC Justice. Mitch McConnell has already promised that he will confirm a Justice if a vacancy arises in 2020.

      • David Condon says:

        Most estimates put the number of deaths below 100,000. Granted, there are limitations to those models, but to get to a really high number would require either an extremely drawn out recovery period or a second wave to hit; neither one of which has appeared to occur yet in countries hit earlier by the virus such as China or Italy. Scott’s estimates seem to depend pretty heavily on a second wave occurring without a treatment method being found first. I put a low probability on such a combination of events. While a vaccine is a long ways away, figuring out a way to treat the illness seems to be nearer at hand.

        On the Supreme Court, I think you’re forgetting what happened the last time a Supreme Court Justice was up for nomination during an election year. The Democrats in Congress certainly didn’t, and neither did the Supreme Court Justices. The justices are aware that they would probably not be able to get away with a nominee being approved before the 2020 election, and this will alter their decision.

        • EchoChaos says:

          No Justice is going to resign in 2020 with 90+% probability, agreed. But there is a pandemic ripping through DC and it’s hit several Congressmen/Senators already. Ginsburg/Thomas/Breyer/Sotomayor are all high-risk category for COVID.

          The idea that none of them dying has only a 10% chance of happening seems low, given that the death rate for 80+ year old people in any given year is close to that high anyway.

          And Mitch McConnell has explicitly said he would move forward with a nomination if a vacancy arises. The Democrats cannot stop him, although they will absolutely scream about it. If they can get 3 Republicans to defect they can stop him, but I doubt they can.

          • David Condon says:

            The issue is whether a new justice will be sworn in before January 1st. Mitch McConnell saying he would fill the job is not the same as saying he would overrule a filibuster. So a Justice has to die, and then Trump has to appoint a new nominee, and then said nominee has to go through the judiciary committee, and then Mitch McConnell has to overrule a filibuster, and then the Republicans have to agree to the nuclear option (including the ones who are at risk of losing their Senate seat), and then the Supreme Court has to agree to seat the new justice before January 1st. A lot of Republican Senators on the bubble will probably not go against McConnell openly, but they will oppose overruling the filibuster behind closed doors. Letting the filibuster stand is probably the Republicans’ best electoral move. It’s not just a Supreme Court Justice dying that fulfills the requirements here. And that’s a lot of steps that have to happen in a 7 month period. Calculate probabilities for each of those events separately, and then multiply together to get the final estimate.

          • EchoChaos says:

            @David Condon

            You have the details wrong. The filibuster was already abolished for Gorsuch and didn’t apply to Kavanaugh at all, so McConnell only needs 50 Senators and there are no other legislative hurdles. There is no more filibuster.

            Kennedy announced his retirement on June 27th and Kavanaugh was sworn in on October 6th, meaning the process took ~3 months (July, August and September).

            Assuming McConnell can’t accelerate that schedule (and given the criticality of the seat, he probably can), that means that a Justice would have to retire or die before about the end of August. Not impossible, but one of nine people, two of whom are over 80, dying in 4 months is certainly more than 10%.

        • keaswaran says:

          > to get to a really high number would require either an extremely drawn out recovery period or a second wave to hit; neither one of which has appeared to occur yet in countries hit earlier by the virus such as China or Italy.

          China seemed to complete a very thorough lockdown that actually brought the daily number of new cases down to the double digits. Italy is only a week or two ahead of the United States on the curve, so I don’t see why the fact that they haven’t had “an extremely drawn out recovery period or a second wave” really doesn’t tell us anything. Especially not about what happens at any point in the next six months before the election. Recall that the first human infection was probably some time in early to mid November last year, so we are only just now reaching the halfway point from the very first human case and the election.

  32. Enkidum says:

    I can’t imagine what it would take for you not to vote Democrat. I guess you don’t really live in a swing state, but it seems very improbably given your previously-stated commitments.

    • Anteros says:

      I was going to say something similar. Unless not voting at all is more likely than we assume.

  33. benf says:

    Which “Tara Reade allegation” are you actually referring to? She’s had multiple independent allegations, so in the spirit of the prediction you should be specific. As far as I know there’s “Staffers in his office made demeaning remarks about me but he was never actually meaningfully involved” and “Oh right I forgot also he forcibly fingerbanged me”. There may be others as well. So you really need to be specific.

    • Enkidum says:

      It is the latter.

      • benf says:

        I haven’t seen him actually clarify his own prediction, I’m not sure why you think you can do it for him.

        • gbdub says:

          You know what he meant, you’re just snarkily adding your voice to the “I don’t find Tara Reade credible” crowd with a side order of veiled insult to our host. Less of this, please (the snark of course, you’re entitled to your opinion on Reade).

          • benf says:

            Yeah but I DON’T know what he meant, because he wasn’t specific and there actually are multiple allegations at issue. If her allegation of sexual harassment by her immediate supervisor is substantiated, but the more dramatic allegations are not, how does he plan to score the prediction? If you’re not actually clear about what your predictions ARE, doing the whole “percentage confidence scoring” thing is pointless.

          • Controls Freak says:

            @benf

            You might be new here, so I would suggest looking into common knowledge, Schelling points, and Sufi Buddhism-lite. If you didn’t live through the last Reign of Terror, don’t ask about it. Just know that you need to pick up the effort level or it will find you.

  34. fion says:

    Interestingly low credence in Kim Jong-Un being in power. Is there something going on in NK that I’ve not heard about? [EDIT: nevermind, I saw some comments above. There are rumours he’s dead? Crikey.]

    On a more trivial note, I do wish everybody would stop calling Boris Johnson “Boris”. He’s not your friend; he’s not even a talk show host. He’s a politician, and the convention is to refer to politicians by their surname or full name. It just plays into his “cuddly eccentric” persona to call him by his first name.

    • zzzzort says:

      I think there’s latitude when you have a relatively uncommon first name and either a very common last name (as with Bernie) or a last name shared with other members of your political family (as with Jeb and Hillary). BoJo has both a very common last name and a brother in parliament.

    • Edward Scizorhands says:

      Your link is busted. Is it this? https://aapsonline.org/hcq-90-percent-chance/

      AAPS is not some random sample of doctors. They are a conservative organization. https://en.wikipedia.org/wiki/Association_of_American_Physicians_and_Surgeons

      I am not saying “they are conservative, don’t listen.” I am saying “don’t trust them on an appeal to authority of a representative sample of doctors.”

      • detroitdan says:

        Thanks. I’ll try again to post the link (it seems to be the same story you posted, perhaps even word for word, but from a different source):

        Hydroxychloroquine Has about 90 Percent Chance of Helping COVID-19 Patients, States Association of American Physicians and Surgeons (AAPS)

        Thanks for the feedback on the AAPS. I distrust them after reading about them at the link you provided.

      • Tatterdemalion says:

        According to wikipedia,

        “…its publication advocates a range of scientifically discredited hypotheses, including the belief that HIV does not cause AIDS, that being gay reduces life expectancy, that there is a link between abortion and breast cancer, and that there is a causal relationship between vaccines and autism.”

        • Lambert says:

          Abortion vs going to term or vs never getting pregnant?
          Because I think pregnancy would be a big risk factor for breast cancer.

          • Edward Scizorhands says:

            There was a really big fight about 30 years ago, with the standard culture war camps doing that you think they’d do.

            I’d be amazed if pregnancy didn’t significantly alter, in some direction, a bunch of life outcomes. And obviously abortion would matter for some portion of those.

            But the culture war totally mind-killed the entire scientific discussion.

          • Lambert says:

            Well the breast cancer rate increases from the pill are due to the body thinking it’s a bit pregnant and making the boob cells divide more. Pregnancy ought to be that on overdrive.
            Conversely, not ovulating reduces risk of ovarian cancer.

        • that being gay reduces life expectancy

          Given the existence of AIDS, that’s surely true.

          • keaswaran says:

            Note that that report was from 2011, based on deaths over the years 1988 to 2011. When you just look at data from the year 2011 itself, they find that the life expectancy of a 20 year old *with HIV* is 70.

            https://www.healthline.com/health/hiv-aids/life-expectancy

            (I can’t tell if that is 70 total, which represents a lower life expectancy than the average 20 year old, or 70 years longer, which is comparable to the life expectance of the average 20 year old.)

            In the years between 1994 and 2011, several new HIV treatments were discovered which have changed HIV from meaning that one is likely to get AIDS and possibly die, to meaning that one has to take a daily pill and can otherwise ignore it. (Note below the gigantic drop in AIDS deaths in the late 1990s.) That this pill *also* makes it basically impossible to transmit HIV infection, so that the likelihood of a random gay man getting HIV has gone down quite a bit during that time, in addition to HIV being much less dangerous now.

            https://en.wikipedia.org/wiki/HIV/AIDS_in_the_United_States

  35. meh says:

    can these be turned into a survey to evaluate the community?

  36. Tatterdemalion says:

    I’m not convinced that flipping sub-50% probabilities is a good idea.

    Yes, it gives you a larger sample size at each of the probabilities you look at. But while “k% chance of X” and “(100-k)% chance of not-X” are equivalent logically, they’re very different psychologically.

    I can very easily believe that a predictor might be subconsciously affected by the framing of questions, and systematically over-optimistic or over-pessimistic about the thing they are asked about, and if you asked them “what is the probability of X” and “what is the probability of not-X” on different days, without them remembering that they’d been asked before and connecting the questions, they might typically give answers that sum to less than or more than 100%.

    But if 0% of your 20% predictions and 60% of your 80% predictions come true, and you make the same number of each, then using flipped probabilities you will register this as 80% of your 80% predictions coming true, and believe yourself to be well-calibrated when actually you clearly aren’t.

  37. cevapcici says:

    ‘Best scientific consensus ends up being that hydroxychloroquine was significantly effective: 20%’

    Why so low when much of the world is already using it – and continuing to use it – as one of their go-tos?

    • Lambert says:

      Go read all Scott’s posts tagged ‘replication crisis’.

    • keaswaran says:

      Source? Do you just mean that this is one of the five or six drugs that are being given to people because they’re the only ones with any hope of doing anything? Or do you mean that there are some hospitals where they actually treat hydroxychloroquine as substantially more likely to help than remdesivir or serum therapy or other alternatives?

  38. TheTurtleMoves says:

    Why do you even post the redacted stuff? Why not just list that stuff on your own private doc for your own edification? I don’t get what the point is?

    Is it because sometimes you can reveal what it was?

    I mean you could just write at the bottom: “I have some I’m not telling you about but if I can later I will and you’ll just have to trust me”

  39. TheTurtleMoves says:

    73. I have specific, set-in-motion plans to quit work / start my own business: 5%

    Just
    publish
    Unsong
    on
    Amazon
    Kindle Direct.

    I swear it would do amazing and support you for at least a year. I almost feel like I could promise this.

    • Bugmaster says:

      Will it, though ? How much of a cut does Amazon take ?

      • Doctor Mist says:

        Not so much, really, at least compared to real publishers. For a Kindle book you have a choice between a 30% royalty program and a 70% royalty program. The latter obliges you to publish exclusively through Amazon, though you can withdraw at regular intervals (I’m thinking monthly, but I don’t really remember), and the royalty is on the “net”: list price minus any VAT and minus Amazon’s “delivery fee” of fifteen cents per megabyte (the book I just published, about 300 pages, was about 3/4 of a megabyte; I don’t know how long Unsong would be). For 70% you are also obliged to price the book between $2.99 and $9.99.

        But with a $2-$7 royalty, Scott would have to sell a lot of copies to “support himself”. I don’t remember how big SSC readership is, but I don’t think selling one to each of us would cut it.

    • Douglas Knight says:

      Scott is a part-time writer and a part-time psychiatrist. As one of the lowest-paid psychiatrists in the country he makes more money than as one of the highest-paid writers (which he already is without selling writing). This prediction is about starting his own medical practice to increase that income. (And maybe more for other reasons than money, like flexibility. His current employers are OK that he’s part time, but employers don’t usually like that.)

  40. Incandenza says:

    It’s too bad you forgot to do these predictions before the coronavirus went bananas. 2020 is shaping up as a real black swan of a year; would that have led your predictions to be wildly off? Or would it not actually have had much of an effect? Either way it would have been interesting to see.

  41. Caledfwlch says:

    Neat predictions. Looking forward to seeing the results

  42. Douglas Knight says:

    On January 1, I will still believe that there had already been 100k US coronavirus deaths on 29 April: 80%.
    On January 1, Scott will believe that: ?

  43. pjiq says:

    Counter predictions:

    Dow below 25,000 on Jan 1 2021: 70%
    S&P closes at least one day below current 2020 lows: 70%
    US official death toll (before corrections based on uptick in deaths) is less than 150,000: 70%
    Cure for covid19 that reduces fatality by 60% or better found by October 1st: 70%
    Unemployment still > 10% Jan 1 2021 in spite of improved coronavirus situation: 70%
    Escalated tensions btwn US and China (defined as higher tariffs or a military skirmish of some sort): 70%
    Someone other than Joe Biden elected: 70%
    If Trump loses the election, lots of drama before he’s gone (riots, Trump claims bad vote count, etc): 70%
    General consensus is we overreacted to covid19: 70%
    Kim Jung Un dead and replaced by someone more stable with closer ties to China: 70%
    Majority of US states no longer under lockdown September 1st. 70%
    Some new scary thing (separate from Coronavirus) is uncovered/ originates out of China: 70%
    Chinese economy comes out of recession faster than US economy: 70%
    But at least no nukes detonated on cities or ww3 started yet: 70%

    Personal:

    If I meet slatestarcodexguy and we have an argument regarding whether or not 50% predictions are lame, I win said argument: 51%
    If I meet slatestarcodexguy and we have a wrestling match, i win said wrestling match: 40%
    If I meet slatestarcodexguy and we have a duel to the death with swords, i win said duel to the death with swords: 90%
    If I meet slatestarcodexguy and we measure our respective brains in order to win the affections of a beautiful woman, said beautiful woman will ride off with me into the sunset: 3%.
    If I meet slatestarcodexguy and we have a thumb war, I will decisively win said thumb war: 1000000000000000%

    • bullseye says:

      What exactly would brain measuring entail? Are you going to use skull size as a proxy for brain size, or something more invasive?

      • pjiq says:

        Ah good point, not the easiest to measure, at least before death. If we both got an MRI we could estimate the respective volumes. But I do feel I should disclose: my small brain intended that as some kind of bizarre joke- I have no intentions of actually losing such a contest to the great slatestarcodexguy any time soon.

        • bullseye says:

          Ok, but your 1000000000000000% chance of winning the thumb war is serious, right?

  44. wayanad Empire says:

    I like your wesite, My thanks for doing such a great information
    Wayanad Empire

  45. goundo says:

    Based on how well informed you are about other things, I’m rather shocked that your planned exercise regimen appears to consist half of sit-ups. Perhaps you have some other perspective, but based on nearly everything I’ve read and practiced, sit ups are kind of a terrible choice of exercise.

    1) The movement you are practicing has little real world application – it simply makes your abs look nicer. Really, how often are you required in life or sport to forcefully flex your spine? Stabilizing your spine, sure, but sit ups don’t help with that.

    2) They’re bad for your back and tighten up likely already tight hip flexors.

    Honestly, I’m conflicted, since generally getting any exercise is better than getting none at all. But due to issue (2), it’s possible that’s not the case, and my experience is that sit ups are of little value for improving any kind of athleticism or general health. There are *a lot* of exercises you can do at home, alone, with no equipment – even more if you are open to going outside with friends and using equipment. Might I suggest glute bridges instead? Or if you’re really committed to that 6 pack, prone leg lifts with the hands supporting the low back?