Open Thread 95.5

This is the twice-weekly hidden open thread. As the off-weekend thread, this is culture-war-free, so please try try to avoid overly controversial topics. You can also talk at the SSC subreddit or the SSC Discord server.

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

446 Responses to Open Thread 95.5

  1. Hence says:

    Hello there. For those who missed it: significance testing (and p-values), and why the practice should be dropped.

    Been a while that I wanted to put the pieces in one place, and this was the idea here.

    Comments? Tell me how you like it.

    (Also consider sending it to some of your “possibly recoverable” researcher friends)

    • bean says:

      Narrowly, you don’t propose any alternative, which is a massive problem in getting people to accept your argument. When I read an article like that, I want to know what I should be doing instead. All you give is some platitudes about “critical thinking”. I’m sure that some of your links contain alternatives, but I don’t feel like hunting for the answer through a bunch of them.
      Broadly, I strongly dislike the tone. You’re acting like p-value is totally useless and borderline immoral. It’s certainly not a number without limitations, but I’ve never yet heard of any metric anywhere that gave a true and complete picture by itself. And anyone who might consider your case (the sort of people who see p-hacking as a problem instead of as a way to get results) already knows it’s not perfect. Proposing abolition (to the point of calling for a total boycott) without giving a better tool is just not helpful.

      • fr8train_ssc says: was suggested on that page. TL;DR: Use a modified version of effect-size calculation that also weighs against sample size, so that there’s less apples-to-oranges comparisons of different studies with different sample sizes for what to determine as truth.

        FWIW, I agree 90% with that article. p-values should be left to the domain of calibration experiments, and even then, with a more stringent threshold of truth. Otherwise we end up with people claiming humans have ESP because they can see porn from the future (

      • Hence says:

        @bean @Nabil @Aapje and others, thank you for the open criticism and questions. Answers on the new FAQ.

        • Nabil ad Dajjal says:

          Pretty much every paper you cite has a sentence like this somewhere in it:

          “There is no one statistical tool which works in every situation; scientists should employ a toolbox of statistical tools and heuristics.”

          Which leads to natural questions. What sort of tools are appropriate to employ in commonly encountered situations in research? What are the limitations of each of those tools?

          Your answer to those questions, judging by your FAQ, is “fuck you.”

          The problem of researchers hammering away with NHST won’t be solved by trading their hammers for empty toolboxes.

          • quanta413 says:

            Have you read Andrew Gelman’s blog? In a lot of cases, there is no statistical toolbox that will salvage a piece of research. His analogy is that a lot of studies are trying to weigh a feather by placing it in the pouch of a kangaroo and then having the kangaroo jump up and down on the scale. No statistical skills can save this sort of experimental design. The solution is better experimental design which is field specific and coming up with better research questions which is also field specific.

            A lot of research should probably just not be done. A different question should be studied that’s in the grasp of the available experimental methods. The other answer is that journals should stop focusing on publishing based upon results and instead look at the methodology and the question.

            So maybe the FAQ should focus on how to use statistics to help estimate if a question is one you can almost certainly answer or get meaningful data for.

          • Nabil ad Dajjal says:


            I haven’t read his blog. I’ve read some of his publications, including some referenced here, but I could never get into his blogging. Maybe I should give it another shot.

            If I’m understanding correctly, he’s saying something like one of my professors: “if you need to use statistics [to demonstrate an effect], it’s probably bullshit.” That a well designed experiment will produce a dramatic result.

            The problem with this is that, in a field like biology, not every result is dramatic. Small changes in complex systems can have very important effects even though the changes themselves are difficult to detect. This restriction would push us back into the early aughts; that means no polygenic traits, no epigenetics, and no -omics of any kind.

            If the proposal is to do statistics better, sign me up! Just show me how to do it better and I will. But that proposal actually calls for some work by the one making it.

          • Aapje says:


            Then what scientists/journals/etc need is a way to figure out what experimental designs don’t work. Just saying: ‘don’t do bad science’ doesn’t cut it.

          • quanta413 says:

            @Nabil ad Dajjal

            If I’m understanding correctly, he’s saying something like one of my professors: “if you need to use statistics [to demonstrate an effect], it’s probably bullshit.” That a well designed experiment will produce a dramatic result.

            That’s definitely not what he’s saying. Hopefully how I interpret what’s he’s saying will be more clear after I respond to your second point.

            The problem with this is that, in a field like biology, not every result is dramatic. Small changes in complex systems can have very important effects even though the changes themselves are difficult to detect. This restriction would push us back into the early aughts; that means no polygenic traits, no epigenetics, and no -omics of any kind.

            My field is biology so this is great. First, small is only relevant to the error/noise so let’s assume that for whatever reason you can’t improve your experiment enough to shrink that compared to the effect you are looking for. There are many ways around that, but I’m only going to cover broad general classes of ways around it. I’ll go from easiest to hardest cases experimentally. The math or theory may be “hardest” for the first case, but that will take a lot less time than the other cases.

            The easiest case is if your theory predicts not just the mean value of some X you are looking for but also the distribution of X. You can then often overwhelm the problem by sheer brute force. If your theory is right, as you collect more data the distribution of X’s you measure should converge to the predicted distribution. Then you can apply pretty standard methods to estimate the parameters of the distribution and the standard error of your estimates. You should use something like a measure of mutual information to make sure your data is actually well described by the distribution or this isn’t going to work. A kolmogorov smirnov test is one option. Of the top of my head, if you’re not sure how “small” the distance should be, draw samples of the same size as your data from the predicted distribution with your estimated parameters and compare the distance of those from the exact distribution to the distance from your data to the exact distribution. Ideally, your data should not do much worse than a typical draw from the estimated parameters.

            Ok, but a lot of the time we don’t know what the distribution should look like. In this case, if the noise or systematic effects are high compared to the signal, you must use multiple techniques with different systematic errors to measure the quantity of interest. The usage of multiple techniques tends to be favored by microbiologists. Each technique should have its own controls that give you at least a vague idea of its systematic error. For example, say you want to measure the level of expression of a protein but you know either the expression level is low or the error is high. You can use the following techniques (not exhaustive). You can translationally fuse the protein to a fluorescent protein to image it to measure its concentration in live cells using fluorescence microscopy, you can do reverse transcription of RNA extracted from the cells followed by quantitative PCR to measure RNA levels (which are correlated with protein levels), and you could modify the protein with an antibody tag and extract protein from the cells and pull down the one you want with antibodies then do blotting tests (I am least familiar with performing this last technique). The important point is that these techniques are all physically distinct in mechanism so from having 3 estimates you get a better idea of your systematic error. If you’re a master of causal modeling, you may have some clever way to narrow down what combination of estimates is best but I don’t think it will usually add a lot of value.

            Ok, worst case scenario. You don’t know the distribution, you only have one technique to apply, and systematic errors you don’t understand are running rampant despite your mighty efforts at dozens of controls. First don’t expect miracles. Anything you do is pretty likely to turn out wrong; you’re in an exploratory stage. So what can you do? One thing is you can still give tentative upper and lower bounds. The second thing you can do is attempt to publish your basically null results honestly with pointers to what factors need to be controlled to improve the results. You can probably rule out something even if you can’t rule anything in. None of these will be very exciting and you’re not going to get into Nature or Cell. But it can help future scientists who have new techniques decide if your problem is now within reach of a new technique. The second thing you can try is to reduce the scope of your question; I don’t have a good example of this off the top of my head. The third thing you can do is terminate the work and move on to a different experiment (this is what I did when I got stuck); this totally blows, but you may learn a new experimental technique in 6 months, you may improve your own technique, or you may figure out how to do the second thing above while you’re working on other stuff. The one thing you shouldn’t do is go full steam ahead, make some dubious claims, and pollute the field with noise mining research. Yes, this may benefit your career but the equilibrium we’re in is terrible. You won’t be ruined by not doing this anyways. Go with the more modest and honest publication strategy. If you entered science for fame or money, god help you because you already made a terrible mistake.


            Hopefully the above clarifies what I’m thinking of for you. But more broadly, I’d say there is no royal road. Some experiments that were well done and looked like a lock are going to turn out as just ????. The solution to this is social (journals should focus more on the methodology and the question so that these results can get published, not just on the result reached), but I don’t really know what to do here. I have some ideas about making this case less common above, but it’s going to happen no matter what.

          • Nabil ad Dajjal says:

            I guess I didn’t make it clear; I’m also a biologist. That’s why I used it as an example.

            I suspect that we’re talking past each other because most of the steps you mentioned are examples of good experimental design. I’d take issue with estimating the abundance of a protein from qPCR data rather than trying to measure it directly, but that’s a nitpick.

            I went to his blog and I’m looking at the chain of posts leading up to the kangaroo quote. Luckily he links one of his papers on the subject (Gelman & Carlin, 2014) so I don’t have to read too much blogging. It looks like he’s talking about the related topic of misuse of power calculations. Boiled down, he makes three common sense points:

            1. Post-hoc power calculations are still bad idea and misleading; calculate power before you perform an experiment.
            2. Don’t just pick whatever effect size makes power the largest for a given sample size; be conservative and pick plausible effect sizes.
            3. Don’t do insanely low-powered studies or you’ll be eaten by a grue; your studies should at a minimum have power > 0.5.

            This doesn’t seem objectionable and is only tangentially related to the NHST discussion so I’ll just agree to agree and leave it at that.

          • Aapje says:


            I think that we need a check on a far more basic level, which pushes people to do what you suggest. That check is replication.

            What seems to happen in some disciplines that bad experiment setups get used again and again, where many give positive results due to one of the many ways in which positive results can be falsely generated. See IAT for an example.

            When replicating, it is far harder to generate the same positive results for these bad experiment setups, so this is then often revealed.

            Then step 2 is that people fix their experiments. I don’t think that you can get some disciplines to jump to step 2 right away, because of their culture.

          • quanta413 says:

            @Nabil ad Dajjal

            Ok, didn’t realize that was your field. Using multiple independent measurements for one thing is indeed obviously good experimental design, and I think most biology papers do reasonably well with this. I listed it because my impression from Gelman’s blog (which is mostly about social science) is that a significant chunk of social scientists who manage to get published in high prestige journals don’t clear this bar. It would seem a mistake to leave it out then.

            I think my first suggestion (test against the whole distribution not just summary statistics) is definitely not standard in my subfield or in the a lot of ones I’m familiar with; it may be in yours. Some people do and some don’t. Some of the “don’ts” have really egregious mistakes in estimating parameters even when the experiments appear competently executed. Things like using the empirical mean as an estimate for a parameter of interest when even in theory that parameter is not equal to the true mean (and will differ by factors of 2, 3, or 10). It doesn’t apply if your data has binary values and thus by definition is binomial distributed over multiple trials. It’s also not super useful on its own if your data is normally distributed (because that could happen for any number of reasons). But there are cases where something should be distributed in some particular nontrivial way, and people just eyeball a histogram and compare means when they should compute mutual information, do some bootstrapping, etc. It’s definitely pretty far from the NHST framework too.


            I agree that replication would help, but it’s not incentivized as is, which is why I say the problem is partly social. Reliably getting a journal to publish your giant replication that says some previously published effect doesn’t exist (or even does exist) is a pain in the ass, and currently it takes way too long fighting with the authors of the original if the effect doesn’t replicate. My hat is off to those who are willing to engage in this slog though. However, if we keep replicating poorly designed experiments, we’ll just see a random walk in estimates over time rather than seeing progress so it’s also important to get people to understand how to do things correctly.

          • Aapje says:


            Biology and many other disciplines may have the internal motivation to develop and enforce better norms themselves.

            However, I think that some disciplines have a broken culture that favor desired results over correct results. The advantage of replication is that it is very easy to understand for lay people and also not a demand that is unreasonable. So society can demand it from scientists, without those scientists being easily able to dismiss it as an ignorant demand.

          • Hence says:

            Thank you, Nabil. See updated FAQ for note on hammers. Specific references there too about CI and Bayes. And the previous comment about reviewing inferential statistics reminded me to suggest this recent article. I hope all useful.

        • skef says:

          An attempt to give one kind of constructive focus the push-back you’re getting:

          Let’s grant, for the sake of discussion, your view that only a variety of statistical approaches can provide an appropriate minimum toolbox. In that case, it seems that what would be needed to go forward is a practical (and not cognitively overwhelming) guide for paper evaluators (including reviewers). Off the top of my head, such a guide would need to include at least these topics:

          1) What tools does every reviewer need to be familiar enough with to evaluate papers that claim experimental results of scientific interest? (This need not cover every possible or actual paper, but needs to cover enough that the process of evaluation itself doesn’t break down.)

          2) How is the problem of “statistical approach hacking” to be addressed? That is, how should an evaluator judge whether one approach was chosen over others because it suggested significance (however vaguely defined) while others didn’t?

        • bean says:

          I don’t feel that’s a good answer, though, because the burning building analogy doesn’t work. A building works fine when it’s not on fire, then suddenly becomes on fire, and you need to get out quickly. We’ve made a lot of progress using NHST, and I don’t see how it could have suddenly caught fire. So I do think it’s reasonable to hold out for a better tool before abandoning it.

          As best I can tell, you seem to think that NHST is actually worse than having no statistics at all. Which seems absurd to me. At the very least, they raise the bar for hucksterism, and give people who are somewhat cynical good clues to games people are playing. (This seems implausible. What’s the p value? 0.045. Ah. I smell p-hacking.)

          Edit: Above all, I’ll repeat. You’re not likely to have much luck recruiting for your crusade against p-values without offering something to crusade for, instead of just crusading against. That may be a couple of different methods instead of just one magical replacement, but I personally am not really interested until I have a replacement. Give me a good one and convince me of it, and I’ll sign up. But I’m not at all sure that something better will spring up from the wreckage of the existing system if you just smash it.

          • Hence says:

            The fire metaphor works if by “you” you think “society”. I changed the analogy anyway.

            > You’re not likely to have much luck recruiting for your crusade against p-values without offering something to crusade for, instead of just crusading against.

            Not recruiting for a crusade. People read, make their minds. The crusade, if there’s one, is for better science, here through removal of a bad tool.

            Seems you believe NHST is flawed but overall does a positive job.

            Wreckage: whoever feels lost without NHST might be too dependent on it.

    • Nabil ad Dajjal says:

      I agree with bean above. It’s not enough to oppose NHST, you need to at least suggest alternatives.

      The last time you posted this comment, I asked your opinion on confidence intervals and Bayes factors. Would either of these satisfy as a replacement or are they also objectionable?

      That said, I don’t want to be entirely critical. Even as researchers continue to use NHST it’s good to raise awareness of its limitations. Also your comment last time got me off my butt to actually read some introductory material on causal inference, which seems interesting so far.

    • Aapje says:


      The issue is still that the scientific community has become addicted to p-values as a clear target to aim for and to judge by. It is quite understandable why, because it gives scientists a way to decide how many subjects they need for their study & some confidence that their papers will be accepted and it gives journals a way to separate the wheat from the chaff. That it doesn’t work well is rather inconvenient, but it’s doubtful whether they will give it up unless they are offered something better.

      I don’t see that in the criticism, so I fear that this is not a good way to deal with the issue.

      I think that a better way is first to solve some other issues, like the lack of replication, the bias towards significant results, etc. When those are addressed, p-values automatically become less important and thereby easier to abandon.

      EDIT: darn, ninja’d

      • Ideally, one needs to get people to understand what a p value is information about–not the probability that the theory is true conditional on the evidence but the probability of the evidence for the theory being as strong as it is conditional on the theory being false in the particular way represented by the null hypothesis. If everyone understands that, p values are useful–just not as useful as they seem to be to people who don’t understand it.

        • Aapje says:

          One of the problems I think exists is that scientists are too much on their own on this. It’s not realistic that a bunch of social scientists will be able to build up enough expertise to do a custom analysis of a study methodology, for example.

          It should probably become more normal to get consultants to come in before the study starts, to help with the study methodology. In general, I think that scientists should specialize more. There is no reason why creating the study methodology, executing the study and doing the data analysis should have to be by the same people/group.

      • Hence says:

        Aapje, see my comment on significance testing and replication, plus the matter of something better, on the new FAQ.

        • Aapje says:

          Note that when I said to address the “lack of replication, the bias towards significant results, etc,” I didn’t necessarily mean that p-values should not be criticized, but rather that you are not going to get rid of them without addressing those issues first.

          My proclamation is first and foremost a practical one: I think that you simply won’t get the buy-in for change by criticizing p-values like you are doing.

          Humans like tools. They like better tools more than worse tools, but they like a bad tool more than no tool.

          • Hence says:

            Thank you. Agree about humans prefering bad tools to none. Agree that this creates resistance. My view should be clearer in the updated FAQ.

    • cactus head says:

      I agree with what this broadly is about. I am currently working on a Masters thesis in statistics where I found myself stuck in the rut of the significance testing framework for reporting results even though my methods of statistical analysis (generalised linear mixed models, Bayesian models fitted with Markov Chain Monte Carlo) make that difficult and pointless. I hope to keep to a higher standard in the future.

    • sty_silver says:

      I feel like you don’t emphasize the main point nearly enough: p values are mathematical nonsense. Probability has no qualitative levels. You should hammer this point home before you talk about practical implications.

      I mean, maybe you intentionally did it differently, but that’s my reaction.

      • Hence says:

        It was somewhat intentional. Implications make people care, which comes first here. Technical details come later on the page and through suggested links, for those who want.

        Note that “probability has no qualitative levels” is about significance testing, not p value itself, which is a continuous variable. That said, p values are problematic for additional reasons.

    • alef says:

      I have to agree with the tone of many of these comments – you are being far too dismissive of the sociological reasons why p-values (and the like) are so entrenched, and of the practical barriers for any single person to start using “proper methodology”. (And I have to say, as someone who couldn’t agree more with the overall point you are making, this phrase itself sets my teeth on edge, doubly so set aside the vague pointers you give to what this might be (yes, including the second FAQ.)) As such, you get to feel right about something important but are taking a very
      ineffective path to effect change.

      Trying to be constructive… here’s something I haven’t personally seen (not to say it doesn’t exist), but I think would be valuable to you (and to many people, if published). Find and dissect a few exemplar papers: papers in fields dominated by p-values etc, and would could have been presented in traditional (p-value, significance, etc) style, but instead which do it “right” by your lights. Show us (and researchers, and reviewers, and editors, and stats profs) some concrete examples of what good (and ideally peer-review-passing and career-enhancing) research looks like! (If you can’t, or if this is painfully hard, there might be a lesson in that too.)

    • CatCube says:

      I went to your first page, and didn’t come away with an understanding of why significance testing is bad. You just assert that it is, and spam a bunch of links.

      Now, I will say that I just skimmed it. However, I think that’s fair, because that’s all you do with the argument. The best way I can explain it is to take one of your points as an example:

      3. Abandon. Because significance thresholds are a bad idea, period. It creates problems that tweaking cutoff levels won’t fix. To understand the damage it does, why it should end, and what to focus on instead, please read now the clear McShane & Gelman 2017, Trafimow et al 2017, and Gorard 2016. For an amusing account, read the alpha wars.

      Ok, no. Don’t just hand me a bunch of further reading. You’re saying that significance testing “creates problems that tweaking cutoff levels won’t fix” Tell me what the problems are. Similarly, summarize what “the damage it does” is, the arguments for “why it should end”, and give at least bullet points for “what to focus on instead.” If you make good, interesting arguments there, then I’ll probably read more about it.

    • rahien.din says:

      One source of trepidation comes from McShane 2017, summarizing a common thread in your papers :

      In science publishing and many areas of research, the status quo is a lexicographic decision rule in which any result is first required to have a p-value that surpasses the 0.05 threshold and only then is consideration—often scant—given to such factors as prior and related evidence, plausibility of mechanism, study design and data quality, real world costs and benefits, novelty of finding, and other factors that vary by research domain.

      This sequence of assessments is presented as a problem. To me that indicates a misunderstanding of how research is conducted. Once data have been gathered, your first task is to determine if they can even be properly interpreted. Only then can you invoke the novelty, utility, and continuity of the results.

      Tests of significance prevent us from being persuaded by uninterpretable data, even when we are tempted by “What a novel finding!” or “Imagine the use we could put this to!” or “This violates my status quo bias!” We require some test or procedure to fence us off from our biases. That’s the exact aim of science, and to the extent that you consider utility and novelty before checking significance, you are not a scientist.

      So the only plausible alternative is a decision rule in which any result is first required to have a [Bayes factor or confidence interval that meets certain criteria] and only then is consideration given to such factors as related evidence, plausibility, study design, data quality, costs and benefits, novelty, and other factors that vary by research domain.

      You might contend that p-values are just a bad procedure, but I am really hesitant to read these papers when every other one of them contains a sentiment of the sort found in McShane’s abstract.

      Another source of trepidation comes from a paper you did not link, but by the same Stern as in your Gelman and Stern 2012 From Stern 2017 :

      [p-value correspondence with confidence intervals]
      A series of significance tests computing p-values for various hypothesized values of the parameter in question can provide the same information as a confidence interval and a single confidence interval tells us which values of the parameter would be found plausible in a series of significance tests.

      [Confidence intervals do not permit probabilistic claims]
      The single confidence interval being computed does not provide the kind of probabilistic guarantee that the editors of BASP believe is required for an appropriate inference. The frequentist argument that supports the confidence interval procedure guarantees that the specified proportion of the 1–a confidence intervals that we create will contain the true parameter value, but it is not possible to make a probabilistic claim for the one interval at hand.

      [Confidence intervals approximate Bayesian posterior intervals]
      For one thing, in the simple setting considered here the t confidence interval corresponds to a Bayesian posterior interval for a diffuse (sometimes called non-informative) choice of the prior distribution of the model parameters which means that the desired probabilistic interpretation is realistic in that case. In large samples the t-interval is approximately the same as a Bayesian posterior interval regardless of the prior distribution which again justifies the probabilistic interpretation.

      Methodologically, this seems extremely muddy.

      It must be pointed out that a big hurdle for using Bayes factors is selecting appropriate priors, and, that this is the exact same hurdle faced by science in general. It’s simply been moved to a different point in the process. If a confidence interval can approximate the Bayesian posterior without any confounding from inappropriate priors, this is an advantage for confidence intervals. And, if a series of p-values does the same thing as a confidence interval… you get it. Did I mention that one proposal for refining Bayes factors is to calibrate them via frequentist techniques?

      Lastly, and maybe most importantly : if we were to switch to Bayes factors, what is the likelihood that researchers will understand or utilize them with any more rigor than p-values? How fungible are humanity’s capacities for bias and metrics hacking?

      Stern goes on to state very clearly :

      I continue to believe that [banning the p-value and all other traces of significance testing] is a misguided and extreme reaction to the fact that people occasionally misinterpret p-values as being more than just a measure of evidence about Ho.

      I wholeheartedly agree.

      I am grateful for your bringing attention to this controversy, and for reminding me of how I should be approaching p-values, and the limitations thereof.

      • Hence says:

        Hi rahien.din. Thank you for commenting. Now a FAQ, if you want to see.

        > Tests of significance prevent us from being persuaded by uninterpretable data, even when we are tempted by “What a novel finding!” or “Imagine the use we could put this to!” or “This violates my status quo bias!” We require some test or procedure to fence us off from our biases. That’s the exact aim of science, and to the extent that you consider utility and novelty before checking significance, you are not a scientist.

        If you are talking about the data before any analysis: it should be randomized from the whole population, otherwise p values cannot be used by definition. See 3.1 of Gorard 2016 linked on the page.

        If you are talking about being objective, or about evaluating if the finding can be trusted, see False objections 1 and 6 in Objections on the site’s first page.

        > So the only plausible alternative is a decision rule in which any result is first required to have a [Bayes factor or confidence interval that meets certain criteria] and only then is consideration given to such factors as related evidence, plausibility, study design, data quality, costs and benefits, novelty, and other factors that vary by research domain.

        Evaluation of data quality and study design after a lexicographic decision whose function is to decide if the data is to be trusted…?

        > Lastly, and maybe most importantly: if we were to switch to Bayes factors, what is the likelihood that researchers will understand or utilize them with any more rigor than p-values? How fungible are humanity’s capacities for bias and metrics hacking?

        Right, agree. The search for some universal tool.
        See Gigerenzer & Marewski 2015 linked there.

        > I am grateful for your bringing attention to this controversy, and for reminding me of how I should be approaching p-values, and the limitations thereof.

        Happy to know it was useful. Thank you for letting me know.

        • rahien.din says:

          What’s most interesting to me is your response to Stern 2017’s claims that :
          – A series of p-values contains the same information as a confidence interval
          – A confidence interval approximates a Bayesian posterior interval, largely regardless of the prior distribution
          – A confidence interval can not be used to make probabilistic claims

          I don’t know enough about the math thereof to evaluate these claims. Can you help me?

          If you are talking about being objective, or about evaluating if the finding can be trusted, see False objections 1 and 6 in Objections on the site’s first page.

          Those are straw objections.

          Evaluation of data quality and study design after a lexicographic decision whose function is to decide if the data is to be trusted…?

          I am simply arguing within the presented framework. If you claim that this is a bad framework, then I agree. That framework will not get fixed by switching from p-values to something else, because it has nothing to do with the choice of metric. Moreover, you didn’t ask me about that. If you did, my answer would be “That’s not how to read a paper in the first place.”

          To me this is relevant because it suggests that the authors either don’t know how to read a paper in the first place, or, they have a serious ax to grind.

    • Hence says:

      Thank you all for the comments.

      I updated the FAQ with additional questions.

    • Douglas Knight says:

      Is this performance art? Is the purpose of the site to be bad in exactly the same way that p-values are bad?

  2. cheburashka says:

    If you were new to a city without a significant friend circle and had to quickly get a date to a formal engagement such as a ball, what would you do?

    • anonymousskimmer says:

      Tell the people organizing the ball your problem? Perhaps there are single people of the opposite sex who are invited and also need a date? Perhaps they know of people within their friend circles?

      Contact a local dance hall/studio and see if they have any regulars or advanced students willing to step up? Perhaps the instructors themselves would be willing to step up (likely for their typical hourly rate)?

      Failing that, escort service?

    • cassander says:

      Assuming you are male, find someone appropriate at work for whom you have no romantic feelings, tell them you have an extra ticket if they’re interested. Ideally, you tell a couple of them at once, in a group, to make it clear you aren’t asking any of them out. If someone is enthusiastic, make arrangements. If they are non-committal, don’t push and don’t ask again. Don’t ask anyone who works for you.

      • Tatterdemalion says:

        I think “ask a group simultaneously” is a really valuable tip. If you ask an individual who doesn’t know you really well, the risk of misinterpretation strikes me as non-trivial, whereas that’s much less likely if you ask a group all at once.

        I regularly go to ceilidhs and contra dances, where the conventions about asking strangers to dance vary slightly, and one of the tricks I sometimes use to avoid being creepy or making people uncomfortable is to ask “would any of you like to dance” to a group, rather than singling someone out and putting them on the spot.

    • The Nybbler says:

      Next city!

    • toastengineer says:

      Call up an escort service and make clear that I’m actually asking for what they publicly claim to be offering (people do actually do that.)

      • Protagoras says:

        Yes, and in some cases it’s possible that you’ll get a reduced no sex rate. But as I understand it, that isn’t a standard practice or anything, just something you might find with some if you get lucky. And it’ll still be expensive.

    • Andrew Hunter says:

      Why, where, and how is this a situation that happens?

      • CatCube says:

        I could imagine it in the Army, if you got assigned to a new unit in a new place immediately prior to the unit ball.

        Edit: Although, in a military context, going stag wouldn’t be a problem, especially in those circumstances.

  3. Vincent Soderberg says:

    Question to everyone that had/has depression, had trouble feeling happy/emotions/motivation, and now has gotten better: what does it feel like when having those things again?

    • anonymousskimmer says:

      I think I got used to it, so I can’t remember if the pleasant emotions felt any sparkier when they first came back.

      • anonymousskimmer says:

        I remember feeling as if my life was on track again, and that things might be good in the long run.

    • Donald Smith says:

      Just read Nietzsche on his depression and aftermath on Brain Pickings by Maria, and some of Werner Heisenberg’s love letters. Great Sunday morning email.

    • Baeraad says:

      I mostly just tend to feel relieved. Like when you’ve been stuck in a really cramped room all day, and you finally get to go outside and take a nice big breath of fresh air.

    • Nabil ad Dajjal says:

      I remember very vividly when my depression started letting up because I was looking out my kitchen window one morning and out of the blue I thought “life is great.” It felt very alien, like it was someone else’s thought, but then I realized that I just wasn’t used to being content. I was used to intrusive thoughts but they weren’t generally happy until that point.

      That was the biggest change I noticed internally. My intrusive thoughts became friendlier and more helpful.

  4. Vincent Soderberg says:

    Has anyone here read Elephant in the brain, and wants to talk about it?

    • Synaspora says:

      Looking forward to reading it–anything new in it?

      • Vincent Soderberg says:

        yes and no. I would say it’s new in how it connects a lot of dots that you probably didn’t notice before. i.e., how we have hidden selfish motives that we don’t wanna acknowledge. Every individual bit is not super new, but by bringing it all together, it paradoxically becomes bigger then it’s parts. I would say that even if you have read tons of psychology, it’s probably worth reading it.

    • fortaleza84 says:

      I read it. As I mentioned in the last thread, I have been paying attention to normal conversations in light of that book and it’s astonishing just how much of human conversation consists of showing off; jockeying for status; and/or trying to get laid.

      • Vincent Soderberg says:

        can you link me to that thread? I just don’t feel like chatting today, but I want to read peoples thoughts on it.

      • Vincent Soderberg says:

        Yeah. I’m high functioning autistic, and I feel like knowing this stuff retroactively makes me get social interactions a lot more.

    • James says:

      Started it and enjoying it quite a lot so far.

    • Urstoff says:

      I’m about a third of the way through it. Trying to whip up the motivation to continue reading, but being familiar with Hanson’s work for a long time, I’m pretty sure I know almost everything that’s going to be said (or at least the gist of it).

  5. johan_larson says:

    More poetry puzzlers. This time it’s all Kipling all the time.

    N T I T L O T J —
    A O A A T A T S;
    A T W T S K I M P,
    B T W T S B I M D.

    A R, A R,
    W S T R A R?
    T L R T G A T,
    T B S F, B N B,
    T K T S T A
    W T O J A R.

    I, D W S O P, W L
    W T T H N T I A,
    S B A T G U,
    A L B W T L —
    L G O H, B W U Y,
    L W F — L W F.

    O, E I E, A W I W, A N T T S M,
    T E A S S P A G G J S;
    B T I N E N W, B, N B, N B,
    W T S M S F T F, T T C F T E O T E.

    “O, W A Y G T, A Y B S,
    W E O C, U A D T S S?”
    “W A G T F Y Y B A Y B,
    Y B, P, A M, E, A, A C.”

    • Donald Smith says:

      I know Kipling. Where are the rules for these puzzles?

      • johan_larson says:

        What sort of rules do you mean?

        Edit: These are quotes from poems by Kipling, with only the first letter of each word shown. Your job is to find the titles of the poems.

        You are welcome to post the answers, but encode them using ROT13 to avoid spoilers.

    • Tatterdemalion says:

      I know three of these.
      1: ABJ guvf vf gur Ynj bs gur Whatyr — nf byq naq nf gehr nf gur fxl;
      Naq gur Jbys gung funyy xrrc vg znl cebfcre, ohg gur Jbys gung funyy oernx vg zhfg qvr.

      4: 0u, Rnfg vf Rnfg, naq Jrfg vf Jrfg, naq arire gur gjnva funyy zrrg,
      Gvyy Rnegu naq Fxl fgnaq cerfragyl ng Tbq’f terng Whqtzrag Frng;
      Ohg gurer vf arvgure Rnfg abe Jrfg, Obeqre, abe Oerrq, abe Ovegu,
      Jura gjb fgebat zra fgnaq snpr gb snpr, gubhtu gurl pbzr sebz gur raqf bs gur rnegu!

      5: “BU, jurer ner lbh tbvat gb, nyy lbh Ovt Fgrnzref,
      Jvgu Ratynaq’f bja pbny, hc naq qbja gur fnyg frnf? ”
      “Jr ner tbvat gb srgpu lbh lbhe oernq naq lbhe ohggre,
      Lbhe orrs, cbex, naq zhggba, rttf, nccyrf, naq purrfr.”

      I’ve also worked out number 2, but only by cheating and using an index of first lines, so I’ll leave it for someone else to claim. Number 3 puzzles me – as far as I can see does not list any poems starting with the words beginning with I D W. Are you using one not on there, or have I just missed it?

      • johan_larson says:

        The quote from number 3 is not from the beginning of the poem.

        Let me suggest you encode your answers, to let others take a crack at the problems, too.

        • Tatterdemalion says:

          I got number 3 in bed overnight:

          Vs, qehax jvgu fvtug bs cbjre, jr ybbfr
          Jvyq gbathrf gung unir abg Gurr va njr,
          Fhpu obnfgvatf nf gur Tragvyrf hfr,
          Be yrffre oerrqf jvgubhg gur Ynj—
          Ybeq Tbq bs Ubfgf, or jvgu hf lrg,
          Yrfg jr sbetrg—yrfg jr sbetrg!

      • Trofim_Lysenko says:

        I knew the same three. The other two are tantalizing but haven’t clicked yet and I’m trying to avoid looking at poems to refresh my memory.

    • johan_larson says:

      Some hints:

      N T I T L O T J —
      A old A A T A T S;
      A T W T S K I M P,
      B T W T S B I M D.

      A R, A R,
      W S T R A R?
      T L R T G A T,
      T bend S F, B N B,
      T K T S T A
      W T O J A R.

      I, D W S O P, W L
      wild T T H N T I A,
      S B A T G U,
      A L B W T L —
      L G O H, B W U Y,
      L W F — L W F.

      O, E I E, A W I W, A N T T S meet,
      T E A S S P A G G J S;
      B T I N E N W, B, N B, N B,
      W T S M S F T F, T T C F T E O T E.

      “O, W A Y G T, A Y B S,
      W E O coal, U A D T S S?”
      “W A G T F Y Y B A Y B,
      Y B, P, A M, E, A, A C.”

      • johan_larson says:

        More hints:

        N T I T L O T J —
        A old A A T A T sky;
        A T W T S K I M P,
        B T W T S B I M D.

        A R, A R,
        W S T R A R?
        T L R T G A T,
        T bend S F, B N break,
        T K T S T A
        W T O J A R.

        I, D W S O P, W L
        wild T T H N T I awe,
        S B A T G U,
        A L B W T L —
        L G O H, B W U Y,
        L W F — L W F.

        O, E I E, A W I W, A N T T S meet,
        T E A S S P A G G J seat;
        B T I N E N W, B, N B, N B,
        W T S M S F T F, T T C F T E O T E.

        “O, W A Y G T, A Y B S,
        W E O coal, U A D T S seas?”
        “W A G T F Y Y B A Y B,
        Y B, P, A M, E, A, A C.”

        And the answers:
        1. Gur Ynj bs gur Whatyr. Abj guvf vf gur ynj bs gur whatyr …
        2. Gur Errqf bs Ehaalzrqr. Ng Ehaalzrqr, ng Ehaalzrqr, …
        3. Erprffvbany. Vs, qehax jvgu fvtug bs cbjre, jr ybbfr …
        4. Gur Onyynq bs Rnfg naq Jrfg. Bu, rnfg vf rnfg, naq jrfg vf jrfg …
        5. Ovt Fgrnzref. Bu, jurer ner lbh tbvat gb, …

    • SamChevre says:

      Finally gave up and looked at an index of first lines for 2–then it was easy.

      I really enjoy these.

      This set reminded me of what a versatile poet Kipling was.

      • A while back I was reading through collections of various other poets, looking for things I could use for my book/web page of literary works with interesting economic insights. What struck me was how much larger a fraction of Kipling’s poems were ones I liked than of almost anyone else–the one possible exception that occurs to me is Millay. But her range is much smaller.

  6. Andrew Hunter says:

    A while back I was asking for book recommendations and someone suggested Torchship. I have since read all three in the trilogy.

    The first was an extremely pleasant surprise: basically hard sf Firefly. Sequence of interesting capers exploring random corners of a reasonably well thought out setting, with clear hints at a larger plot. Characters are all sketched archetypes and the prose is a little pedestrian, but highly enjoyable, and the main plot line grows nicely. The setting, while vague, has a lot of cool aspects, and a great mix of high and low tech (including the best justification for human involvement in spaceflight I’ve seen in a while.) My favorite sequence was the xvqanccvat frdhrapr ol gur tnf tvnag, where we really saw the limits of available warfighting tech in great detail, and a sequence of interesting reveals.

    Second book tries to be the same, but more so. I like seeing more corners of the setting and more characterization, and there’s much more of a cohesive single plot. The implied diversity in book one was a bit more fleshed out here, which was cool, without at all being diversity for its own sake (very much not “nothing happens, but at least it doesn’t happen to minorities.”) The plotline about gur Shfvba snxvat bzrtnf was a genuinely novel societal idea, and one worth investigating further. Definite flaws, though: the pacing is off and a lot of scenes felt more like “roll d20 on random encounter table” than should. The McGuffins are a bit painful, especially the zntvp unpxvat plorejrncba which just felt off. And most importantly, gur obkrq NV whfg gbgnyyl ehvaf gur srryvat bs gur frggvat, naq gur fcrpvnyarff bs gur greensbezref vg pnzr sebz. N pbhcyr enaqbz rkgenf frrz hcfrg nobhg vg orvat gurer naq rirelbar ryfr whfg npprcgf vg–jung gur shpx? Guvf fubhyq or n zhpu ovttre qrny guna vg vf, ohg gur nhgube jnagf gb erirny gur bzrtnf naq gb irevsl n srj snpgf bs gur Orgenlny naq pna’g svther bhg n orggre jnl gb qb vg, fb whfg oevatf va n crg NV gura qebcf vg fhqqrayl. Netu. The ending, while a cute reveal, also felt very out of place.

    Third book is kind of a disaster. A few things were good: seeing our main characters become more and more important in the world felt like a payoff. Some brave choices shown in exploring wildly divergent politics, even if they were mostly dumb (gur serapu eribyhgvba erqhk jnf cnvashy rabhtu jura Qnivq Jrore qvq vg.) I liked the exploration of Harmony pretty well (and Pbashpvna ubarlcbg puvpx was more interesting than I expected when she first showed up.)

    On the other hand: the ending was fucking awful; the main antagonist just goes over like a jobbing wrestler, taking all the menace out of them retroactively. (And also really breaking the realism to anyone who takes nv nyvtazrag seriously.) And these characters really did fit better on a small scale. We kept inventing reasons for their tramp freighter to be there at all.

    Also, some truly unnecessary and what-the-fuck scenes. Frevbhfyl, gjb gbgnyyl bhg bs abjurer naq haarprffnel guerrfbzrf ba fperra? Jung gur shpx? Bxnl, jura Thra pbzrf vagb Zvgpuvr naq Thb’f pnova, V wbxvatyl fnvq gb zlfrys “Bu zna, guvf vf tbaan raq va n guerrfbzr!” orpnhfr V’z onfvpnyyl guvegrra vagreanyyl. Gura jr ernyyl qb trg n cybgyvar bhg bs cbea, naq purrfl shpxvat cbea ng gung. Jr unir n ornhgvshy gevyyvbanver *grrantre* jub fnlf, jub YVGRENYYL fnlf, “V whfg qba’g haqrefgnaq frk…pna lbh gjb cbffvoyl *tvttyr* fubj zr?” Naq gura jr trg n tenghvghbhf qrfpevcgvba fgenvtug bss bs Yvgrebgvpn? Jub cbffvoyl gubhtug guvf jnf n tbbq vqrn?

    Ybbx, V trg gung Zvgpuvr jnf n cebsrffvbany ubarlcbg, naq gur nhgube vf gelvat gb frg hc fbzr qenzn nebhaq vg, ohg frevbhfyl, zl vzzrefvba va jung bgurejvfr jnf n uneq fcnpr bcren nobhg NVf naq gbepufuvcf naq fbzr cbyvgvpf, rira va n frevrf gung unq dhvgr n ybg bs pbhcvat-hc, jrag gb n fperrpuvat unyg.

    Gura gurer’f gur frpbaq guerrfbzr bhg bs abjurer (frevbhfyl, abg n cuenfr V fubhyq unir gb hfr nobhg nal obbx abg fbyq haqre cynva pbiref!) Naq V zrna, lrf, gur vqrn bs Zvgpuvr univat gb grnpu n ubbxre nobhg ubj gb cergraq gb or ure vf shaal. V’yy tvir lbh gung. Vg’f rira n ovg pyrire. Ohg ntnva, gur cbvag ng juvpu lbh, gur nhgube, qrpvqr vg’yy or vagrerfgvat gb unir n punenpgre fnl “Lbh xabj, V guvax V’z tbvat gb rng lbh bhg, fb V pna yrnea ubj gb fbhaq yvxr lbh univat na betnfz?” V’z fgnegvat gb guvax guvf puncgre jnf jevggra jvgu bar unaq.

    Okay, I maybe went on about the guerrfbzrf scenes more than I should have, but seriously, more than anything else in the book this confused the fuck out of me. I cannot overestimate how fucking weird that was.

    Overall: first one is not a book for the ages, but nevertheless an extremely fun little gem which hooked me deeply. The second tries to be more of that and a bit grander, mostly hits but has a little less charm. Third book is a fucking mess. I’d highly recommend the first. The second is a positive reading experience but shouldn’t be your highest priority unless you’re into completionism. Third is skippable unless you *really* want to see these characters more (which you well might.)

    • John Schilling says:

      I wish I had written this review, and will have to settle for agreeing with it on all points. Especially the gratuitous guerrfbzrf, which is only welcome when I’m not trying to take the characters in question seriously.

    • I did not have the same reaction. A good deal of what bothered both of you struck me as plot relevant, and not wildly implausible given the situation.

      • Andrew Hunter says:

        Anything in particular really stand out as well justified that I didn’t accept?

        • The whole bit with the prostitute role-playing the protagonist, and then the way in which the protagonist made use of her. The love/sex problems originally created by the protagonist’s job, then spreading later into a potential love triangle and how it was dealt with. And I liked the author’s idea of the potential problem with a UBI even in a society rich enough to be able to afford one.

          I never figured out why the historical parallelism was there–presumably the academic responsible had some reason for creating it. But it worked tolerably well.

          And the ending feels like a wet firecracker only because the reader has bought into the erroneous picture of the situation held by almost all the characters.

          Torchship is a first novel, so the series isn’t as good as it could be if written by a much more experienced writer, but it was still good enough so that I read and reread it with pleasure. The particular feature that seems to have bothered both of you most didn’t strike me as a problem, except to the extent that it limited whom I could recommend the book to.

          But then, I’m very fond of Casanova’s Memoirs, which might be viewed as having a similar problem.

  7. bean says:

    Today at Naval Gazing: HMS Dreadnought, and why she’s important.
    Also, it’s been one year since I wrote the first post in what became Naval Gazing.

    • cassander says:

      random question, do you have the equivalent of a TO&E for the iowa that shows all the billets on the ship and the ranks they’d hold? Or if not the whole thing, at least the officers?

    • Urstoff says:

      Is there a decent book that talks about the development of battleships from ironclads to WWII?

    • Vermillion says:

      I enjoyed the post but this bit confused me,

      She was laid down on October 2nd, 1905, and launched only 5 months later, on February 10th, 1906. The 14 months exactly from laying down to commissioning set a record that has never been broken for capital ships, and was intended by Fisher to send a message to the world.

      I feel like there should be a sentence in the middle talking about what happened in the 9 months between launch and commissioning.

      Also RE:the Dreadnought hoax, I find it hilarious that one of the pranksters was Virginia Woolf! I first heard about that just last week on the Cracked Podcast here @43 minutes.

      • bean says:

        I feel like there should be a sentence in the middle talking about what happened in the 9 months between launch and commissioning.

        That’s fitting-out and trials. Particularly at the time (modern ships are constructed rather differently), the bare hull was built with minimal systems, and most of the equipment was installed afloat. The turrets are the most obvious examples, but on some ships, engines and the like also were installed after launch. And then you have all of the tests to make sure that the ship works well. There’s not a one-sentence explanation, although it’s due for treatment under “so you want to build a battleship” when I get around to it.

      • johan_larson says:

        BUNGA BUNGA!

    • Chevalier Mal Fet says:


  8. Synaspora says:

    Any comments on Steven Pinker’s new book?

  9. Odovacer says:

    What Winter Olympic sport would you be most likely to do or want to compete in?

    Suggest another sport to add to the Winter Olympics.

    Edit: Here’s a better list of the events:

    • Odovacer says:

      I’d like to compete in either the 20 km biathlon or get a team together for the 4-man bobsled. I really like cross-country skiing and the idea of the biathlon appeals to me. Maybe I could pretend to be a Finn in Winter War. Also, bobsledding looks fun and not as crazy as the skeleton or luge.

      If I were to suggest a sport, I suppose mountain climbing or ice climbing seems wintery.

      • johan_larson says:

        I’m thinking there should a climb/ski event where you first run up the hill with your skis, and then ski down it.

    • The Nybbler says:

      Speed skating, certainly; an easy choice since I used to compete in amateur inline speed events. I would really love a chance to try long track speed, but there aren’t many tracks and I’ve never made it work out.

    • fortaleza84 says:

      Probably curling since I like to drink beer.

    • Brad says:

      Probably Super G.

    • gbdub says:

      Obviously you know I’d say curling. But I do already play that at an amateur level.

      I also do downhill skiing, but it would be awesome to be good enough to do super G or slope style.

      Of the sports I’ve never tried in any form, bobsledding.

    • Paul Zrimsek says:

      I’ve been told that I’m a born luger. I think that’s what they said, anyway.

    • ohwhatisthis? says:

      The one that relies on gravity the most. Luge sounds fun.

      I would try the skiing events, but i’m only the worlds most basic pizza french-frier. When the entirety of your skiing knowledge comes from a south park episode, you’re gonna have a bad time.

      Curling is the sport I want to do more then any other sport, so I can empirically test if it isn’t some strange placebo sport. If it is a placebo sport, sign me up.

  10. johan_larson says:

    Does anyone here work in collections? I’m wondering what happens with people who have student loans they can’t realistically pay off. Suppose for instance someone borrowed $150,000 to go to Fancy Pants College expecting a great job at Big Company. Only they never got the job, and are now faced with the prospect of trying to pay off the debt with a $10/hour job, which probably isn’t feasible.

    What happens next? I seem to remember student loans aren’t typically dischargeable through declaring bankruptcy.

    • Brad says:

      They aren’t dischargable in bankruptcy unless you can meet a legal test which, as interpreted, is very nearly impossible.

      However, if you are talking about public student loans, which are the overwhelming bulk right now, there are income based repayment plans. In the scenario you outline, with the person making say $20,000, with a household size of one would have a monthly payment of $15. With any larger household size the monthly payment would be $0.

      After 20 years worth of qualifying payments (including $0) the remaining principle and capitalized interest would be forgiven. Under current law that forgiven amount, which would have grown quite substantially in the meanwhile, is treated as income for tax purposes. But the IRS also has provisions for people that can’t afford their tax bill which includes an ultimate discharge. Further, IMO it is likely this tax treatment will change when forgiveness starts kicking in.

      Because of these programs, in general there isn’t a good reason to be in default of direct government student loans. But some people are anyway. In those cases the servicer will seek to garnish wages, tax returns, and eventually social security checks.

      I’m not sure what happens with private student loans. I’ve heard that despite not being discharable lenders will negotiate reductions on them, something the DoE will not do at all. But that’s second hand.

      • outis says:

        That sounds really generous. Why do people still act like student loans ruined their life, then? Are they trying to finance their lifestyle with credit card debt and they’re upset that their credit score is bad?

        • johan_larson says:

          Yes, it does sound generous. I had expected something much more severe.

        • MrApophenia says:

          I think a big part is all of the above only applies to direct federal loans. There are strict limits on how much the government will directly loan to a given student, based on financial need. If that’s not enough to pay for school, you’re getting loans from a bank instead, and they aren’t so nice.

          (I mean actually my loan holder was still very reasonable so I don’t have a lot of complaints on this, but it was a lot of money to need to pay back.)

          • Brad says:

            Private loans are around 10% of the total outstanding student loan dollars. That includes loans made for undergraduate, graduate, and professional education.

          • A Definite Beta Guy says:

            Private loans might be a significant factor among people who are complaining, though. I’m not an expert, but it looks like the maximum federal loan at the graduate level is something like $20,000 per year? My wife had 4 years of graduate education that probably cost around $45k/pop, so a good $100k right there.

            Private loans make up the bulk of our liability, I am guessing because federal aid was quickly exhausted due to out-of-state tuition rates.

            We’re on an income-based repayment plan that does not appear to have been updated for a while (so we should be paying more money than we are), but the private loan payments are almost as high. Basically doubling the debt burden.

          • Brad says:

            Those loan limits are for Stafford loans. Grad PLUS loans are capped at the school’s published total cost of attendance (minus other loans/aid).

            Private student loans to grad students, especially professional grad students, are more likely to have a lower initial interest rate and fees, but they are generally going to be adjustable rate and don’t qualify for income driven repayment plans.

          • A Definite Beta Guy says:

            Ouch, those Grad PLUS loans look expensive. There’s a 4% loan fee and a 7% interest rate.

            The adjustable rate private loans we have are Prime+.5%.

            Edit: though, yeah, no income repayment plan available. Wonder how those numbers would actually work out, but I wasn’t around when these loans were incurred, so it’s all guess work on my end.

          • Brad says:

            And don’t forget Public Service Loan Forgiveness, that’s another wrinkle in the decision making process.

          • A Definite Beta Guy says:

            Well I just spent 40 minutes on absolutely pointless debt repayment spreadsheets, so thanks for that, Brad! (I jest, I jest…hopefully that was obvious…)

        • Brad says:

          In general the people complaining aren’t making $20,000 a year with no hope for significantly more in the future. After 150% of the poverty level it’s 10% of income. Think about how much the people making enough money to pay the top marginal rate endlessly bitch and moan about their taxes. Effectively people at much lower, though still decent to very decent, income levels are paying those same rates. And keep in mind the interest capitalizes, so if the number is large and you make the minimum payment you may well be locking yourself into 20 years of payments (25 years for grad loans). Yes, people exaggerate about “ruined my life” but exaggeration is in style right now.

        • anonymousskimmer says:

          I’m not one of those making this complaint, though it sort of applies to me:

          Here are the relevant lines in my budget:
          Earnings 5408.00
          Bef Tax Ded 486.72
          Tax Gross 4921.28
          Total Tax 895.72
          After Tax Ded 103.64
          Net Pay 3921.92
          Student Loan -471.31

          CC #1 -200.00

          I’m also still paying off a credit card (~9% APR, $200/month, $8200 current balance) that was racked up in large part while going to school (I had to take time off work during school). The other credit card is being paid off in full every month thanks to a small inheritance that allowed paying it’s $10k+ balance off.

          For the student loan I’m on the “Graduated” repayment plan, which means that the payment goes up to $612.71/month next year. (I couldn’t afford the regular repayment plan after graduation, as the credit cards had to be paid and I only received a $1.08/hour pay increase at my prior job after graduating.)

          While the current student loan payment is 12% of net, it is much less than 10% of gross, and it’s the gross income which counts. Next year my student loan payment will be above 10% of gross income, but it’ll also be less than 3 years left to pay off, so it’s not worth switching to an income-based plan which would add years and interest to the payoff, if I’d legally be allowed to do that after choosing the graduated plan in the first place.

          This is why I do not have a smartphone, and am also not paying the $121/month for a non-catastrophic medical insurance plan, and do not have a bicycle, and etcetera.

          Taxes + student loan payment = 25.28% of gross
          Taxes + student loan payment + mandatory 9% pension payment = 34.28% of gross

          • Brad says:

            The REPAYE formula is based on AGI* not gross income. It’s 10% after 150% of the FPL. So your REPAYE amount would be:
            (5408 * 12 – 18210)/(12*10) = $389.05/month.

            You are always allowed to make higher or additional payments, there are no prepayment penalties. I see no downside to switching from graduated to REPAYE. Though if you are happy enough with the current payments, perhaps no upside either.

            *actually MAGI

            Edit: Re-reading, there would be an advantage. Your CC is at a higher rate than you student loans, if you put the difference between the REPAYE amount and the graduated amount towards the CC debt you’d come out ahead.

          • anonymousskimmer says:

            Thanks Brad, I’ll look into this*. And that’s another reason why people complain about student loans – they lock in in one situation and don’t realize down the road that things have changed such that they can switch to another.

            Those with student loans are less likely to be able to afford an accountant, after all. 🙂

            You are always allowed to make higher or additional payments, there are no prepayment penalties.

            The “penalty” to making additional payments is that your next autopay amount is decreased by the amount you pre-paid (at least with my servicer – Nelnet). So if you prepay once, and want to keep the bonus of having prepaid, you have to keep manually paying every month.

            * – the estimate on Nelnet’s website is -$54/month on income-base ($417.28), -$190/month on REPAYE ($278.18), and -$210/month on income-contingent ($258.58) from last year’s AGI of $57,742 and married-filing-jointly.

            Taking this and paying my credit card down faster will reduce the credit card payoff time from 50 months to 22 months. 😀 😀

          • Brad says:

            Didn’t realize you were married, that’s even better.

            Glad I was able to provide a useful pointer. I agree that this stuff is needlessly complicated.

            One thing to try re: the prepayment is see if you can push the ACH payment from your bank side instead of pulling it from the servicer side. Though you may want to leave the pull on, even if it is pulling zero, because there may be an interest rate discount for having automatic payments set up.

          • anonymousskimmer says:

            Yep, 0.25%, which is significant when talking about student loan APRs.

          • A Definite Beta Guy says:

            It might not be relevant, but this is probably a good place to note that the actual interest rate is not the APR. APRs are lower than the actual interest rate you are paying.

            APRs are just an annual rate. IE, I tell you I will charge you 10% interest…you might assume that if I lend you $100 at the beginning of the year, you will owe me $110 at the end of the year.

            That’s not how it works.

            Interest compounds throughout the year, normally at the end of every month, but possibly more frequently. So, the APR is broken into 12 discrete periods (in this case .8333%).

            So, really, at the end of January, you owe me $100.83. Then you owe me THAT plus .833333% interest at the end of February.

            Your 10% APR actually costs you 10.47% in interest every year.

            For my 6.2% loans, they actually cost 6.38%.

        • The Nybbler says:

          As long as the debtors are low-income, the student loans are not a large issue. But as interest keeps accruing and being capitalized (compounded), it doesn’t take very long for the hole to be too big for them to dig out of in less than the 20-year limit. If their income goes up, the payments go up too, and they can’t get ahead.

  11. onyomi says:

    Haven’t looked into the details, but this story about study showing children work harder when dressed as Batman seems relevant to local interests.

    My summary: asked to work at a boring task while resisting the temptation to play with an available iPad, children spent more time on task if they told themselves “Alice is working hard on this important task” than if they told themselves “I’m working hard on this important task.” And supposedly they did even better if they got to dress up as a favorite superhero and tell themselves “Wonderwoman is working hard on this important task.”

    Implication seeming to be: it’s easier to concentrate/exert willpower when thinking of oneself from a third person, objective perspective, and especially when thinking of an idealized version of oneself?

    • Well... says:

      I wonder if there’s some provable truth to the “dress for success” adage. Do bankers who wear suits perform better than bankers in slacks and polo shirts? Personally I hope it’s not true. I would wear t-shirts and cargo shorts every day if I could.

      When I was a kid I played sports. I remember feeling energized just by putting on my uniform. Also, I performed better when playing organized sports outdoors at night, because it was kind of a big deal to have the lights illuminating the field and it made me feel like a professional. No clue if any of that would still work now.

      • Matt M says:

        I’ve had a somewhat inverted view of this theory lately. Maybe I’m crazy, but I feel like in certain situations, if you dress poorly, people will respect you more. They’ll look at you and say “Wow, he must be a REALLY good X if he’s that casual and unprofessional and they keep him around anyway!”

        Call it the M*A*S*H theory?

        • onyomi says:

          I think it’s highly dependent on individual and cultural context.

          Famously in New York you look like you’re a serious person if wearing a nice suit, whereas in California you look like you’re trying too hard/are a butler/waitstaff-level person.

          Within any given context it’s probably a signalling game like “Right is the New Left“: dressing nicely distinguishes you from people who are just poor, lazy, bums, but if you’re confident enough you can’t be mistaken for a poor, lazy, bum, dressing less well may send a signal that you’re so secure in your status as a not-poor, lazy bum that you don’t need to look the part, and so on.

        • engleberg says:

          Montgomery let exceptionally competent Brits slide on crappy military appearance, and it spread, as you describe.

        • There’s an overcoming bias post on this, although I’m not sure I’ve dug the right one up.

          • Garrett says:

            I’d also add that there’s a difference between dressing up, and dressing well. The high-powered folks who are dressing down are almost certainly not wearing untailored clothes they bought from a second-hand store (unless that’s specifically their shtick). I tend to see them wearing well-tailored clothes which might as well have cost as much as a fancy suit, but just happen to be less formal.

            Now I want to see someone wear real formal wear including cape, etc. (think Count of Monte Cristo movie) to a board meeting. Bonus points if it’s a new-hire junior front-line customer service manager.

          • A Definite Beta Guy says:

            I’d also add that there’s a difference between dressing up, and dressing well.

            Seriously, “chic casual” or whatever it is, is a great way for the rich to flaunt their wealth and the fashionable to flaunt their style while remaining quite casual.

            Most people know this intuitively. I have some really baggy sweaters and bright blue levi jeans I wear on the weekends. These are not fashionable. I have a quarter-zip and some dark blue jeans that are more fashionable, even though both outfits are the same level of “casual.”

          • anonymousskimmer says:


            In the fantasies where I’m attending a televised red carpet event I dress full sun-king fop.

          • Michael Handy says:


            Gasp! Capes and White Tie are EVENING WEAR, and in any case capes and opera pumps are only acceptable for the Opera and functions where there will be dancing.

            Ultra-Formal day wear stands out a little less, though they might ask who you’re getting married to

            I too would like to see someone in a Swallow-tail coat, waistcoat, and stripy trousers in a business meeting though. You do occasionally see it worn on the continent, especially in Italy

          • Another Throw says:

            I too would like to see someone in a Swallow-tail coat, waistcoat, and stripy trousers in a business meeting though. You do occasionally see it worn on the continent, especially in Italy

            Do bankers still wear morning dress in The City? Does the Attorney Solicitor General even still wear it in the Supreme Court? Like, honestly, the only place I think you can reliably find it is at the Royal Ascot.

          • Brad says:

            Male attorneys in the Solicitor General’s Office that appear before the Supreme Court wear morning dress. Female attorneys in the Solicitor General’s Office can optionally wear a version of men’s morning dress or can choose something else. When Kagan was SG she wore a regular pants suit.

      • onyomi says:

        I find making detailed to-do lists makes me more productive, not only because it makes me take account of time usage, but, I think, because it makes me think about my day from a “third-person” or perhaps “long run” perspective: rather than deciding moment to moment how I’d like to spent the next few minutes, I try to decide how I’d be happiest, at the end of the day, if I spent my time, and plan accordingly (sort of like how it’s easier to make your RPG/”Sims” character have a productive day than yourself because you can simply decide he will spend all day practicing the violin and it is as good as done).

        On the other hand, I think dressing nice probably reduces my productivity because, subjectively, dressing nice feels “ego depleting”/willpower draining, meaning I have less willpower left for work than if I’d stayed in sweatpants. Going to the office when I could have worked from home is a mixed bag: there are fewer relaxing distractions, but being there is also mildly willpower depleting relative to being at home.

        That said, it does seem like getting into “uniform” or otherwise approximating a high-pressure “performance” context may be useful in getting oneself to practice seriously/concentrate (to give “game day effort” in practice, so to speak).

      • A Definite Beta Guy says:

        I figured this would be easily googleable. Individual performance assessment or competence evaluation based on dress.

        Couldn’t find anything. My Google-Fu must suck.

        IME (since I don’t have any hard evidence), “dress for the job you want” has basically zero value outside the interview stage. I won’t say performance evaluation and promotion is based on “merit,” but there’s a merit element, and subjectivity and bias tends to get introduced based on personal relationships, confidence in speaking, speaking in a way people can understand, getting lucky breaks, your manager having lots of political clout (a highly underrated aspect).

        There are some guys who definitely try to dress well to get ahead. If anything, there’s an inverse correlation between their attempts to dress well and their promotion efforts. They have poor work ethics, come in late, are usually only of middling competence, and speak in a really condescending manner. They also tend to think quite highly of themselves, and then they DON’T get promoted, they tend to get pissy, which is a GREAT way for management to automatically discount you for any future promotions.

        Plus these kinds of folk tend to look down on YOU for not dressing well. Like, dude, we are allowed to wear jeans to work, and we have a laid-back office where everyone wears jeans. Hell, the f’in VP is wearing a goddam hoodie. YOU are the one who looks like an idiot in your suit and tie.

        And when I say “guys,” I mean “guys.” There are probably women who believe in the “dress for the job you want” mantra, but I just haven’t met any of them yet.

        • John Schilling says:

          And when I say “guys,” I mean “guys.” There are probably women who believe in the “dress for the job you want” mantra, but I just haven’t met any of them yet.

          I used to have one of them working for me, before she lateraled into a job even better than the one I was trying to swing for her in my section. To be fair, she did more than just dress for the job she wanted, but I’m fairly certain that her dress style was calculated for professional effect.

        • SamChevre says:

          There are some guys who definitely try to dress well to get ahead.

          I’m one of them, and so far it has worked well for me.

          Actuaries are stereotypical geeks, good at explaining things in detail, and somewhat likely to answer the question you asked (in great detail) without really helping much*. And with that reputation goes a tendency to dress somewhat poorly. (One former colleague–a very senior, well-respected actuary–was notorious for buying used monogrammed dress shirts and picking the monogramming out. Another gave the distinct appearance of having bought all his clothes in one shopping trip in about 1975.)

          I started doing two things a decade ago. One, wear clothes like the people who get asked to meetings, not the ones who spend all day every day at their computers. And two, of those people, dress like the ones who meet with people who are my boss’ boss–dress like someone one level higher in the organization than I am.

          Over time, I’ve gotten a reputation as a go-to person for jobs that need someone who is good at the explaining part of the job. I think looking like I’d be good at that/willing to do that helped.

          *There’s an actuarial joke about a man in a balloon shouting to someone on the ground “where am I?”, and getting the answer “you are 26 feet above the ground, 1127 feet above sea level, 30 feet north and 11 feet west of me.”

          • A Definite Beta Guy says:

            Hope it continues working out for you. Honestly, the only people I’ve ever seen in suit and tie at my offices were the CEO (once), salespeople trying to make a pitch.

            Even the CEO comes in wearing blue jeans and a pull-over. Right now the VPs are wearing a hoodie, a patterned-dress that looks like came from H&M (not a high-end retailer), and a dude wearing a standard gray dress shirt and khakis.

            Directors are pretty much the same.

            I remember one time a CFO sat right next to me and I had no idea. I thought he was an intern. Zippered hoodie undone, ill-fitting blue jeans, chowing down a Subway sandwich and getting crap everywhere.

            I think dressing “smartly” should at least be a leg-up, at least if you are somewhat competent. I think a big factor in your success might also be those shitty explanations you mentioned. That’s a HUGE problem. When we say people can’t communicate, that’s normally what we mean. I think media/pundits/colleges spin this as students being unable to write boring-ass 5 paragraph essays or whatever.

        • Michael Handy says:

          A trick is to dress ONE level above the rest, Ie. If everyone is jeans, t shirt and hoodie, Jeans and a short sleeve button up and a fancy looking coat is the maximum you can wear to “impress”. A suit just makes you look like a wanker.

          • Le Maistre Chat says:

            This is where I again tell the story of the rationalist guy I met at the University of Oregon who wanted to get a trenchcoat and brimmed hat as rational rain wear, so I warned him that he’d have to wear slacks and collared shirts to not look like a stupid hipster, and he ended up going around overdressed all the time.

          • Nornagest says:

            What was his reasoning? I like vintage looks myself (though I own neither trenchcoat nor brimmed hat), but I’d have a hard time selling them as “rational”. Trenchcoats are traditionally cotton garbardine, which is heavier, less durable, and less water-resistant than synthetics, stains more easily, and wrinkles if you don’t store it right: less practical on all counts than a North Face parka. And they’re not exactly fashionable right now. Unless you really want to look like a nerdy Humphrey Bogart.

          • Le Maistre Chat says:

            @Nornagest: That jackets don’t protect your legs from rain/snow, and the hood on one still let rain get on his eyeglasses.
            Also I think the trenchcoat was a water-repellent synthetic, not cotton gabardine.

      • Mark V Anderson says:

        I had to wear a tie for the first 20 years or so of my career. I haven’t had to wear a tie for the last 15-20 years (and I guarantee I will never wear a tie if I don’t have to). It is totally subjective, but I feel I am more productive without it. Of course ties are restrict the air passage way, so maybe ties are a special case. But it does seem like dressing “sharp” usually means less comfortable than I prefer, and I think better when I am comfortable.

        • Well... says:

          A tie shouldn’t literally be choking you. Some restriction happens, but it’s really from buttoning the top button, not from the tie. Unless you’re yanking that tie way too tight.

          I hate wearing collared shirts, dress shirts most of all, but as long as I’ve got a dress shirt on for whatever reason, I’m usually happy to put on a tie as well. The opportunity cost is low at that point, and ties are kinda fun.

          • Le Maistre Chat says:

            What Well said. A necktie doesn’t have to be tighter than your shirt collar, and it’s one of a man’s few opportunities to wear colorful silk. But if you feel uncomfortable with your collar done, that’s understandable.

          • Mark V Anderson says:

            A tie shouldn’t literally be choking you. Some restriction happens, but it’s really from buttoning the top button, not from the tie. Unless you’re yanking that tie way too tight.

            Well, yeah, but part of wearing a tie is buttoning the top button, unless one really does want to look like a slob. Maybe it’s that top button that I really hate, but it comes as a package deal. I wear a dress shirt every day to work, but I never button that top button. I am way too fond of my breathing (plus of course you look like an idiot if you do that without a tie).

          • Don P. says:

            Sorry to reply in an expired OT, but: does everyone know about collar extenders?

        • Michael Handy says:

          You may be like me, off the rack shirts that sit right on me have too-small collars, so ties choke.
          I get my shirts custom now (not as expensive as you might think.), but a tailor should be able to alter the collar for a fraction of the shirt cost.I can’t say this enough, Get A Tailor, no exceptions.

          As for comfort, I find smart clothes to be MORE comfortable in many settings. They fit better, they don’t ride up, the fabric feels softer, they breathe better. But you do pay for it, it can’t be faked with a $50 H&M polyester blazer.

          • yodelyak says:

            Echoing everyone else… As kids, all the guys in my family had thick necks, and hated wearing collared shirts with the top button buttoned. A tie was a bit further of an insult, but the damage was in the top button. But we all slimmed up as adults, and/or started having control over what neck size our shirts would be, and collared shirts stopped sucking.

          • Nancy Lebovitz says:

            I’m surprised that apparently no one has figured out making cheap or cheapish shirts with larger necks.

          • Well... says:

            @Nancy Lebovitz:

            Think about it: it would look weird. One of the things a man’s shirt or blazer is supposed to do is make his shoulders look big, or at least not make them look smaller. If you widen the neck-hole, first of all it’s not going to fit most men who have proportional necks, and then for those few guys with really thick or fat necks, it’s going to make them look like they don’t have shoulders, and considered over the life of owning a shirt, that’s probably worse (i.e. less desirable to the average clothing consumer) than being a bit uncomfortable for part of the day.

          • Aapje says:

            @Nancy Lebovitz

            That’s only worthwhile if enough men are willing to pay a premium and/or sacrifice choice AND know enough about fashion to know that this will make their lives better.

            I have my doubts…

          • Nancy Lebovitz says:

            My impression is that quite a few men find ties to be uncomfortable, though the proportion who know that the problem is really the shirt is smaller.

            Anyone have a feeling for the numbers?

          • I find a tie plus top shirt button uncomfortable. When I was an undergraduate coat and tie was required in the dining hall, so I would carry a rolled up tie in my pocket to put on for meals. I attributed my reaction in part to having done judo, where choke holds are legal, for many years.

        • A Definite Beta Guy says:

          I think wearing a tie automatically inflicts a 30% productivity hit. I can think of very few guys that enjoy wearing a tie, and pretty much everyone rips it off the moment they are able.

          As one of the other posters mentioned, dressing smart in correctly fitting, high quality materials…is actually quite comfortable! It probably won’t match the comfort of a baggy sweater and sweat pants, but I find some decent chinos and an ironed shirt are substantially more comfy than blue jeans (which I think are highly overrated) and a baggy t-shirt (or low quality dress shirt).

          EDIT: Should add that I am a cheap-skate, so I buy all my clothes off-the-rack at Kohl’s or occasionally Macy’s. I like those Calvin Klein “steel” shirts, or whatever they are called. They tend to cost twice as much as the other dress shirts, but they are far, far more comfortable, keep up their looks better, and are much easier to iron.

          Though I have discovered the local goodwill has a ton of awesome stuff for about $5-$10, so I am going to shop there soon. Yeah, I am definitely cheap…

          • Well... says:

            You sound about as cheap as me. I refuse to even step into Kohl’s or Macy’s unless I know the shirts I’ll be looking at are marked down 50% or more.

          • A Definite Beta Guy says:

            Birds of a feather. I also only do the 50% sales!

          • There is a place in Boston called “The Garment District” that sells used clothing by the pound. My daughter likes to shop there when we visit Boston. It provides a wider range of styles than shopping for whatever is currently in fashion, at a very low price.

            As another inexpensive source, let me recommend, which I have used mostly for pants for fifty years or so, and my father before me. Mail order clothing for cheapskates. They have a wider range of pants sizes than most places and better prices.

            Quality varies–the flannel shirts are noticeably less nice than the more expensive ones from Lands End or L.L.Beans, but I like the pants and some of the other things.

          • Aapje says:

            My very limited thrift store experience in a different country is that it seems a very viable way for women to pick up nice clothes, but less so for men. I presume that the reason is that women tend to wear their clothes less, get tired of perfectly good clothes more easily and are more willing to accept having to dig through more clothes to find a good deal (aka they tend to like shopping more). So demand and supply is probably better.

            I ended up helping the woman I was with to pick out quite a few ‘new’ clothes from quality brands (at least, that is what the salesperson said, neither of us are fashionistas).

    • anonymousskimmer says:

      I could see this backfiring for adults, as years and years of extrinsic motivation starts screwing with a sense of duty toward one’s self image.

    • JustToSay says:

      I haven’t read the linked article yet, but just yesterday I was talking to a woman who said that when she was in elementary school she did better on timed math fact tests when she pretended to be someone else. I don’t know why I didn’t ask her who she pretended to be; I’ll have to ask her sometime.

      If I imagine imagining myself in the third person as I work (if that isn’t too meta), I think I’d perform better because it triggers the idea that someone’s watching me. Like I’m a productive human animal on a Discovery Channel show. “Watch…as she completes her tasks with record efficiency.”

      • Michael Handy says:

        I’ve done that at Uni. Isolating Malaria in blood samples becomes much easier when you’re Rothbert Thrippingdon, Master Haemotologist!

    • outis says:

      I’m not a child but I’m going to give it a try anyway. Batman is going to work really hard on that project.

      …does the activity have to fit the character’s personality, though? “Goku is working hard on this project” doesn’t seem like it should work.

      • Whatever Happened To Anonymous says:

        Just pretend the project involves punching really strong guys.

      • MrApophenia says:

        Fortunately, even if it is character specific, Batman should work for basically everything except “Batman is going to maintain a healthy work/life balance.”

    • powerfuller says:

      This sounds like a good argument for school uniforms, assuming the uniforms are Batman costumes.

    • andrewflicker says:

      I’ve always been fond of a particular exercise when faced with a challenging intellectual problem: Try to think of what a *smarter* person would do here- what steps would they take, etc.

    • baconbits9 says:

      I feel like the simplest explanation would be that it is easier to do boring tasks when you make them a little less boring. The question I would have to resolve this is how many times can you pretend to be batman and have that positive edge?

      • Matt M says:

        Right. Definitely seems like this is subject to diminishing returns. I’d expect if they kept repeating this experiment with the same kids day after day, the effect would be smaller each time, until eventually disappearing entirely.

        • baconbits9 says:

          Then you just run the reverse and make kids do boring stuff but tell them they aren’t allowed to pretend they are Batman.

          • anonymousskimmer says:

            Batman has a secret identity he must use subterfuge and lies to keep. Telling the kids they can’t pretend that they are Batman would just feed into the roleplaying that much more. This would also lead them to the Batman-fed understanding that deceit is good and necessary.

            Before you know it the kids are stealing from their parent’s purse to buy toys the same way Bruce Wayne ‘steals’ from his company to build his crime-fighting equipment.

    • Nornagest says:

      See if you can sell Google on this. It’d be more fun than open floor plans.

  12. Well... says:

    I intend this as a philosophical inquiry, but it might be culture warry. I’m not sure. If it is, then someone say so and we can just leave it alone.

    Is the suspension of our basic legal rights ever justified? If so, when?

    • The Nybbler says:

      Is the suspension of our basic legal rights ever justified? If so, when?

      Actual imminent existential risk (for the nation, at a minimum) if and only if they are not suspended. I cannot think of a plausible example.

      • Well... says:

        Do you think the Union during the Civil War counts?

        (As I’ve heard it,) Lincoln suspended certain Constitutional Rights such as protesting the draft. Without the draft he may not have been able to raise the army he needed to win the war.

        • JonathanD says:

          I believe that one is justified in the Constitution itself, though. “The Privilege of the Writ of Habeas Corpus shall not be suspended, unless when in Cases of Rebellion or Invasion the public Safety may require it.”

    • Evan Þ says:

      Yes. You have a basic right to life; it can be suspended when you commit a crime that deserves capital punishment. (Or, at the very least, when killing you is necessary to defend against other crimes you’re about to commit.)

      Similarly, you have rights to liberty and property; they can be suspended to some degree when you commit just about any crime and incur fines or a prison sentence.

      • Well... says:

        Oops, I should have clarified, I meant “our” more literally. Everyone’s at once.

        • Matthew S. says:

          (The modern) Battlestar Galactica had a number of episodes about this question, including one extremely culture-warry one.

          • Loquat says:

            If I’m thinking of the same extremely culture-warry one you are, I don’t really recommend that episode. The central issue, and supporting details, were all introduced in that episode and never addressed again after it, and most of the arguments about it felt like the writers just transplanted modern America’s arguments into the setting.

            Also, speaking of never addressing the issue again, the final decision was heavily undermined by a later episode in which we’re told that fvtavsvpnag ahzoref bs lbhat zbguref ner univat gb ghea gb cebfgvghgvba gb fhccbeg gurve puvyqera, juvpu frrzf yvxr n cerggl tbbq ernfba gb jnag gb nibvq orpbzvat n zbgure.

            The show was okay when the liberty-vs-security issues were directly relevant to their main plot, not so much in the stand-alone filler episodes.

    • fortaleza84 says:

      In theory, yes, for example Nybbler’s standard of “actual imminent existential risk.” But as a practical matter, the problem is that evil people will convince themselves and others that their desire to suspend their out-group’s legal rights (and conveniently accumulate treasure and/or power for the in-group) falls into such a category. Since this thread is CW-free, I won’t give specific examples of “major threat to public health/safety” being offered as an excuse to take away peoples’ rights, but they are easy enough to think of.

      So I would say no, it’s never justified.

    • Baeraad says:

      In theory I would say yes, because nothing is ever absolute and trying to make it that way is a recipe for stubborn idiocy.

      In practice, though, I can’t think off-hand of any situation dire enough to justify it that would actually happen in real life with any frequency, so for all intents and purposes I think we can say that it’s never justified.

      So I’m idealistically pragmatic, but pragmatically idealistic, I guess?

    • gbdub says:

      The draft seems like the obvious one – the army is not known for being particularly interested in your rights as an individual.

      We seem to have mostly settled on the draft being okay for survival-of-the-nation wars, or for times when it’s basically just civil service with little risk of actual warring, but not okay for things that feel like wars of choice.

      • bean says:

        We seem to have mostly settled on the draft being okay for survival-of-the-nation wars, or for times when it’s basically just civil service with little risk of actual warring, but not okay for things that feel like wars of choice.

        I’ve seen suggestions that one of the big drivers in discontent over the draft in Vietnam was how few people had to go. In eras when basically everyone went, it was seen as a form of civic duty, but when lots of people get to get out of going, you get resentment. Of course, the big reason behind the low number drafted was rising cost per man, so the draft as we knew it is almost certainly dead forever.

    • Wrong Species says:

      There’s no neutral way to talk about this. Either you think a certain right is important, in which case the suspension of rights necessitates extreme circumstances or you don’t, in which case, it’s only useful to the extent it serves other purposes. If you want to talk about which rights are important, you bring in the culture war.

      • Well... says:

        Maybe we can evade both horns of the dilemma by focusing on a historic example: Lincoln’s suspension of certain Constitutional rights during the Civil War.

        • Wrong Species says:

          That’s not a good example because the Civil War is very politicized. A better example would involve something that almost everyone agrees is a basic right that we should have strong protections for. I’m struggling to think of a good example though.

          • orihara says:

            I’ve never seen anyone argue against self-incrimination?

          • Well... says:

            @Wrong Species:

            The Civil War is politicized insofar as it deals with race and slavery and the legacy of the South and so forth. But I don’t think the topic of what sacrifices the Union took in order to win is hotly political these days.

            And besides we can put all that aside and just say “suppose the Union’s winning the Civil War meant ending slavery and that slavery is as evil as everyone says.” So we put that as non-contestable and discuss whether or not Lincoln suspending people’s basic rights was justified and, more importantly, whether any principles from that discussion can be generalized.

            For instance, suppose a ban on guns prevented X% of deliberate gun deaths per year (I’m not saying I believe it would, but for the sake of argument let’s say it would), but at the risk of making it harder to resist tyranny. Is it worthwhile?

          • Wrong Species says:


            Are you stipulating that suspending those rights is necessary to win the war or is it uncertain? Because if it’s necessary, then yeah, I would immediately say it’s was the right thing to do. But in a world of uncertainty, I’m much more skeptical.

            The meta-principle is that we should all be moral pluralists to some degree. If someone is willing to sacrifice all other values for one single value, they might as well be a paper clip maximizer.

          • Well... says:

            It is uncertain in an absolute sense, but I’m supposing you believed that suspending those rights is necessary to win the war. It sounds like in that case you’d say Yeah, go ahead and do it. Which of course is what Lincoln did.

            Did Lincoln believe it was absolutely necessary, do you think, or was there something else going on?

          • Wrong Species says:

            A lot of times philosophers will make some stipulation that leads you to one conclusion, while in the real world, you would never do it. The best example would probably be the variant of the trolley case where we stipulate that pushing a certain fat man on to the tracks would save five people further down the track. Sure, within the confines of the experiment it’s equivalent to simply changing the trolley track. But in the real world, if you choose to do that, you’re probably either an idiot or a sociopath.

            With the Civil War example, it’s not enough to have the belief that it’s necessary. You should have a justified belief, and even then you might be wrong. Lincoln may or may not have believed his arguments but it sure was convenient for his position. I’m very skeptical of the claim that it was necessary to suspend free speech for him to win the war. But if I was somehow convinced, then it seems clear to me that abolition is worth the temporary suspension of certain Constitutional rights.

    • John Schilling says:

      As with torture, or running trolleys over innocent bystanders, this is one of those cases where in theory it might be justified but in actual practice it’s about 100% Evil Villains trying to justify their Evil Villainy by propping that theory in front of some very dubious facts. Conveniently, your right to examine and question those facts is one of the first to be disposed of in the name of dire necessity.

      So a strict policy of Never Ever Do This, seems best.

      Conscription might have been an exception in the days when a nation’s survival was best ensured by lining up the largest possible number of men drilled in basic musketry; one of the benefits of modern warfighting techniques is that this doesn’t work any more.

      • Mark V Anderson says:


      • Well... says:

        Conscription might have been an exception in the days when a nation’s survival was best ensured by lining up the largest possible number of men drilled in basic musketry

        Indeed. During the Civil War, Lincoln locked up thousands of people merely for protesting the draft. Suppose his doing this helped the Union win the war and end slavery. Was it justified?

        • John Schilling says:

          Well, it clearly didn’t end slavery, because we were still conscripting people a century later.

          But of course you mean it ended people other than the government owning slaves which is somehow extra specially bad and so justifies slavery by the government to stop. Yeah, I’m going to call that one a fine theory that leads me to suspect the underlying motives, ask whether maybe you’re in it merely to preserve the almighty power of the Union, and suggest you look real hard at ways to end slavery without conscripting millions of people to fight a bloody terrible war.

          • Wrong Species says:

            It should be obvious that chattel slavery is “somehow extra especially bad” compared to the draft. Do you actually believe that being conscripted is just as bad as being a slave in 1865 Mississippi?

          • Sfoil says:

            Why is it obvious? Conscripts were almost certainly more likely to die violently, for one thing.

          • Well... says:

            Conscripts were (in most cases, I think) paid for their service, which only lasted a few years. And then after WWII they were showered with benefits through the GI bill. Plus, veterans have long commanded a special respected social status, even back when there was a draft.

            There are tons of other comparisons we could make to show how different conscription is from chattel slavery, and how ludicrous it is to equate them.

            Seems like a very disingenuous argument to me. Libertarian virtue signaling maybe? And in a non-CW thread! Tsk tsk.

          • Sfoil says:

            I was specifically talking about men drafted into the Federal army during the American Civil War, a conflict which was unusually dangerous to its combatants, and who certainly did not benefit from the Montgomery GI Bill. If you want to talk about any form of conscription anywhere ever, then you also have to talk about any form of slavery anywhere ever. There are worse times and places to be a slave than the American South in the mid-19th century, and better. Likewise for conscript soldiers.

            Conscripts, like slaves, are paid (when at all) at below-market rates. This is as true for modern armies as historical. Further, the social status of veterans is not at all universal across space and time. Palace eunuchs enjoyed elevated social status in many historical societies as well but that’s not an argument for castrating slaves.

            Both conscription and slavery are involuntary servitude requiring coercion to enforce and maintain, and it requires more than assertion to demonstrate that well of course slavery is always bad but conscription isn’t. The standard libertarian position is to declare them both morally wrong.

          • Well... says:

            I was specifically talking about men drafted into the Federal army during the American Civil War, a conflict which was unusually dangerous to its combatants, and who certainly did not benefit from the Montgomery GI Bill.

            OK, good. I think it’s important to specify cases because you’re right: the nature of both slavery and conscription have varied widely over history and geography — and sometimes have been quite literally one in the same, as when a conquering people forced captured men to fight for them in subsequent battles.

            So, comparing conscription into the Federal Army during the Civil War vs. slavery as practiced in the South during the Civil War, I’m not sure which institution was worse on the whole. But slavery still seems obviously less moral to me.

            But you’re right, I don’t think I can completely articulate why, I would only just be asserting. If I tried to articulate it, I guess I would start with the fact that as a conscript in the Federal army, you are at least born free and get to return to your family free, provided you survive. Between conscription and slavery I don’t know which path is less likely to result in death or maiming.

            Would you say that suspending the right to protest the draft during the Civil War was unjustifiable, even if we take it as a given that allowing the protests would have compromised the war effort and have allowed the South to win the war and continue practicing slavery?

          • dndnrsn says:


            This is a separate nit to pick, but, was the American civil war unusually deadly? Around 1/5 of Civil War soldiers died; this isn’t necessarily deadlier than usual for a major war in that period (of 3 million French soldiers under Napoleon, just under a million died, which is about 1/3 – medical care was about the same, which was the major factor in deaths in both; numbers for non-French are spottier) or in general (even with better medical care about 1/6 of soldiers in WWI died; I’m having a harder time quickly coming up with a number for # of soldiers in WWII, but I would guess overall it’s about the same – while medical technology was better in WWII, it was not evenly distributed; the Western Allies, esp. Americans, had the best, but they did a minority share of the fighting).

            The Civil War was deadlier than usual for wars Americans fought in, but the US suffered considerably less than the norm in both World Wars.

          • John Schilling says:

            I would say that the Civil war was of roughly average badness (normalized for scale) for wars of the mid-19th century, but the mid-19th through early 20th century was about the worst time to be stuck fighting a war. Conscription and logistics had raised mass armies to as large a scale as they could be, firepower had reached the point of greatly impeding maneuver and making a frontal attack across open ground nigh-suicidal so we couldn’t settle a war with a few decisive battles, we didn’t yet have armor or air power to break the stalemate and at least put an end to the bloody mess, and we didn’t have antibiotics to deal with the disease problem of having millions of people stuck in trenches for years on end.

            I also think that being too quick to argue the relative badness of various forms of slavery/conscription/war to justify one over the other, risks turning into a blanket justification for the allegedly lesser evil when you ought to be looking for not-evil solutions.

          • dndnrsn says:

            Armour and air power broke the trench deadlock, but having a faster-moving front line seems mostly to have made life more miserable for civilians (eg, Russian military:civilian deaths in WWI were about 2:1; Soviet about 1:1.75 or so, with the Soviet civilians being killed by direct action alone roughly equalling military deaths).

          • Iain says:

            I’m no expert here, but:

            The United States first employed national conscription during the American Civil War. The vast majority of troops were volunteers; of the 2,100,000 Union soldiers, about 2% were draftees, and another 6% were substitutes paid by draftees.
            Of the 168,649 men procured for the Union Army through the draft, 117,986 were substitutes, leaving only 50,663 who had their personal services conscripted.

            Is it obvious that 4M people condemned to a lifetime of slavery is worse than 50K people forced into combat, roughly 10K of whom were statistically expected to die? We report; you decide!

          • Wrong Species says:


            I explicitly asked about slavery in the south, not palace eunuchs. I would much rather be conscripted in the Civil War than be an antebellum slave. Wouldn’t you?


            It would be nice if there was always a third option but sometimes there isn’t. Yes, we should always look for one and people will make dubious arguments to expand their powers. But sometimes, it really is just the two bad options and you have to pick the lesser one. Of course, you can argue that temporary bad things are worth it to uphold some value but you still need to make that calculation yourself. Also, I don’t think conscription is nearly as bad, relative to antebellum slavery, as you do.

          • Chalid says:

            The fact that so many people volunteered for the war is itself an indication that the experience of being a Civil-war era soldier is much better than the experience of being a slave. You didn’t see many volunteers for slavery.

          • John Schilling says:

            That’s about right, but we’re apparently being asked to hypothesize that the draft was necessary, which presumably means that we’re in an alternate universe where you can’t fill 98% of your ranks by way of idealism or money.

            Mostly, it’s probably best to reject hypotheticals like this, particularly ones where some parts but not others are designed to match controversial real scenarios down to the name. Or the general form, “Assume $UltimateEvil will happen unless we do things my way. We should do things my way, right?”

          • Sfoil says:

            I would rather be a Union conscript than an antebellum slave, but then I also volunteered in real life so I might be biased. The odds of dying as a Federal soldier were somewhere between 20-25% depending on how you account for non-combat casualties. Choosing this over antebellum slavery requires placing a relatively low value on one’s own life. On the other hand, antebellum slavery was hereditary and manumission seems to have been relatively uncommon.


            This is a separate nit to pick, but, was the American civil war unusually deadly?

            For an American? Absolutely. For a gunpowder war perhaps about average.


            Is it obvious that 4M people condemned to a lifetime of slavery is worse than 50K people forced into combat, roughly 10K of whom were statistically expected to die?

            Actually, I do think it was worth it (by the way, you’re forgetting to take into account Southern conscripts, who would have kept their freedom and lives if there wasn’t a war). However, claiming that it’s “obvious” relies on a lot of assumptions. What if the slave states had agreed to emancipate every slave on the condition that 10,000 military-age male slaves were first selected at random and shot? The conditions and treatment of the freed slaves would be no different than historical, of course.


            You didn’t see many volunteers for slavery.

            No, but plenty of people volunteered to engage in agricultural labor or domestic service, which is a more relevant comparison.

          • Well... says:

            @John Schilling:

            Some background on why I created the OP…not for discussion because it’s probably too close to CW, but just to explain why I’m framing my hypothetical the way I’m framing it. (TL;DR: I’m steelmanning something.)

            A family member of mine regularly writes outside editorials to major newspapers, and emails them to me and several others for review before he submits them. These email chains often turn into discussions.

            His most recent editorial was in favor of some pretty harsh restrictions on gun ownership as a preventative measure against further mass shootings. He conceded that he was calling for a reduction of 2nd Amendment rights but supported his argument by claiming we are in a crisis, and pointing out that Lincoln suspended constitutional rights during the Civil War, specifically when he locked up thousands of draft protesters, merely for protesting the draft.

            I wrote him back that we are not obviously in a crisis, but regardless, it isn’t clear that Lincoln was justified in suspending basic Constitutional rights. After I sent that email to him I was still thinking about it, and I wondered under what circumstances suspending those rights might be justified, but especially what others thought specifically about that Civil War scenario.

            So I set it up so that all the priors one might need to make my relative’s argument were already there as a given, to see if anyone had any other objections.

            What’s funny is when I typed the OP I included a lot of this background, but decided to shorten it because who wants to read all that? Just get to the point! Which is what I tried to do.

          • Wrong Species says:


            The relevant comparison isn’t between the lives of conscripts and slaves during the war. It’s for their entire lives. If you want to argue that being a slave is preferable, you have to start from the moment they were born. How do you think their life expectancy compares to the average conscript?

            What if the slave states had agreed to emancipate every slave on the condition that 10,000 military-age male slaves were first selected at random and shot?

            First, who would have expected this outcome? That’s such a weird hypothetical to try and even the odds. Second, we’re talking about the events as they played out, not in some alternate world.

            No, but plenty of people volunteered to engage in agricultural labor or domestic service, which is a more relevant comparison.

            Why would you think that’s a relevant comparison?

            You would have chosen conscription. I would have chosen conscription. Everyone in this thread would have probably chosen conscription. Why is it so hard for you to admit that it’s obvious that being a conscript is preferable to being a slave in the Antebellum south?

          • Sfoil says:

            I think we’ve gotten bogged down in specifics about the American Civil War. The point is that any difference between “conscription” and “slavery” is basically one of degree rather than kind. I agree that for most people a two-year term in the Federal army, even with a 25% chance of death, is better than being born an antebellum field slave. I’m much less sure that a 20-year term in the army of Imperial Russia is obviously preferable to being the domestic servant of some Mediterranean spice merchant.

            Why would you think that’s a relevant comparison?

            Volunteering for the Army:Volunteering to pick cotton::Forced into the Army:Forced to pick cotton. That being said, I’m going to back off of that one because I don’t think it captures what actually made slavery worse than conscription, in America in the 1860s.

            Of course that hypothetical about shooting random slaves and releasing the rest is silly. However, it’s nearly identical to the claim made by Iain that it was “obvious” that 10,000 people dying in involuntary servitude in order to free the slaves was worth it.

          • Iain says:

            I didn’t claim that 10,000 dead conscripts was “worth it”. Note that “obvious” was borrowed from Wrong Species, not my own word choice.

            I simply pointed out the actual magnitude of the bad things under discussion. Contra John Schilling, I think it is critical to consider relative badness, because the world is not so kind as to only ever give you easy choices. On that narrow ground, it seems clear to me that American chattel slavery circa 1860 was a greater evil than Civil War conscription.

            Of course, that doesn’t mean the lesser evil was justified. Everybody in this discussion would have preferred it if the Union could have won without conscription. Was that possible? It’s not an easy question, particularly at the time without the benefit of hindsight. What I will say is:

            1. If you postulate that conscription was necessary, I think the trade-off is worthwhile.
            2. If you postulate that sfoil’s “murder 10,000 slaves and set the rest free” was genuinely offered, then I think it would have been worthwhile.
            3. #1 is a couple orders of magnitude more realistic than #2.

    • Orpheus says:

      Does quarantining a large city/are in case of a pandemic count?

      • Well... says:

        Yeah, that might, and it’s an interesting example. What do you think about it?

        • Brad says:

          Quarantining a whole city seems to be more than is needed to answer the question. What about a single individual — one person has virulent and deadly disease and the government locks him away in an isolation ward. He hasn’t done anything wrong and certainly been convicted of doing anything wrong, but his civil rights are pretty clearly being circumscribed.

          This seems like one of those situations where if your ethical / political theory gets the wrong answer then there’s a probably a problem with your theory, not with the answer.

    • SamChevre says:

      I would say suspension of basic legal rights is justified given three criteria:
      1) The suspension is evenly applied
      2) The risks are very high
      3) The suspension is temporary and short-duration

      So, justified cases:
      1) A curfew or no-entry order after a natural disaster
      2) Mandatory evacuation before a natural disaster
      3) Quarantine for a contagious disease

      • albatross11 says:

        As a practical matter, it seems very important to work out why this special emergency power/rights violation/etc. won’t go on indefinitely. Otherwise, we start out with a discussion about how torture would be okay in this one really contrived case involving a hydrogen bomb set to incinerate Manhattan, and we end up with every city police department running a torture chamber to extract confessions in small-time drug cases.

  13. Collin says:

    At some point in the last year Scott had a post that covered, if I recall, various ways to look at the world and prioritize the preferences of all living beings. He showed different ways that some of the perspectives were broken by making examples using small populations (for instance, increasing the average well being was a poor goal because it could be achieved by killing unhappy people which would be against their preferences), and where others succeeded and could scale.

    Does anyone know what post this was?

  14. VirgilKurkjian says:

    Anyone hear about the link between PFASs and body weight? Seems like potential support for the leptin theory of obesity.

    • outis says:

      What’s the best way to avoid endocrine disruption if I have to live in America?

      • VirgilKurkjian says:

        No idea, unfortunately. I’m a cognitive scientist, so your guess is as good as mine. Love to hear if anyone else has any ideas.

  15. rahien.din says:

    A few OT’s back, there was a brief discussion of harsh vocal styles, some related to certain folk music traditions, other related to black metal and death metal singing. Some of us were trying to learn how to sing in those styles, and I was wondering how y’all were progressing?

    • pontifex says:

      Is anyone working on a screamo version of the Sequences?

    • Thegnskald says:


      I have the basic technique of throat singing down. Working on expanding my pitch range there.

      I am still experimenting with death metal growls; I can get some cool noise, but haven’t developed any consistency yet, so I can’t necessarily repeat it.

      Black metal growling still somewhat eludes me. I haven’t found the muscle movements to produce the noise in any kind of consistent way.

  16. We are having another South Bay Meetup on Saturday, March 10th, starting at 2 P.M. Details webbed.

  17. johan_larson says:

    Goodness, the Norwegians are just running away with these games, aren’t they? Right now they have 26 medals, ahead of second-place Germany with 18. Very impressive for such a small country.

    • Randy M says:

      Asked to comment, the Gold-medalist replied “Batman just gave 110% on this Luge, as did the entire Justice league, in fact.”

    • Mark V Anderson says:

      Yes, I’ve been wondering myself why the Norwegians are doing so well. Are they so rich with all their oil that lots of folks (or at least 20 somethings) spend their time on athletics? Does Norway spend a lot of money on Olympic sports?

      • johan_larson says:

        If I had to guess
        1. Norway has a strong grass-roots tradition of skiing.
        2. There are a lot of medals to be won in skiing, rather like running or swimming in the summer games.
        3. Norway has spent plenty of money cultivating its elite athletes.

      • Fossegrimen says:

        Norway is really weird.
        An ambitious parent in the US could tell a teenager to work hard and get into Harvard. In Norway where tuition is free and available to ‘all’ and nobody cares which university you went to, the parent would tell the child that ‘if you want a good job in finance, you need to shave 10 seconds off that pace of yours’.
        I have consulted on several places where it is common to conduct job interviews on the cross-country ski tracks (or on a bicycle in the summer months).
        This is also the reason for DavidFriedmans observation above; If the weather is good and you spend your time in the office, you can forget that promotion in some companies.
        Norway has a lot of prepared ski track and a lot of people use them. Every little hamlet has miles and miles of prepared and illuminated tracks so you can ski in the darkness.
        Then they take all those very fit and ambitious 15 year-olds and throw money at them. I read somewhere that every single Norwegian cross-country skier is backed by more money than the entire Swedish team. (Since this was journalism, you probably have to twist the actual facts into pretzels to make the claim true, but there is at least a significant difference.)
        It’s cold up there.

  18. fr8train_ssc says:

    Now that SSC is home of the the world duplication illusion, does this mean that “This is the the day” is now the official theme song? (

  19. tayfie says:

    Damn it, Scott. It’s on the header now and I still can’t see the duplicates the first time.

  20. tayfie says:

    This is more of a policy and political incentives discussion, so I think the risk of culture war is minimal.

    One of the common criticisms against consumption taxes is that they are bad for the poor who consume a greater percentage of their income. Consuming a greater percentage of income means a greater percentage of income goes to taxes. This is a regressive tax, which is widely agreed to be unfair. Right or wrong, most people accept the proposition that those with more ability to pay the tax burden should take on more of it.

    Enter the progressive consumption tax. There is nothing about a consumption tax that necessitates it be regressive. The wealthy still consume more as an absolute dollar amount, and they consume many more luxuries. A fairly simple first order approximation is to have a sales tax percentage that increases with the cost of the goods sold.

    Why is there no mainstream political proposal for such a thing? It sounds to me like a better way to handle inequality than a progressive income tax. People only care about income inequality in the first place because it enables a different sort of lifestyle. It makes more sense to just tax the fancy lifestyles directly rather than taxing income as a proxy. This is also less destructive than other taxes because it does not discourage earnings, savings, or investment.

    Such a proposal would also give people much more control over the taxes they pay as spending is something people control immediately with their own decisions about what and how much to buy whereas they don’t have the same kind of control over their income. It would be conceptually simpler than the income tax. It would decrease the price of labor relative to other goods, which would increase employment. A competent politician could sell a policy like this ten different ways.

    An obvious issue would be the large transition costs, but is there anything else I am missing? This feels like it could be a big opportunity for someone who wants new ideas.

    • Fluffy Buffalo says:

      This sounds like it would be exploited in ways that would add a lot of inconvenience for the sake of saving a few bucks: sellers would try to split one item into two, three, or a dozen, adding up to the same pre-tax value, but incurring less sales tax. Incentivizing people to play stupid bookkeeping tricks instead of using economies of scale as far as possible doesn’t sound like a good idea.

      • tayfie says:

        Are you saying that the current tax system doesn’t encourage stupid bookkeeping tricks? Why would the tricks you fear be any worse for this than what we have now?

        I’m not even sure a tax system could be designed to prevent such “tricks” because the incentive to pay less taxes would always be the same. A gigantic mass of people looking for every possible loophole is nearly impossible to plan against.

    • This is a regressive tax, which is widely agreed to be unfair. Right or wrong, most people accept the proposition that those with more ability to pay the tax burden should take on more of it.

      Your second sentence doesn’t support your first. Even with a regressive tax, richer people pay more than poorer, so take on more of the burden. They pay a larger sum but a smaller fraction of their (larger) income.

      • episcience says:

        Steelmanning, I think tayfie was reflecting the orthodox consensus position that tax rates should be inversely proportional to the marginal utility of additional income. Or, in other words, instead of “dollars more” of the tax burden read “proportionally more”.

        • I think the common phrasing that tayfie was echoing originates as a motte and bailey argument. “Pay more” sounds more obviously right than “pay a larger fraction of their income.”

          Which is why I bothered to point out the error.

          Scott made a related mistake in his ancient Non-libertarian FAQ, when he wrote “50% of what a person with $10,000 makes is more valuable to her than 50% of what a billionaire makes is to the billionaire.” That does not follow from declining marginal utility of income, which he was trying to base it on.

      • tayfie says:

        You are correct that my phrasing was not exact. I should have specified “proportionally” more.

        And neither of those sentences are my personal views. I was merely trying to state in simple terms what the majority of people believe in my experience.

    • Aapje says:


      Almost all of Europe already has 2 (or more) tariffs, where food is commonly taxed with the lower tariff. The only exception seems to be Denmark.

      • anonymousskimmer says:

        In the US food is generally exempt from the sales tax, or taxed at a lesser rate (depending on state).

    • outis says:

      Many European countries have variable sales taxes, with various categories of products (e.g. food) getting a lower rate. One possible downside is that this gives politicians yet another level to arbitrarily favor some things over others, e.g. books will typically be taxed at a lower tax rate (because education! culture!), but computers will be taxed at the luxury (general) rate, even though they’re strictly more useful than books.

      BTW, I strongly dislike the fact that a flat tax is called “regressive”, even though it still means that richer people pay more. I don’t think it’s right to concede that anything not maximally progressive is automatically regressive. Therefore, I support having at least one actually regressive tax on the books (e.g. one where poorer people pay more), which is the only way from keeping the term “regressive” from being abused.

      • Aapje says:


        In the EU, e-books that are downloaded are taxed at the higher tariff, because the low tariff is for printed works, but e-books that are downloaded are consider a digital service. The lower tariff does apply if you send someone a memory stick with an e-book on it.

      • Therefore, I support having at least one actually regressive tax on the books (e.g. one where poorer people pay more), which is the only way from keeping the term “regressive” from being abused.

        You are now abusing it in the other direction. An actually regressive tax is one where poorer people pay a larger fraction of their income, not one where they pay more money. A flat tax is the dividing line between regressive and progressive. Your version would make a head tax the dividing line.

      • The Nybbler says:

        We have at least one regressive tax in the US — Social Security, 6.2%* from bottom dollar up to $128,400 in wage income.

        * yes, I know, it’s roughly double that because of the hidden “employer contribution”.

        • Nornagest says:

          Most sales taxes are regressive in practice, too — richer people buy more expensive stuff, but not enough to make up for lower consumption as fraction of income. And the most expensive stuff they consume (housing, education) doesn’t get hit by it.

          Food’s exempt too, but prepared foods usually aren’t.

    • episcience says:

      See also the Hall-Rabushka tax, which combines the administrative simplicity of a flat tax with the economic benefits of a consumption tax.

      One thing you elide over is that the economic preference for a consumption tax over an income tax is not for equity reasons but because of how distortionary the tax is. Capital is mobile but consumption is (to a large degree) not, so the view is that by exempting capital from tax the distortionary effect of the tax is reduced.

      There are a few reasons why this might not be persuasive from a public policy perspective, though:
      – The idea of direct taxation (i.e. taxes on income and capital gains), where people are taxed on their net increase in wealth from year to year, has some intuitive appeal. If Anna earns £50,000 and spends all of it on food, shelter, and education for her kids and Bob earns £100,000 and spends only half of it, from a public policy perspective we may want to tax Bob on all £100,000, even if he invests the other half.
      – It’s not clear that the distortionary effects of income taxation are that high — I recall a study that estimated elasticities high enough that marginal tax rates would have to reach something over 60% before there was a significant impact.
      – There are some difficulties (except where you are using the Hall-Rabushka system) in administration. Simply charging more for more expensive goods doesn’t make the tax progressive on increasing consumption, but flat. (If your rate is 20% for iPads, someone buying one iPad pays the same rate of tax than someone buying 20.) So you need some mechanism for tracking annual consumption.

      But I think it’s something worth exploring further — as a tax lawyer I definitely agree that policy wonks should spend more time discussing different tax systems.

      • All of these discussions of taxation share a common error–they confuse the question of who hands over the money with the question of who bears the burden of the tax.

        To see why this is a mistake, consider two possible versions of a sales tax. In one version, every time the store sells a widget it pays the IRS a dollar. In the other version, every time a customer buys a widget he pays the IRS a dollar. Call the price of a widget under the first system P1, under the second P2.

        In the first version, the amount the customer pays for a widget is P1, the price, the amount the seller receives is P1-1. In the second, the amount paid is P2+1, the amount received is P2.

        The amount supplied by producers is a function of the amount they receive when they sell a widget, the amount purchased by consumers is a function of the amount they pay when they buy a widget. Since quantity supplied equals quantity demanded under both systems, P1=P2+1. Hence the effect on consumers and producers is identical–it does not depend on which one hands over the dollar.

        The obvious real example is the Social Security tax, theoretically paid half by employer, half by employee. That is pure window dressing–the effect would be the same if it was paid all by one or all by another. Who actually bears the burden depends on the relative elasticity for supply and demand of labor.

    • fortaleza84 says:

      Occasionally there are proposals for luxury taxes but yeah, it seems like they haven’t made much headway in the United States.

      I would guess part of the problem is that consumption taxes are a lot easier to evade than income taxes. So for example, if you are a fancy New York City attorney, it would be very difficult to convince the taxation authorities that you actually live and work in New Hampshire. On the other hand, it would be much harder to stop you (or monitor you) from driving up to New Hampshire, buying a $100,000 Patek-Philippe watch, and wearing it in New York City; then commissioning a yacht in the Carribean for $300,000, keeping it moored there, and using it only in the Carribean; then going on a fancy ski trip overseas; etc.

      • Nornagest says:

        There was a federal luxury tax in the Nineties, but most of its provisions got repealed when they failed to deliver on expected revenue. Taxes on expensive cars stuck around for a while, though.

      • tayfie says:

        If a wealthy attorney could go to New Hampshire to buy a watch, he would pay the same tax as in any of the states if things were implemented at the federal level. If it were only implemented at the state level, that kind of property sounds like a feature rather than a bug. It would prevent states from setting the taxes too much higher than their neighbours for fear of encouraging people to buy goods in different states. The same is true with the Caribbean ship.

        I am not sure your example makes sense. A person like you propose may really have a house in New Hampshire and may convincingly argue they live there most of the year.

        It also doesn’t seem at all obvious to me that a consumption tax is harder to track. The United States avoided income taxes for a long time thinking they would be impossible to track. It is hard to measure income, which happens over a year. Consumption creates actual exchange of goods or services, which can be directly observed, as well as the records from the sale, allowing the tax to be calculated and paid almost immediately. Are current sales taxes easier or harder to administer than current income taxes? I don’t know, but it could shed some light.

    • SamChevre says:

      Isn’t this very similar to the FairTax proposal? (Flat consumption tax with an exempt amount of consumption.) That may not have gotten mainstream economic support, but it certainly got mainstream political support.

      What I’d note is that there are several different things that one might want a tax system to maximize, and the same tax system will not optimize all of them.

      From a left-ish position, you might want a tax to maximize: (1) government revenue (2) consumption equality (3) wealth equality (4) Equality of INCREASE in wealth. A consumption tax does really well at #1, and OK at #2 (see European VATs); it does very badly at #3 and #4. Allowing tax-free investment tends to benefit investors, who are already wealthier than average. (This is why I sometimes refer to the “pro-equality” movement in the US, with it’s focus on people who earned generational wealth in the last generation, as the Rockefeller protection movement.)

      From a right-ish position, you might want a tax to maximize: (1) visibility (2) proportionality of benefits and burdens (3) alignment between of payers-for and voters-for. A VAT-type tax is very good at #3, but tends to be very low visibility, and less good at aligning benefits and burdens (especially for something like retirement benefits) than a wage tax for retirement benefits.

      I think that on both left and right, a mixed tax system is likely to meet the entire set of goals better than any single tax system. I would agree that the US system is probably over-weighted toward income tax relative to consumption tax, but the Right does not trust that adding a tax category will not increase total tax burden rather than shifting it around (and the experience of states that added an income tax in the last 30 years supports that view), and the Left is not likely to support a tax system that decreases taxes on “the 1%”, so I don’t expect a move in that direction at the national level any time soon.

      • tayfie says:

        Maybe I misunderstand the leftish positions, but I think they care a lot more about consumption equality than wealth or increase in wealth. The rhetoric is always about the rich having their mansions with diamond studded indoor swimming pools while the poor huddle together in a tiny, decayed apartment. Wealth is an instrumental, rather than terminal, target. It’s ok to be rich as long as you don’t enjoy it too much.

        • Rob K says:

          I think you’re missing a piece of the leftish positions.

          A highly unequal balance of wealth in a society begets a highly unequal balance of power in a society. This is a large part of what Piketty and other recent observers of growing inequality are worried about as they’re describing the rise of patrimonial capitalism. Large inequalities of wealth generate disproportionate social and political power, creating the potential for runaway feedback loops;

    • The Nybbler says:

      Trying to draw the line between consumption goods and investment goods seems like an issue. Suppose I buy a $50,000 car now, and in 5 years I sell it for $10,000 and buy another $50,000 car. What’s my tax on that? $50,000 twice? Depreciation over each year? $50,000 in year 1 and $40,000 in year 6 (with the $10,000 being negative consumption)? Same for other durable goods, especially houses, but any other “consumption” good that retains value. Gates’s taxes are no doubt complex already, but making taxes more complex for ordinary individuals seems like a big negative.

      • tayfie says:

        You have brought up a complexity of a hypothetical system without showing it would be more complex than the existing system. How would this be more complex than existing income taxes?

        Furthermore, there seems a rather straightforward solution to the complexity. The tax always takes place at time of sale on the sale price.

        So yes, you would pay tax on the $50,000 twice, once when you bought the first car the first year and once when you bought the second the fifth year. If you sell the car for $10,000, whoever you sold to is responsible for the tax on the $10,000.

        • The Nybbler says:

          The way this consumption tax is described, is I pay taxes on Income – (Net Savings). This is all the complexity of the existing income tax, plus the complexity of figuring out what “net savings” is.

          What will happen is that everyone fills in a tax form, very similar to the 1040 of today. You record your income, as usual, but all of that income that you have saved is deducted before whatever the tax rate is is applied. You can think of this as being a little like all of your savings, of any type at all, as being in a tax-free 401 (k) if you like. However, all of the money that you’ve taken out of your savings that year is added to your taxable income.

          This implicitly requires a split between things which count as “savings” and things which don’t. Real estate and durable goods are edge categories; there are others, such as collectible items. Taxing these as consumption items at time of sale not only results in a huge one-year hit for several-year consumption (because the tax is progressive buying a house or car may bump you up several brackets), but also fails to account for value retained — that is, after 5 years I’ve consumed only $40,000 worth of car, not the $50,000 I was taxed on.

    • Mark V Anderson says:

      Usually consumption taxes are considered regressive because tax is calculated as a percentage of income, and the rich usually invest a higher percentage. I don’t see why we don’t use consumption as a denominator, so that a non-discriminatory consumption tax would be neither progressive nor regressive.

      I am absolutely against a progressive tax. I agree that it is a good thing when the government subsidizes the poor so that no one is destitute. But at least in the US, there are so many different methods of subsidizing the poor (including progressive taxes), that it is impossible to determine how much welfare is really paid by the government. Also all the different programs means that those that are able to game the system receive a lot more welfare than those that can’t or won’t. I favor there being one welfare agency in one government entity (in the US, probably the state level is best). With one agency it is then transparent how much welfare is being paid, and the amount paid can be made in a more rational manner.

      It is true that progressive taxes also even out the income of the rich versus the middle class, which welfare does not do. I don’t think this evening out is a good idea. There are many income differentials where the higher earning person deserves a higher income: they work more hours, they work harder, they are willing to accept worse working conditions or nasty supervisors, they take more risks, they sacrifice to earn more in the future with more education. It is true that more income is often due to undeserving factors, such as inheritance, connections, inborn talent, luck. I don’t trust the government to ever do a good job of sorting out the deserving from the undeserving. Plus of course I haven’t even discussed the factor of incentives, where higher earnings will incent a person to do a better job, which is a benefit to everyone.

      • Le Maistre Chat says:

        I am absolutely against a progressive tax. I agree that it is a good thing when the government subsidizes the poor so that no one is destitute. But at least in the US, there are so many different methods of subsidizing the poor (including progressive taxes), that it is impossible to determine how much welfare is really paid by the government.

        That’s an interesting point I hadn’t considered before. I’m socially archconservative, but always took it for granted that a strongly progressive income tax was the only just kind of income tax, and that consumption taxes were to be avoided so long as any income tax existed.

        • but always took it for granted that a strongly progressive income tax was the only just kind of income tax

          I’m curious as to the theory of justice that leads to that conclusion. There is an obvious utilitarian argument, but that doesn’t feel to me like an issue of justice. There is the argument that income largely depends on things the individual can’t claim credit for, but if you take that line of argument seriously the natural conclusion is not that everyone deserves the same income but that nobody deserves anything, making the income distribution morally empty.

          • Chalid says:

            Couldn’t you get that from “income partially depends on things the individual can’t claim credit for, and partially depends on things he can claim credit for, and the fraction that depends on things he can’t claim credit for rises with increasing income”?

          • John Schilling says:

            and the fraction that depends on things he can’t claim credit for rises with increasing income

            Citation needed.

            And if you’re just going to enumerate a list of external factors that contribute to a billionaire industrialist’s fortune, explain how those aren’t equally important to the salaries that industrialist pays his employees.

          • Chalid says:

            Obviously it’s not citable but it’s plausible. Employee #50 at Facebook got tens of millions of dollars, employee #50 at whatever the #2 social network was got approximately nothing for his efforts, and that’s primarily due to luck and not employee ability or effort. A milder version of this in which pay is tied to company success goes on in most companies in America but only at the medium-to-high levels in the company – middle and upper management are paid in equity and get bonuses when the company does well, while line workers don’t. Equity pay increases the luck involved in your outcomes, and most people paid in equity are not actually meaningfully impacting the stock price.

            Winner-take-all competitions will especially lend themselves to this dynamic – there’s usually going to be a ton of luck involved in who gets to actually be the winner since ability is often well-matched – and winner-take-all games become more and more prevalent as you look toward the right tail of the income distribution.

            To be clear I believe in progressive taxation not due to the above but mainly because of what DF refers to as the obvious utilitarian argument; I mentioned it mainly because it seemed to be a natural response to DF’s post.

          • Couldn’t you get that from “income partially depends on things the individual can’t claim credit for, and partially depends on things he can claim credit for

            The problem is that, once you take the argument seriously, it’s hard to see what one can take credit for. Everything you do is a result of a combination of your innate characteristics and your environment, neither of which you caused.

            One could try to make a free will response in which your decision to work hard is a result of your willing it–but what is the cause of your having strong will power?

            My response to that line of argument is that when we attribute desert to someone, whether for good or bad, we are judging the person as he now exists, not the disembodied potential person that got put into a particular body with particular genes and a particular environment. John deserves to be punished because he is a nasty person who goes around hurting people. The potential pre-existence John doesn’t deserve punishment–but he isn’t the John we are going to punish. Similarly, mutatis mutandis, for the virtuous, generous, productive person.

          • Chalid says:


            Sure, I’m personally completely sympathetic to the argument that everything is luck; we are all a product of things outside our control. But one can reject that argument without then going to the opposite extreme of claiming there is *no* luck in outcomes.

            Most people would say that some fraction of outcomes is deserved and some fraction is not.

          • albatross11 says:

            Yeah, it feels like this is kind-of obvious, and is obscured more by what’s easy to argue for or what’s rhetorically satisfying than what anyone believes. Almost everyone acts in day-to-day life as though they believe that some of your life outcomes are luck/external stuff, and some are the results of your choices.

            For example, most parents lean on our kids to do their homework and stay away from drugs, because we think that we can maybe influence our kids to make certain kinds of choices, and that those choices will likely improve their lives. And also, most parents make sure our kids get vaccinated and hang around with a reasonably good peer group, because we think that external stuff like illnesses and hanging out with the wrong crowd are likely to make their lives worse.

            And it seems to me that when you think about the balance between the two, things work rather like when you think about the balance between nature and nurture–the more you restrict the range of circumstances and external stuff, the more peoples’ choices drive outcomes. So if you consider all humans ever, you’ll conclude that most of your life outcomes are driven by external stuff. Compare the life outcomes of the guy who was born and died a serf in 1600s Russia with a guy who was born into the middle class in the US and ended up as a successful schoolteacher with a wife, a nice house, and a couple kids, and choices don’t look so important relative to circumstances. Compare the life outcomes of two next-door neighbors growing up in Southern California in 1950, and it looks an awful lot like choices and good decisions leads to a substantial difference in outcomes.

            And on the gripping hand, it may be personally or socially useful to put more emphasis on the impact of your choices than the impact of external factors on your success. This Interfluidity piece discusses how that might work.

          • Le Maistre Chat says:

            @DavidFriedman: I am not a utilitarian, but sometimes social issues seem clearer in that frame than in virtue terms.
            As Samuel Johnson put it, “Decent provision for the poor is the true test of civilization.” Of the ways to achieve that end, income tax redistributing wealth from those who need it least to those who need it most is one of the most practical. It creates the injustice of the IRS’s existence, but compare it to other taxes whose end is to keep lower-income people from dying in the streets, like the regressive Social Security payroll tax…

          • Mark V Anderson says:


            To be clear I believe in progressive taxation not due to the above but mainly because of what DF refers to as the obvious utilitarian argument; I mentioned it mainly because it seemed to be a natural response to DF’s post.

            But how progressive? I want to reiterate my point that with all the different welfare schemes of the government, it is impossible to know how much welfare is going to the poor. Doesn’t it make more sense to pay all welfare out of one source, so it can be quantified and controlled and transparent, so that voters and our civil servants pay the amount they deem as correct? Progressive taxes not only complicate everyone’s tax return, but are one more thing making it more difficult to determine how much welfare goes to each person.

            Without an overall guideline of how much welfare we should give to each person, how can you determine how progressive the taxes should be? It seems like the left always wants taxes to be even more progressive, and the right often wants them to be less progressive, but how can anyone make an informed judgment without knowing all the other welfare going on in government? In practice it’s just a tribal thing, with no rationality involved. That’s why I’d like to see a flat tax, and pay welfare from a different source.

          • Le Maistre Chat says:

            It seems like the left always wants taxes to be even more progressive, and the right often wants them to be less progressive, but how can anyone make an informed judgment without knowing all the other welfare going on in government? In practice it’s just a tribal thing, with no rationality involved. That’s why I’d like to see a flat tax, and pay welfare from a different source.

            Welfare and Social Security are pretty expensive programs. I dunno what revenue stream would be more just to fund them with than income tax. The payroll tax we use now redistributes wealth from younger proles to older ones and disabled people, which seems less just than a sort of noblesse oblige tax.

          • Chalid says:

            @Mark V Anderson

            Isn’t the desire for simplicity essentially orthogonal to the discussion about the tax code’s progressivity? There’s nothing inherently complex about progressive taxes.

          • Mark V Anderson says:

            Isn’t the desire for simplicity essentially orthogonal to the discussion about the tax code’s progressivity? There’s nothing inherently complex about progressive taxes.

            Progressivity does add to complexity. Yes, when you fill out the return, you just go to a table that tells you the tax, or the computer does so. But any planning for taxes is a lot more complicated, because the tax from any particular change depends on the earner’s tax bracket, and may be more than one bracket. OF course my larger point isn’t just the extra complexity of tax returns, but the fact that it is one of dozens of programs that pay to the poor. It is most complicated from a societal point of view. It makes policy much harder to make.

            I dunno what revenue stream would be more just to fund them with than income tax.

            ??? General funds. That’s where the funds for progressive taxes come from now. Take a proportional amount from each person to fund the government. This funding includes all government spending including welfare. Assuming we keep similar funding as we do now, most of Federal funds would come from income taxes. Most of state funds would come from income taxes and sales taxes. Hopefully all this funding would be somewhat proportional to income or spending of its residents. Then welfare is paid out of these funds based on policy determined by politicians, which would be more transparent to voters than our current system of dozens of welfare systems.

    • tayfie says:

      If anyone is still reading, it seems like most commenters missed what I wanted to be the main point.

      The point I wanted to bring up is not if it is good policy on the object level. My point was *Why does there seem to be no serious support in that direction?*

      Some commenters have already pointed out the European VATs, Fairtax, and the luxury taxes of the 90s, but that seems like a lackluster intellectual landscape.

  21. Aapje says:

    What is the best way to rank countries by the number of medals they won in the Olympics?

    Currently there seem to be two common methods:
    1. Add up the medals, independent of color
    2. Go from large to small, so countries are first ranked by gold medals, then if that is a tie, by silver and if that is a tie, by bronze

    Both methods have problems. The first method ignores that gold is a greater achievement than silver and silver greater than bronze, so it overvalues the lesser medals. However, the second method seems to undervalue silver and bronze medals, by not counting them at all unless there is a tie in the better medals. Having both in use means that countries can end up way higher on one kind of ranking than on the other. For example, Russia is now 5th on rankings that use method 1 and 20th on rankings that use method 2 (they won some silver and many bronze, but no gold).

    I would argue that an intermediate ranking is better. Perhaps a good method is give each type of medal a value and then add it up like so:
    – Gold = 4
    – Silver = 2
    – Bronze = 1
    Then a country with 1 gold medal and 1 bronze medal would have earned 5 points total, placing them ahead of a country with two silver medals.

    The 4,2,1 scheme is based on the single-elimination tournament style of deciding winners, where you have 1 winner, 2 finalists and 4 semi-finalists. My preference is questionable, since many Olympic sports do not use this type of tournament and even those that do, often have an extra game between the losing semi-finalists to decide 3rd place. However, it feels like the right level of respect for each medal.

    Of course, such a scheme seems unlikely to be adopted by the media, since it requires an ability to understand elementary math.

    • Orpheus says:

      You should also probably normalize for population size. Surely we don’t expect Greenland to do as well as, say, Canada?

      • outis says:

        No, that would just be ranking a completely different thing.

      • Aapje says:


        I have considered such a ranking in the past, but it cannot work well, because there are per country caps on how many athletes/teams they can send. So larger countries are disadvantaged by this. On the other hand, team sports are generally a bit of a disadvantage to smaller countries*, because half of a great team doesn’t result in half the number of medals.

        These rules also result in calculated behavior that brings medals to countries that did relatively little to raise winning athletes. For example, the 10 km speed skating gold medal was won by a person who was raised in The Netherlands and trained there to the point where he won the Dutch National Allround Championships. However, it was difficult for him to get sent to the Olympics or other big sports events because there was so much competition in The Netherlands. His father was born, but not raised in Canada, which gave him and his son dual-citizenship. So the athlete could and did compete for Canada, which he would never have done without the caps on the number of athletes per country.

        So there is really no fair way to normalize how well countries do given the population they can draw from.

        * Although some small countries can utterly dominate a sport, if they put a lot of resources in them, so then they can also field a dominating team.

      • Wrong Species says:

        I saw a video about this problem. If you rank adjusted by population, then a small country can win one medal and countries with large populations would have to win more medals than exist in order to be “better” than them. So he uses a probabilistic formula that comes out with Norway and Australia being on top for Winter and Summer Olympics, respectively.

    • gbdub says:

      One trouble with “gold only” is that it undersells dominance by one country – having multiple athletes on the podium is more impressive than having a single superlative athlete.

      Also, gold only can oversell the impact of a single athlete who can compete in multiple events (e.g. Michael Phelps, although honestly I’m annoyed that swimming has so many medal events in the first place – if you can be good in so many disciplines (and most of the Olympians race multiple events) then they aren’t really separate disciplines).

      It’s interesting how countries take different paths though. Norway is small, but absolutely dominant in cross country skiing. There are a lot of events in that sport, and they will likely medal (or win multiple medals) in almost all of them.

      Meanwhile the USA doesn’t really dominate any winter sport, but has competitive athletes in almost all of them.

      • Aapje says:

        Focusing on a (mostly) individual sport with many medals like swimming, running, speed ice skating or cross country skiing seems like the best strategy for a small country.

        The worst return on investment seems to be team sports with just one event. You have to train and support an entire team and together they can still just win one medal.

        If a bigger country wants to win a lot of medals, it may be best to have different regions specialize in different disciplines. Having a good density and quality of infrastructure/events/etc for the sport seems far more important than to be able to select the best athlete from a large population. For example, Jamaica produces a huge number of great runners from just 3 million people. Most of the top speed skaters from The Netherlands come from one of the least populated provinces, with only 600k people.

        Of course, from an international/Olympic perspective, the US wastes enormous amounts of money and talent on sports like American football, baseball and basketball. Because baseball is so huge in the US and so focused on North America, US teams even refuse to send their best to the Olympics. So the US has mediocre results at the Olympic baseball event, despite their enormous spending on the sport.

        • Matt M says:

          Focusing on a (mostly) individual sport with many medals like swimming, running, speed ice skating or cross country skiing seems like the best strategy for a small country.

          Not sure if true – but I remember hearing during the Beijing Olympics that the Chinese ministry of sport (or whatever they call it) made a huge effort in the run up to try and maximize medal count by identifying good young athletes and pressuring them towards the “best” sports for this. They mainly tried to optimize on sports where a lot of medals were at stake, and where international popularity was considered pretty low – all of the minor, forgotten sports that never reach American TV, but still count as a gold when you list the “medal count.”

          • Odovacer says:

            I believe China also tried to focus on women’s events more as well.


          • Aapje says:

            Women’s events are easier because quite a few countries give less opportunities to women, so the competition is less.

            Also, women’s events provide better doping opportunities.

          • quaelegit says:


            I’ve been told that’s the reason for the US Women’s Soccer* Team performing so much better than the US Men’s Soccer* Team at World Cup and such… but since our main competitors are Western European countries** that seems unlikely? My other guess is that higher-profile American sports (NFL,NBA, etc) are drawing in the most “athletic/serious” people who might potentially be great soccer* players.

            *football for civilized types 😛
            ** Well also Central and South American countries, which I’d naively guess don’t emphasize womens’ sports much, but I don’t know.

          • gbdub says:

            “*football for civilized types ”

            “Soccer” is just short for “Association Football” the way “Rugby” is short for “Rugby Union Football”.

            I assume you think people who call rugby “rugby” are civilized? 😛

            But seriously I have no idea how soccer became the one true football when there are like a dozen games all called “football”.

          • Aapje says:


            Women’s soccer is a tiny, but fairly fast growing sport in The Netherlands, while men’s soccer is already huge. In 2014, there were 1.1 million male soccer players in my country vs 150k female players.

            An issue is that women in many sports and especially in soccer, have trouble attracting a large viewing audience (and in so far that they have an audience, it’s mostly men, with women not showing much solidarity), so becoming professionals is difficult.

            Male professional players from the Dutch top league earn 266k euros on average. The top female league is an amateur league, where the players merely get costs reimbursed. They don’t get a salary.

            In other countries, some female players do earn a salary, but those are typically quite low. In Sweden, female salaries are 5k euro on average. The American top female league (WSP) actually has relatively high salaries, around 22k euros on average.

            So the good performance by the US Women’s Soccer Team probably just reflects that all the national team players are professionals.


            I hope we can agree that the sport where people primarily use their hands should not be the one that can claim ‘football’ 😉

          • quaelegit says:

            @gdub — people on the internet keep telling me to call it “football”, so I obliged. Maybe the mistake is assuming they are the civilized ones 😛

            But thanks for the etymology tip! I was wondering where the word “soccer” came from…

            @Aajpe — ah, that makes even more sense! I remember when the USWT won the world cup in 2015 there were a bunch of articles on Medium and similar asking “why do we pay the men’s team more when the women’s team is the one that’s actually good?” but I don’t follow the sport so no idea if that makes sense economically or if anything happened.

          • gbdub says:

            @Aapje – I suppose if you must limit to one “football”, soccer is the “footiest”. And gridiron football is probably the least footy. But Aussie and Gaelic football are still mostly kick and catch (limited if any throwing). Soccer is not the oldest form of football (it would be most accurate to say they all share a common ancestor), so I don’t see why it should get preference.

            You should just be happy Americans call it “soccer” instead of a more accurate name like “flopball” or “snoozejog” or “who-is-more-corrupt-FIFA-or-the-Mafia-ball” 😛

          • John Schilling says:

            You should just be happy Americans call it “soccer” instead of a more accurate name like “flopball” or “snoozejog” or “who-is-more-corrupt-FIFA-or-the-Mafia-ball” 😛

            Such a pity that “moneyball” is already taken.

          • Deiseach says:


            Is that meant to be a trick question? Because I think we all know the answer 🙂

            John Schilling, I thought “moneyball” was American football, given the huge salaries that top players of a sport that realistically is only played in one country can earn?

          • Nornagest says:

            Women’s MMA seems to be doing pretty well, but I hope it’s never an Olympic sport.

          • Aapje says:


            People who want the athletes who get ‘better*’ results to get more income don’t understand how it works.

            You get prizes for winning. You get money for generating income for the people who pay your salary. These are not the same. The worst player of the worst NFL team generates more income than the best lacrosse player.

            * The comparison is also typically apples to oranges.

            @gdbub & John Schilling

            I’m very much Against modern football, so you guys can piss over it all you want 🙂

            Of course, it is still a better sport than American Ad Break Football. >:)

          • John Schilling says:

            John Schilling, I thought “moneyball” was American football,

            I would have thought “moneyball” a better match for American football myself, or maybe basketball. But the name was officially claimed for a book and a movie about some scoundrel who ruined revolutionized the game of baseball (aka American-rules cricket) by pursuing mathematically-optimized, cost-effective techniques for winning baseball games.

          • dndnrsn says:


            However, women’s MMA at its highest level is packaged with men’s MMA at its highest level. So it’s not a clear comparison to, say, women’s soccer.

          • Andrew Hunter says:

            Economically MMA is a game of stars. (People say this about basketball and football and the like, but it’s not to that extent.) Combat sports live and die on single headliners, for multiple years at a time. An event has 4-6 main card events (and another 8-10 undercard fights) but no one buys the PPV for anything but the main event, and the success of a promotion boils down to: do you have super popular champion(s) who draw focus? You don’t really control who these people are: new champions show up unpredictably, and talent doesn’t always imply marketability.

            Women’s MMA got super popular because of Ronda Rousey, who was photogenic, gave great press conference, and had a great judo pedigree, and hit the lucky combo for becoming super popular. She had a great run until people worked out her flaws. (I don’t care for her as a person, most notably because she seems deeply unaware of how unexceptional she was in everything but judo, but I’ll admit she made some great highlight reels. And then got made into a great highlight reel, which I enjoyed somewhat more. :)) Now that Ronda is gone, the new female champs are much worse as PPV draws; I enjoyed watching Cyborg fight, but she doesn’t lead to millions of buys.

            (I actually rather like women’s MMA for a different reason: the sport is younger and less deep which means you see a lot more one-dimensional specialists. Every men’s champ is a great wrestler and striker and BJJ player; they have different specialties and preferences, but they can all do everything. Women’s division still sees very interesting fights between wrestlers who can barely strike and strikers who will get their back taken easily and so on; I find that sort of thing enjoyable.

          • Aapje says:

            @Andrew Hunter

            I think that some sports/games can be ruined by the participants becoming too good. They stop making the kind of mistakes that can be exploiting majestically and instead, winning becomes about minor details and luck.

          • dndnrsn says:

            @Andrew Hunter

            While I’d disagree that the top male fighters right now are great all-around (there’s a few, even at the top level, who make up for some weakness somewhere through skill elsewhere, or tactics), broadly, you’re right. Top female fighters are more two-dimensional than one-dimensional, though – you’ll get people who are good strikers and wrestlers, but not great at BJJ, or good at BJJ and striking but can’t get a takedown – compare to top male fighters in the early 2000s.

          • Andrew Hunter says:

            @dndnrsn Yeah, I agree with your comment; I was somewhat oversimplifying for non-fans.

      • Matt M says:

        (e.g. Michael Phelps, although honestly I’m annoyed that swimming has so many medal events in the first place – if you can be good in so many disciplines (and most of the Olympians race multiple events) then they aren’t really separate disciplines).

        I feel like this makes judging Olympic performance, at the country level, almost entirely pointless and futile. It seems almost random to me how it was decided what disciplines get how many medal events, how many runs people get, etc. Like, the Lugers get four runs added together, but the skiiers get one single run – why? Figure skating gets short and long programs added together rather than judged separately – why? Snowboarders have to race each other (where crashes are common) but skiiers get individual time trials – why? Speed skating has medals for some seemingly bizarrely huge number of different distances even though a lot of the same people compete (a “team” of skaters could presumably win five medals) – but hockey only gets one medal for the entire team of 20+ – why?

        I mean sure, it’s fun, who cares, it doesn’t really matter – it’s an entertainment program masquerading as a serious athletic competition – but let’s not try to apply statistical rigor to a process so poorly designed…

        • gbdub says:

          Those are all good points. A couple quibbles – I know at least Super G has two runs, with the times added together (more like luge).

          What’s weirder to me apart from number of runs is that in the freestyle skiing/snowboard events, usually you get some number of runs and only take the best score. This seems to make the competition more exciting – competitors are more willing to take big risks to land difficult tricks, knowing that they have another more conservative run in their pocket.

          But in the speed events with multiple runs, the runs are added together, so any major mistake can put you totally out of the competition early.

        • AnonYEmous says:

          I feel like this makes judging Olympic performance, at the country level, almost entirely pointless and futile.

          yeah, but I don’t think this criticism validates your conclusion that

          it’s an entertainment program masquerading as a serious athletic competition

          After all, the athletes themselves are still competing against the best, and most of your criticisms don’t touch on that, aside from perhaps that if there’s an unnecessarily split event and different people win the golds for the different events, that the athletic skill shown might be pretty narrow in scope. But at the end of the day it’s still impressive.

          • Matt M says:

            I phrased that very poorly. What I meant there was that my own personal enjoyment of the Olympics went up a lot when I started treating as an entertainment program and spectacle rather than an athletic competition. I suppose results may vary, but that’s certainly how NBC televises it. There’s a reason they use tape delays, jump from sport to sport, only show the most relevant athletes, highlight the sports that appeal most to American sensibilities, etc.

            The way they present the Olympics is wholly different than the way they’d present any other sport. Even in the most popular sports you don’t get full coverage of the event. And when there’s far more coverage than they can possibly show, they still spend assloads of time on prepared human interest video packages, and Mike Tirico sitting in some sort of bizarre ski lodge set calmly reviewing the results you care about.

          • AnonYEmous says:

            Guess those are good points. To be honest, I don’t really watch the Olympics anyways, so it’s tough for me to comment on presentation.

          • Aapje says:

            @Matt M

            It’s not atypical for the level of competition at world championships to be higher than at the Olympics, because the latter usually has more stringent rules about the number of competitors per country, so it is less meritocratic.

            The Olympics focuses more on having a more diverse group of competitors, which is also the financially superior choice (since it increases viewership when athletes from more countries compete).

            Being more cynical makes it hard to see these competitions as being that meritocratic or the winners to truly be the best that humanity could produce, but it opens up a third dimension where you can start to understand some of the politics behind what happens.

          • gbdub says:

            NBC coverage sucks. It always has. They refuse to cover it as a sporting event and instead just create stupid narratives about whatever couple (mostly American) athletes they’ve decided to favor (or hate) this time around.

            Their excuse is that the Olympics tends to draw a lot of people who aren’t normally sports fans (a lot of women for one). But I’m not clear on why women would stop watching if they actually gave a decent live call for women’s super G.

            Best advice is to watch on their web app and focus on stuff that is not being shown live on NBC.

        • Random Poster says:

          It seems almost random to me how it was decided what disciplines get how many medal events, how many runs people get, etc.

          It actually is kinda random, although “decentralized” would be a more accurate term. Each sport is governed by an international federation (IF) which determines the disciplines of the sport and the rules within each discipline for the purpose of international competitions; the various IFs do this work independently of each other, so it shouldn’t be surprising that there’s a lot of seemingly random variation. The International Olympic Committee chooses what sports/disciplines get to be in the Olympics based on several criteria, but “number of medals per sport” is not one of them.

      • ohwhatisthis? says:

        Its simple. Perform genetic extrapolations from given known biological variables that correlate strongly with performance and think about what other factors influence performance in these events, like access to snow.

        Don’t. That takes the magic out of things. Wee~~~

    • rlms says:

      How about this? Find a linear regression model for the number of gold medals based on numbers of silvers and bronzes. Then score each country by their actual number of golds plus their predicted number. Doing this on the 2016 Summer Olympics gives these results. Highlights: New Zealand and Canada jump from 19th/20th to 13th/14th (in comparison to just using golds); Azerbaijan goes from 39th to 19th; Switzerland falls from 24th to 34th; Malaysia goes from 60th to 44th; and Norway goes 74th to 54th.

    • Mark V Anderson says:

      I would like to see the Olympics totally separated from countries. Why can’t the Olympics just be a gathering of the best athletes in the world, without trying to tie them to countries? The Olympics have become about countries instead of athletics.

      Of course that would likely be smaller spending on Olympics, but that is probably a benefit not a bug.

      • Matt M says:

        Ratings and interest would plummet. Without a “rooting interest” nobody cares, and given that most of these sports are relatively obscure, very few potential viewers would come to the games with a “favorite bobsledder” in mind or some such thing.

        • A large part of what organize sports are selling is the pleasure of partisanship. That’s why U.S. sports teams are linked to cities or universities–a preexisting body of partisans.

          I like to offer that as an explanation of why people vote, given the obvious argument against bothering to. Every four years a game is played out across the country with the fate of the world at stake, and you not only get to cheer for your side you get to play for your side, even if in a very minor role.

          • Aapje says:

            I would argue that this is one of the better ways to channel the biases of the human ‘monkey brain’ into a safe outlet (which in turn provides the emotional bonding at the national level that allows for better cooperation at that level).

            Getting rid of it seems hopelessly idealistic.

      • Nancy Lebovitz says:

        I’ve thought of a milder approach. Divide the world into equal population regions, or possibly equal GDP regions for the purpose of the Olympics.

    • ohwhatisthis? says:

      Rank how you please, really.

      Total number of golds? Ok. Total number of medals in total. Ok. Point system? Ok. Hottest athletes? Even better. Most steroids? Probably closely correlates with medals, so that’s already accounted for. Total number of medals per capita? Makes sense, somewhat.

      I think you’ve underestimated the number of people who passed elementary school.

      • Aapje says:

        I think you’ve underestimated the number of people who passed elementary school.

        I think that an large portion of the population has an aversion to math that goes beyond just adding numbers up, independent of their ability to multiply.

    • Iain says:

      The CBC ranks countries based on total number of medals, with ties broken by number of golds followed by number of silvers. This seems reasonable to me: the difference between winning a medal and not winning a medal is much larger than the difference between gold and silver.

  22. Bobby Shaftoe says:

    How’s this for a new curling topic: Doping in curling!?! To sweep harder I guess.

    When I saw the headline I got worried that somehow curling was rife with this kind of thing, but everyone interviewed in that article seems as non-plussed as I am about it.

    The real question here is what should be done about the Russians.

    • Aapje says:

      What should be done will not be done, because there are financial and political reasons to keep Russia relatively happy.

    • gbdub says:

      I mean I don’t see why they wouldn’t have doping in curling, although it’s been rare so far. Particularly the discipline in question, mixed doubles – it’s a faster paced game, and with only two people, you often have to throw and sweep on the same stone. That leads to a bigger need for endurance. Also, being a newer discipline, there is probably more leverage for players with particular skills – no one is really an expert, so some upstart with a ton of strength could maybe make waves.

      The Reuters version of that article (not paywalled) makes the common error of assuming that no physical fitness is involved in curling, although certainly it’s more like “good general fitness” rather than “freak athlete”.

      Anyway I’m guessing what really happened is that the Russian program is just so rife with doping that everyone at the gym is using, even the curlers, and he’s just the one that (so far) got caught.

    • Matt M says:

      The real question here is what should be done about the Russians.

      Well they tried “force them to wear pink and forbid them from waving their flag” and that didn’t work!

      • Evan Þ says:

        Take Brad Templeton’s idea: instead of calling them “Olympic Athletes from Russia,” call them “Olympic Athletes from Cheating Nation.” Better yet, call them “Refugee Athletes Formerly Associated with Cheating Nation.”

    • quaelegit says:

      Just speculation (I don’t know how doping works really, or anything more about the story than NPR’s headline): Could it just be the downstream effects of having such a large and state-integrated doping system? Like if doping is such a big part of professional athletics there, maybe you still see it even in circumstances that taken by themselves wouldn’t seem to make sense.

    • dndnrsn says:

      Some PEDs aid in recovery and injury healing, which helps deal with the wear and tear of practice, especially the degree of practice high-level athletes undertake. So, really, any athlete can benefit.

      • The Nybbler says:

        Got anything that repairs cartilage?

        • Aapje says:

          Not AFAIK. Many athletes take inflammation suppressors like cortisone, which seems to damage cartilage. The kind of damage that PEDs can help repair is muscle damage.

          Being a high-level athlete is typically very bad for the body and many end up with permanent damage.

        • dndnrsn says:

          Nope. Joint damage is the reason most top athletes are done by their late 30s.

  23. gbdub says:

    Continuing my curling series from the last couple OTs:

    Broomgate Part I: The History of Sweeping

    Curling has existed in some form since the early 16th century, although it wasn’t until the mid-19th that it started to resemble the current game (prior to that, the rocks were not standardized, there was no sliding delivery, no intentional turn was applied to the stones).

    For most of that period, sweeping, if done at all, used brooms basically similar to common household cleaning implements: wooden handles, corn straw bristles, swept in a side-to-side motion. I don’t know this for a fact, but I’m guessing that for early curlers, playing outside in the snow on frozen lochs, the actual cleaning effect of the sweeping was the most important part. (Side note: one version of the history of broomball has it that the players originally used worn-out curling brooms with the bristles cut down)

    It wasn’t until the late 1950s that brooms really started to change, as the brush started to become more popular than the broom, originally in Scotland, fully taking over everywhere by the 70s. This basically resembled the modern broom/brush (the terms are now used interchangeably) with a small rectangular head, held at an angle, pushed in a back-and-forth motion, like a small pushbroom or a Swiffer. The heads used either horse- or hog-hair bristles, or fabric over a foam pad (the original “Brownie Brush” was invented by a curler/upholstery supplier).

    Since then the broom has continually evolved, with the handles switching from wood to fiberglass and then carbon fiber (for weight – lighter brooms are much easier to sweep quickly). The heads switched from fixed-angle wood to plastic heads attached to the broom by a multi-directional joint allowing the angle to be infinitely adjusted to the curler’s preference. Hair-bristle brooms fell out of favor as it became obvious they were less effective, plus they have a bad habit of shedding hairs which might “pick” on the stone and send it off course.

    So that gave us basically the modern broom: straight-shafted round carbon or light fiberglass handle, 1 to 1.5” in diameter (some brands have a bit of taper), with an adjustable angle rectangular or ovoid plastic head 7-9” wide and 2-3” front to back. The pad in contact with the ice is a layer of foam padding faced with nylon fabric.

    • quaelegit says:

      Why do so many sports originate in Scotland? (Ok only golf and curling come to mind, so maybe I’m miscalibrated, but it seems like a lot…)

      • gbdub says:

        From Wikipedia:
        “Scots, and Scottish emigrants, have made several key contributions to the history of sport, with important innovations and developments in: golf, curling, football, rugby union (the invention of rugby sevens, first international, and first league system), Highland games (which have contributed to the evolution of modern athletics events), shinty (the predecessor of both ice hockey and bandy), cycling (Kirkpatrick Macmillan invented the pedal bicycle), basketball (the inventor, James Naismith, was Scottish-Canadian), and water polo (first set of rules, games and internationals).”

        I have no idea why Scotland seems to have invented so many sports.

    • gbdub says:

      So the US men pick up a surprise win over Canada to give themselves a slim chance to make a playoff. They do control their destiny: win their last two games against Great Britain and Switzerland (both are 5-3), and they are in at least a tiebreaker scenario. Up to 6 teams could theoretically end up tied for the last 3 playoff spots! Sweden is already a guaranteed qualifier. (I’m pretty sure USA comes out of the tiebreaker regardless of who else is in the tie, as they would hold 3 wins over the tied teams)

      On the women’s side, where they’ve played one fewer game, it’s also tight. Canada bounced back from their 0-3 start and is now 3-3. The home team South Korea is a surprise leader, sitting at 5-1. They have 3 games remaining, and their last two opponents have only 3 wins between them. Japan and Sweden are tied for 2nd, and the USA and the Scots sit at 4 wins in a tie for the 4th and final playoff slot.

      • gbdub says:

        The US women fall 9-6 to Korea, and will now need to win their final game against Sweden and get a little help from the field to make the playoff.

        The US men are right now in the 9th end of a game against Switzerland, in what is essentially an elimination game for both teams (Switzerland has 5 wins but will lose most 5 win tiebreaker scenarios). USA is up 7-4 with the hammer.

        UPDATE: Team USA scores one in the 9th and Switzerland concedes. In a good example of what a game of inches this is, Switzerland’s bid to steal one point came up short when their final stone, meant to freeze against a US stone that was touching the back of the button, left just enough space (literally a couple inches) for Shuster to take it out with his final shot.

      • gbdub says:

        OK, so I misread the Olympics tiebreaker rules, which can be found here

        Basically, if there are more tied teams than there are available playoff spots (e.g. if 3 teams are tied for 3rd) then they will play a maximum of one round of tiebreaker games. Tied teams are ranked after the round robin based first on wins against teams they are tied with, then their performance in the last stone draw (a draw to the button contest before each game to determine who gets hammer in the first end), then by WCF ranking.

        Tied teams will only be eliminated without playing in a tiebreaker if a single round of tiebreaker games could not determine who would advance. However, if there are an odd number of teams for the remaining spots, it is possible for teams to advance without having to play a tiebreaker.

        Examples (remember there are 4 playoff spots):
        2 teams tied for 3rd – both advance, no tiebreaker.
        2 teams tied for 4th – they play a tiebreaker, winner advances
        3 teams tied for 4th – lowest ranked team is eliminated, two remaining play a tiebreaker
        3 teams tied for 3rd – highest ranked team advances automatically, 2 remaining play a tiebreaker
        4 teams tied for 3rd – 2 tiebreaker games are played, with seeding determined by ranking. The two winners advance.

        Norway just fell to Italy. So If you’re a team USA fan, your rooting interests for the final round robin match are:
        1) USA beats GBR. This MUST happen, or USA is eliminated. But if they win, they are guaranteed at least a tiebreaker game opportunity.
        2) Canada beats Denmark
        3) Korea beats Japan
        4) Whoever in Sweden vs. Norway – this game won’t affect playoffs as Norway is already out and Sweden is guaranteed the first seed.

        These results would create a 3-way tie for 3rd place, from which I believe USA would advance automatically, as they would hold wins over both GBR and SUI. GBR and SUI would play a tiebreaker for the last spot.

        The messiest scenario is USA, JPN, and DEN all winning. That would create a 5-way tie for 3 spots at 5-4. That’s your TEAM CHAOS rooting interest. Actually, this scenario is a bit better for USA than the above. I’m pretty sure what would happen is that USA would automatically get the 2 seed for the playoff, as they would be the only team with 3 wins against the other 4 tied teams. CAN, GBR, and SUI would be 2-2 against the tied group, with JPN at 1-3. Those 4 teams would play two games with the winners advancing. However, I think this scenario is quite a bit less likely than the above – a 4 way tie for 2 spots is more likely, which would force USA to play a tiebreaker.

        If you like things simple, root for wins by GBR, CAN, and KOR. The playoff would then be SWE, CAN, GBR, and SUI with no tiebreaker game.

        • quaelegit says:

          Is “SUI” an abbreviation for Switzerland?

          And if the U.S. doesn’t make it into the playoffs, who are you rooting for? 😛

          • Protagoras says:

            Yes. Apparently because the IOC favors French or something, they use an abbreviation based on the French version of the country’s name.

          • gbdub says:

            I like Sweden because the skip Niklas Edin has a kick ass beard and Koe (the Canadian skip) seems like kind of a dick.

            On the women’s side it would be kind of neat to see the home team Koreans win, otherwise I’ve always had a soft spot for Eve Muirhead.

          • Aapje says:


            The IOC was established by a Frenchman, Pierre de Coubertin. The IOC is also based in Lausanne, Switzerland, where French is the official language.

            The proclamations that are made during the Olympics are (usually) in French, then English and then the language of the host nation (unless that is French or English).

            Fun fact: de Coubertin won the gold medal for literature at the 1912 Summer Olympics.

          • AlphaGamma says:

            On French names, the unified Korean team use COR (from the French Corée)- the South and North Korean teams are KOR and PRK respectively.

            At the 1992 Olympics, the former Soviet team competed as a “unified team” using the French-based code EUN (Équipe unifiée), while Yugoslavian and Macedonian athletes competed as “Independent Olympic Participants” with the English-based code IOP.

        • gbdub says:

          AAAGH I totally biffed the rooting interests.

          Team USA wants CAN to lose, period. A USA win and CAN loss would create a 4 way tie yes, but for 3 spots. With 4 teams for 3 spots, they would advance the first ranked team automatically. That would be USA, by virtue of being 3-0 against that group.

          So now you have 3 teams for 2 spots. CAN, SUI, and GBR are all 1-2 against that group and 1-1 against each other. So their ranking would come down to last stone draw scores. Right now SUI is winning that. So they would be the 3 seed.

          Then CAN and GBR would play each other for the last spot.

          • gbdub says:

            Team USA (the men) defeat GBR, CAN wins and JPN loses.

            USA is through to the medal round and will face CAN in the semi. GBR and SUI will first play a tiebreaker to determine who will face SWE in the other semi.

            Sadly for USA fans, the US ladies fall to IKEA and are done. They needed a win to make a tiebreaker game against Japan.

          • Iain says:

            And despite being up 3-2 with the hammer when I left for pub trivia, Homan lost to Muirhead, meaning that the Canadian women will be the first Canadian curlers not to come home with an Olympic medal. I still don’t know what went wrong there.

          • gbdub says:

            I don’t know either. They looked atrocious in their 0-3 start, but I thought they’d started to pull it together. Maybe just dug too deep a hole.

            Frankly the CAN women failing to reach the medal round is a bigger surprise that if the men had – an upset either way, but coming into the games the men’s field was considered much more evenly matched.

            The USA men really did things the hard way – they went 3-1 against the top 5 teams and 2-3 against the bottom 5. Hopefully that bodes well for them in the medal round, and they are certainly on a hot streak. They already beat Canada in the round robin, but it was super close.

  24. Le Maistre Chat says:

    If supernatural martial arts existed, how would you rationalize them? I’ll use some empirical facts from, oh, let’s use Street Fighter. 🙂
    Guile can throw a punch that breaks the sound barrier at sea level. That’s 761 MPH, or 18 times as fast as our world’s record holder Keith Liddell. Weighing in at 86 kilograms, putting all his weight into a 761 MPH punch calculates to more than 322 kilojoules. At the highest level of mixed martial arts, his most fragile opponents such as Karin Kanzuki take ten of his strongest punches to knock out.
    Forced to confront such facts, how would you explain this superhuman durability?

    • John Schilling says:

      We live in a simulated reality programmed by a comic-book geek. Fortunately, He has deigned to allow me a mouth which with I shall commence screaming.

    • anonymousskimmer says:

      You wouldn’t need to rationalize them, as the laws of biophysics would have been worked out in advance taking them into account.

      It’s really amazing that a bacterium can withstand thousands of Gs just fine (with thousands of other bacteria piled on top). Just postulate a somewhat stronger extracellular matrix (which can be made even stronger through training) and a person can withstand a supersonic punch just fine.

      Throwing the punch is something else entirely. Biologists would be looking for ATP storage vacuoles in muscle tissue (along the lines of the native fauna in:

    • ohwhatisthis? says:

      I would somehow explain it by Guile coming from a tribe that has coexisted with the mantis shrimp for thousands of years, until gradually its techniques of hunting were found by mankind.

      Seriously. The shrimp can throw the fastest punch in the world, it breaks the sound barrier, emits superheat and a flash of light, and takes both hands. Its a sonic boom.

      Now guile doesn’t throw normal punches that way. I explain that by Guile knowing his arm would break if he did, hence he only uses that ability for range techniques. I think that works.

      I explain Karin Kanzuki by her being a BAMF.

      What about Goku? Now how would you explain Goku?

    • Protagoras says:

      Yeah, I don’t think it’s possible. Damage scales seem to be weirdly compressed in super combat, so that the difference between a guy who can do a sonic boom punch and one who can do a regular punch (or between someone who can bench a battleship and someone who can bench a few hundred pounds) turns out to be very small in its effect on hand to hand combat (and if the weaker guy has more skill, that can easily make up for even such vast differences in power). There’s really no way to rationalize it. Even just saying everybody in super world is super tough doesn’t work, because that would mean that the weak (well, ordinary strength), exceptionally skilled guys wouldn’t be able to do anybody any harm (which is not how it works in these settings), and also it would mean that everybody would be nearly bulletproof (since anybody who can take multiple hits from supersonic fists isn’t going to be much bothered by much smaller, lighter supersonic bullets; also obviously not how it works in these settings).

      • beleester says:

        One way to rationalize it is to have all the heavy lifting done by magic/ki, rather than physical force. Combat consists of wearing down your opponent’s magical shield until the squishy human underneath is exposed. Perhaps there’s some sort of “only magic can kill magic” thing going on, or perhaps magic is just that much stronger than mundane weapons, but either way, your offensive and defensive power mainly depends on how much mana you put into the attack rather than the actual physical force you can generate.

        Exalted works this way – Exalts are ludicrously powerful, but if they’re out of essence, or don’t have an applicable Charm, then they’re as squishy as any mortal.

        • Le Maistre Chat says:

          My Street Fighter example might work this way: Ryu’s martial art explicitly so, and Bison learned “Psycho Power”, which just sounds like evil New Age ki. They could all be tapping into the same thing whether or not they frame it so,

        • Protagoras says:

          Possible if everyone is supposed to have the same power source; gets a bit strained if some people are supposed to be magical, some genetically engineered, some super-skilled due to extraordinary training, etc.

          • ADifferentAnonymous says:

            “Magical” people have a mutation that gives them an innate ability to access ki. Generic engineers also stumbled onto this, but they don’t call it magic. It can also be learned by exploring the limits of human capabilities through intensive practice of various disciplines.

          • Jaskologist says:

            IIRC, in the early days of research into electricity, static electricity, biolelectricity, and two other categories (might be chemical and magnetism?) were thought to all be different. Perhaps our understanding of “ki” is similar; right now we regard the different manifestations as being different forces, but in actuality it’s all tapping into the same stuff.

    • Jaskologist says:

      Also, what do we make of that fact that swords and knives don’t seem to be any more powerful than fists?

      Heck, in Smash Brothers, futuristic laser guns are often weaker than fists.

      • Protagoras says:

        Well, obviously fists just are better than swords which in turn are better than guns. Clearly technological progress has been all about reducing the risk of people getting hurt in fights.

      • Le Maistre Chat says:

        Also, what do we make of that fact that swords and knives don’t seem to be any more powerful than fists?

        I’m not sure of the physics here, but historical steel plate armor ~1mm thick nullified knives and most swords unless one found a gap. So skin impervious to even just pistol rounds would be better attacked with blunt trauma or swinging a pick, just as knights traded in blades for those hammer/pick weapons and spiked polearms.
        Punching and kicking are a weak type of blunt trauma, so they’d be at least as effective against armor skin as swords?

        EDIT: No, that’s definitely not right. If you can thrust the mass of your arm at Mach 1, that’s approximately 18 KJ, the kinetic energy of the most powerful .50 caliber machine gun rounds. You’d definitely do better against armor if the impact is concentrated into the surface area of a stiletto rather than a whole fist.

      • ADifferentAnonymous says:

        Probably points to a system where ki and mechanical force are roughly on par. Using a weapon reduces how much ki you can channel but compensates by doing more mechanical damage. And I’m pretty sure you don’t even have to leave hard SF to conclude that laser pistols just kind of suck.

  25. johan_larson says:

    One person in history has won both a Nobel Prize and an Olympic medal: Philip Noel-Baker, a British politician and diplomat. He won a silver in the 1500 metres at the 1920 Olympic games and a Nobel Peace Prize in 1959 for his work in promoting disarmament.

    And yes, you are indeed wasting your life.

    • cassander says:

      I’d feel worse if it were a real Nobel, and not the one won by woodrow wilson, the european union, and yasser arafat.

      • Aapje says:

        For the ‘hard science’ medals, the Nobel committee is increasingly afraid to give the medal to an undeserving winner and they now typically wait decades after discovery to give out the Nobel. Increasingly, a major requirement for winning a ‘hard’ Nobel is staying alive long enough.

        Yet in contrast, the Peace Prize is often handed out to the hype of the year, seemingly at least as much with the intent to influence politics than with the intent to reward achievement.

        The most absurd one was when the newly elected Obama got the prize, because the committee liked his promises.

        • cassander says:

          the way I like to put it that to win as a sitting US president, you have to screw someone really bad. Teddy Roosevelt got it for screwing the Japanese, Woodrow Wilson got it for screwing the Germans, and Obama got it for screwing John Mccain.

        • A Definite Beta Guy says:

          My favorite is the peace prize for the 1973 Paris agreement that “ended” the Vietnam War.

          It sure ended the war, alright!

    • quaelegit says:

      Reminds me of Fridtjof Nansen, who was also notable in sports and diplomacy. Also a couple Arctic expeditions. I know about him from the Nansen passport.

    • John Schilling says:

      Yeah, yeah, but what was his Erdős-Bacon-Sabbath number?

    • John Buchan was a top thriller writer–Greenmantle and The Thirty-Nine Steps are his most famous novels. Also Governor General of Canada.

      Dick Francis was a top jockey and, after retiring, another top thriller writer.

      David Ricardo was a spectacularly successful stock market speculator, one of the most important figures in the history of economics–I think the first great theorist–and a member of parliament.

    • SamChevre says:

      Roger Bannister never won an Olympic medal, but he’s a significant athlete (first sub-four-minute mile). He was also a very capable neurologist and “wrote one of the first on the autonomic nervous system”.

  26. maintain says:

    I noticed some people here try to post links to articles they’ve written on their blogs, looking for readers or feedback. They usually don’t get a ton of feedback, which I imagine this is pretty frustrating. Nobody wants to spend hours pouring their heart and soul into an article, and the only response is “Yeah, pretty good.”

    So, would anyone be interested in creating some sort of group where we review each other’s writings, give feedback, and try to improve?

    • quanta413 says:

      Could we also review things like notebooks for teaching code? I make Jupyter notebooks for teaching python workshops on programming for scientists, and I could definitely use (more) feedback. And I already read like crazy, so I don’t think I’d have much trouble reading something ~10 pages or less on a fairly timely basis.

    • ohwhatisthis? says:

      Yeah, some form of that can exist in the weekly subreddit threads. Try the friday fun thread or something like that.

      I don’t think *this* place is the one to post your blog. I mean, you can link to it in your username. Write something intelligent in a comment here and have people read it. That’s how to attract people reading it, not *ahem* spamming up a place.

      Now if you have an *absolutely* relevant research-blog article on a topic, maybe that’s acceptable.

  27. littskad says:

    What is it about the winter olympics that lends itself to goofy sounding event names? “Super-G”, “slopestyle”, “skiathlon”, “ski cross”, “skeleton”, “big air”, “curling”? Are there any events at the summer olympics that can compete with these?

    • quaelegit says:

      Well if like me you find “badminton” or “slalom” a funny word… otherwise not really.

      My guess is that the summer sports are more well known/familiar, so things that would potentially sound funny (“dressage”) have been normalized.

      Oh wait! Under the athletics tab you’ve got track and field events, some of those definitely have slightly funny names: pole vault, discus, hammer & javelin throw?

      My favorite is currently “50KM RACE WALK MEN” because I’m imagining racing music players 😛

      • gbdub says:

        some of the little graphics on your link are unintentionally hilarious. Particularly the modern pentathlon, which looks like a fencer trying to stab a horseback rider and a gat-wielding man shooting a fleeing man in the back. Or triathlon, with a runner astride a bike on the water. Or synchronized swimming, which really looks like two people shrugging off their impending death by upside down drowning.

        • quaelegit says:

          Hah, indeed!

          The one that legitimately worried me at first were the diving and synchronized swimming icons. After I thought about it for a second I realized those were reasonable representations of the event, but my gut reaction was “dead or drowing stick figures”…

          (oh I see you mentioned synch. swimming already…)

        • John Schilling says:

          At least there’s no graphic for the decathalon; you’re right that the pentathalon glyph is too busy as it is.

          Hmm, maybe either a stylized ‘Z’ or a skull-and-crossbones for the pentathalon, on the ground that the event is basically just “Be the most awesome swashbuckling action hero possible”.

          Also, for the post-modern pentathalon we replace the horses with dirt bikes, the running has to be upgraded to parkour, and the fencing becomes a round-robin MMA tournament.

      • beleester says:

        “Equestrian eventing” definitely gets my vote for silliest name. It sounds like it’s the most generic possible sport. “I’m an eventer – I do… you know, events. In my sport, things happen.

        Bonus points for verbing a noun.

        • quaelegit says:

          Oh man I meant to include this one in my original post!

          At least “equestrian” tells you it involves horses (which I hope it does?) but I also chuckled at the verbing of the noun. Also the icon for the event is weird — do they really have the horses swim?! Seems like a bad idea…

          But for vagueness, I think “Athletics” takes the cake, although if you click on it you’ll quickly realize that it’s the grouping of Track & Field events.

          • gbdub says:

            Yeah some of the event breakdowns were weird. All of track and field in one box, but Greco-Roman and freestyle wrestling get separate icons?

    • Nornagest says:

      It’s the opposite of a goofy name, but my favorite Winter Olympics event is biathlon. You’d think it’d be a truncated triathlon, right? Nope. Skiing and riflery.

    • gbdub says:

      “Dressage”, “fencing”, “Taekwondo”, “badminton”, “trampoline”, “canoe slalom”, and “skeet” are all at least as funny sounding.

  28. Random Poster says:

    Since many people here seem interested in cryptocurrencies, here’s something related: a webcomic author announces that readers can support the comic by mining Ethereum and gets a lot of both positive and negative comments; next day the author responds to the comments and gets more comments. I wasn’t aware that there are some people who hate cryptocurrencies that much.

  29. nimim.k.m. says:

    Approximately a month ago I made a decision to cease all activities here in OTs and significantly reduce participation (both reading and writing) in non-OT comments too. This has coincided with a minor but noticeable change in my mood. (I have come to view my participation here as a form of procrastination. Unfortunately I still procrastinate more or less as much as prior to starting this … let’s call it an experiment … but it’s been more emotionally satisfying kind of procrastination: on the margin, I seem to spend more time reading fiction.) Because the change has been mostly positive while other aspects in my life have remained constant (so it it’s not attributable to them), I plan to continue with the experiment.

    I’ve also been toying with the idea of a similar “vacation”, but from all internet news sources, and might as well start today. (I remember someone already did something similar before, but I don’t remember what kind of results they reported.)

    • Well... says:

      I’ve also been toying with the idea of a similar “vacation”, but from all internet news sources, and might as well start today. (I remember someone already did something similar before, but I don’t remember what kind of results they reported.)

      Might have been me, though my “vacation” from news (all news, internet and otherwise) has been permanent. I have nothing but positive results to report from that, and I highly recommend it.

  30. Nathan Leveille says:

    I would be interested in laying out my life story to someone well-oriented in relevant fields. Specifically I mean to discuss possible connections between Autism Spectrum Disorder (possibly with ADHD elements, possibly with avoidant/schiz- elements) and the combination of dextromethorphan and diphenhydramine as a remedy.

    Posting this here because I know there was a previous discussion about ketamine/mdma(?) as possible drug therapy. I just want to raise the possibility that I’m not alone in this specific issue by making the analogy in a forum where people are likely to have interest in it.

    • Thegnskald says:

      IIRC, autism symptoms decrease when you have a fever; did the drugs remedy the condition, or did they eliminate symptoms during an illness, unmasking the effects of a fever?

  31. Well... says:

    When I think of “conscientiousness” I think of cleaning up after yourself, folding your linens neatly and stacking them on the arm of your friend’s couch when you’ve spent a night on it, handing over a pair of scissors with the handles facing the person you’re handing it to…in general, trying to view your own actions through other people’s eyes with the aim of making their experience better.

    When psychologists talk about conscientiousness as one of the Big Five personality traits, it seems like they mean almost more of a plodding stubbornness. Showing up on time every day, keeping the inside of your car immaculate, working unwaveringly on a single problem for years on end, that kind of thing.

    These seem like radically different meanings. Is the word “conscientiousness” really understood that differently by psychologists, or am I interpreting the word incorrectly?

    • maintain says:


    • quaelegit says:

      I don’t understand your confusion. Your examples seem to line up with your “psychologist examples” for the most part. “Cleaning up after yourself” vs. “keeping the inside of your car immaculate”? I suppose the distinction you are trying to draw is between stuff that is explicitly done to convenience other people (“your” group) versu stuff that is done for yourself (?) (“Psychologists'” group). I had the impression that “conscientiousness” covered both cases…

      (Although I’m definitely much better at the former group, so maybe the distinction is worth drawing…?)

    • AeXeaz says:

      In the Big Five Aspects Scale, conscientiousness is broken down into orderliness (keeping things tidy, following schedules, not being bothered by messy surroundings) and industriousness (finishing plans, doesn’t waste a lot of time, doesn’t postpone decisions / plans).

      • James says:

        This is handy. Conscientiousness is probably the one of the Big Five that I find myself hardest to rate on, because I feel like in some ways (room-tidying, scheduling) I’m waaay below the mean and in others (working on long-term personal projects) I’m above it. I suspect I’m low orderliness and high industriousness. (Though even that industriousness I developed quite late—I was an atrocious slacker across the board until my mid-twenties.)

      • Well... says:

        orderliness (keeping things tidy, following schedules, not being bothered by messy surroundings)

        See my emphasis. Did you put the word “not” there by mistake? If you’re orderly, shouldn’t messy surroundings bother you?

        In any case, neither orderliness nor industriousness map well to my personal notion of conscientiousness as being about trying to make things easier for other people, rather than any particular inner drive for order or getting things done. So, I’m wondering if my definition was wrong all along or if there are in fact two definitions, where one is more of a distinct term of art operationalized in psychology.

    • Deiseach says:

      When I think of “conscientiousness” I think of cleaning up after yourself, folding your linens neatly and stacking them on the arm of your friend’s couch when you’ve spent a night on it, handing over a pair of scissors with the handles facing the person you’re handing it to

      That’s not “conscientiousness” for me, that’s “good manners, civility, consideration” – if I were tying it to any of the Big Five it would be more under “agreeableness”. If somebody handed me a scissors points-first so that I stuck myself on them, I’d be more inclined to think “he’s a jerk/she has no consideration” rather than “they’re not very conscientious!” Someone with excellent manners need not be conscientious, someone very conscientious could have bad manners.

      Conscientiousness to me has more of self-discipline and reliability and keeping your word – if you make a promise/say you’ll do something, you’ll follow through on it. You know you should start work now on that project rather than slacking off and doing it in a rush at the last minute – “don’t put off till tomorrow what you can do today”. Neatness and orderliness can be part of it, but it needn’t be “keep the inside of your car immaculate” – I’m not very conscientiousness, I admit, but I do try to be better at that at work, so I’ve always tidied up my desk, “a folder for everything”, kept the working environment orderly (home is a different matter).

  32. johan_larson says:

    Moscow’s stray dogs are evolving to fit different niches of their urban environment. Some of them have figured out how to use the subway system.

    • hyperboloid says:

      One should be very careful about using the word “evolving” in this context. Dogs take about a year to reach sexual maturity, and all in all live around ten years. The Moscow feral dog problem dates from the fall of the Soviet Union in 1991; that’s enough time for some changes in population genetics, but not a lot.

      • johan_larson says:

        The article talks about the stray dog population reaching back to the mid-1800s, which seems like an odd time to start counting. What happened at that time? A sudden fad for pet dogs? A sudden distaste for dog-meat?

        • quaelegit says:

          Mid-19th century is when the human population began growing very fast? Makes sense that feral dog population would track that.

          (Although not sure where this data comes from and exponential functions always seem to grow fast at the end…)

          Edit: also Wikipedia has no source for that data and other websites I checked just seemed to copy Wikipedia. But numbers seem reasonable given the Industrial revolution right?

        • quaelegit says:

          Actually the answer is in the article: “The evolution of Moscow’s stray dogs has been going on since at least the mid-1800s, when Russian writers first mentioned the stray dog problem in the city.”

          • johan_larson says:

            I guess the question is whether that’s when the phenomenon started or that’s when the phenomenon made it into the historical record.

  33. dndnrsn says:

    Can you link to the current ads? The last Canadian Forces ad I saw, years ago, was the “Fight with the Canadian Forces” series, which in at least one form showed a facsimile of a combat situation.

    Different militaries and branches have different ads with different themes. Compare, say, the USMC ads to the other branches’ ads in the US.

  34. Thegnskald says:

    Some numerical patterns, in, by my judgement, ascending order of difficulty to figure out:

    6 12 18 24 30 36 42 48 54

    5 15 30 50 75 105 140 180 225

    1 6 18 40 75 126 196 288 405

    9 24 42 60 75 84 84 72 45

    231 420 570 684 765 816 840 840 819

    1 12 45 112 225 396 637 960 1377

    231 380 459 480 455 396 315 224 135

    1 4 9 25 49 121 169 289 361

    13 26 39 65 91 143 169 221 247

    667 437 323 221 143 77 35 15 6

    29 46 57 85 91 121 91 85 57

    1 6 21 65 133 319 481 731 1007

    23 57 119 169 209 203 185 129 106

    10 55 385 3025 25333 220825 1978405 18080425 167731333

    10 101 1557 27971 540009 10884851 225951777 4792175531 103306557609

    • Iain says:

      I can’t tell if I’m missing something obvious, or if the difficulty ramps up very quickly and then plummets again somewhere in the middle.

      1. Zhygvcyrf bs fvk.
      2. Nqq gra, nqq svsgrra, nqq gjragl, …
      3. Cragntbany clenzvqny ahzoref? Abg fher jung gur “rnfl” cnggrea vf urer.
      4. Gur avagu ebj bs N193897? Guvf, gbb, qbrf abg frrz “rnfl”.
      5-7. ???
      8. Fdhnerf bs cevzrf.
      9. Cevzrf gvzrf guvegrra.

    • tayfie says:

      The cheap way:

      For each pattern, interpolate an nth degree polynomial where n is the length of the existing pattern.

  35. AnsisMalins says:

    Hello, Bay Area SSC people? I’ll be visiting the place for the first two weeks of April, and I’m looking for people to meet, dinners to cook, and couches to surf. I want to stay away from hotels and tourist traps as much as I can. I’m looking to experience the thing Scott means when he writes “Bay Area”.

    I can be summarized as a programmer of video games and lover of ponies. I come from Germany right now, and from Latvia originally. I try to be interested in everything. I’m 32 years old, and I’m a guy. I’ve read less than half of the sequences.

    • Nornagest says:

      I’m looking to experience the thing Scott means when he writes “Bay Area”.

      This might be a little on the blunt side, but if you’re not a rationalist celebrity, you’re probably not going to be able to have the same experience of Bay Area rationalist culture that Scott did.

      • AnsisMalins says:

        Uh, that is correct.

        But there must be something else going on for you to have made that comment. Am I being rude?

        • Nornagest says:

          This is informed by the spate of critical posts about the Bay Area community that came out around six months ago: this one is typical. A pattern has been noted where people from various corners of the rationalist diaspora visit or move to Berkeley in search of the rationalist scene that various writers have written so glowingly about, and find that it doesn’t really exist, or rather that it’s more or less limited to a single social circle whose members had all pretty much maxed out their available social connections and stopped talking to newbies sometime around 2012. So they set out to create their own rationalist culture, and end up with something that’s very different and looks more like a freakish mashup of Geek Social Fallacies type stuff with the sort of norms you’d expect when someone outside the community mentions “Berkeley”. Some of them have thrived there, but some haven’t.

          I’m not trying to be a buzzkill. Okay, I’m sort of trying to be a buzzkill. But I’ve seen a lot of people get disappointed.

          • Le Maistre Chat says:

            A pattern has been noted where people from various corners of the rationalist diaspora visit or move to Berkeley in search of the rationalist scene that various writers have written so glowingly about, and found that it doesn’t really exist, or rather that it’s more or less limited to a single social circle whose members had all pretty much maxed out their available social connections and stopped talking to newbies sometime around 2012.

            Is this that Grey Tribe I’ve heard about, which turns out to be so tiny that most people in the Bay Area have never heard of it as they go about their Blue lives?

          • Nornagest says:

            Maybe. Depends how you draw the lines.

            A lot of Scott’s friends belong to that scene, so the Gray Tribe as described in “I Can Tolerate Anything Except The Outgroup” picked up a lot of traits that are properly unique to it. But I do have a sense that there’s something to the idea of a novel, independent cultural tendency with a lot of the same traits, older than rationalism but probably no older than the mid-20th century. It’s not geek culture, exactly, but you can see it in “hacker culture” as described by ESR (another thing that people often go looking for in the wrong places, by the way), in old-school literary SF fandom, and in certain corners of the IT scene among other things. It’s poorly delineated, though, and unlike Blue and Red culture it doesn’t have a consistent politics attached. (Libertarians of various stripes are disproportionately represented, but I’ve also met basic Democrats, Eisenhower Republicans, commies, and even the odd Mormon.)

          • Le Maistre Chat says:

            Hmm, I’m skeptical because “hacker culture” and old school SF fandom + libertarianism don’t add up to a demographic large enough for capitalists to notice. Blue has Whole Foods and hybrid cars, Red has Wal-Mart and pickups…

          • Paul Zrimsek says:

            Great Cryptonomicon reference even if you are trying to be a buzzkill.

          • Nornagest says:

            @Le Maistre Chat — Oh, it’s definitely a lot smaller. And with fuzzier boundaries, like I said. But I do think it’s different enough to qualify as a culture in its own right.

            @Paul Zrimsek — Hah, I wasn’t even trying to do that. There’s just a lot of older Neal Stephenson floating around the back of my head.

          • quaelegit says:

            What’s the Cryptonomicon reference? The closes I’m coming up with is the “Ireland is a state of mind” joke/passage and I’m not sure that’s from Cryptonomicon…

          • AnsisMalins says:

            Ah. Thanks for the warning.

            How true could this be for the Bay Area in general? That all the things that made its reputation are gone and have been replaced by impostors?

          • Brad says:

            What distinguishes a culture and a subculture? You wouldn’t ever say: “red tribe, blue tribe, and goths are …” right?

          • Paul Zrimsek says:

            What’s the Cryptonomicon reference?

            Nornagest did (rather skillfully) for the second-generation Berkeley rationalist scene what Stephenson did for the second-generation Seattle music scene.

    • If we have a South Bay meetup in early April–at the moment the next one is scheduled for March 10th–you are welcome to attend. That should give you a bit of what you are looking for.