Open Thread 86.5

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.

282 Responses to Open Thread 86.5

  1. sandoratthezoo says:

    This is a picture that won some wildlife photo award: http://www.nhm.ac.uk/visit/wpy/gallery/2017/images/behaviour-mammals/5194/giant-gathering.html

    It’s of a pod of sperm whales.

    I don’t understand how everyone is is not terrified of whales. That is a frightening photo! There are these giant sea monsters that go in huge groups and whenever you’re on a boat or swimming in the deep sea, you could be like 100 feet from all these giant sea monsters and you don’t even know it because they’re underwater and thus silent to you, and invisible. That’s frightening, dammit.

    • Randy M says:

      And with that urban camo, they could be right behind you as you walk down the street and you wouldn’t notice.

    • psmith says:

      It certainly puts Nantucket whaling in a new perspective. And see also the story of the Essex–not just a whaleboat, but an entire ship.

    • Conrad Honcho says:

      Giant squids are the real nightmare fuel.

    • Nornagest says:

      Sperm whales are scary! I don’t think that picture really captures it, though; there’s no sense of scale, and it doesn’t show the rather disturbing teeth they’ve got.

      Fortunately, I don’t look like a squid, so I probably have nothing to worry about unless I try to harpoon one for some reason.

      • sandoratthezoo says:

        All whales are scary. It doesn’t have anything to do with their teeth. They’re gigantic, and yes, I understand that most of them probably don’t particularly want to eat me, but neither do hippos or cape buffalo or wild boar, and they’re all still dangerous.

        What if that sperm whale — or for that matter a humpback whale — just gets kind of pissed off that you’re up in its business? When you don’t even know that you’re up in its business, because you’re effectively blind and deaf in water. Do not tell me that a humpback whale’s lack of teeth mean that it could not kill you super dead if you were in water deep enough for it to swim around in.

        • Nornagest says:

          Oh, a humpback whale could absolutely kill me if it wanted to. So could the semi-trailer in the next lane over, or the US Government, or a herd of cows. But they probably won’t unless I give them a reason to, so I’m not particularly scared of them on that account. I find sperm whales scary not because they’re likely to kill me, but because they ooze alien lethality. They look like torpedoes as designed by David Cronenberg.

          Contrast hippos, which look like oversized plush toys but have the aggression of a bathtub full of starving hyenas.

          • sandoratthezoo says:

            I recognize that I am not 100% rational about this, and of course the percentage of people who have been killed or even harmed by whales is microscopic.

            But I also think that most people aren’t 100% rational about this, and have this sense that whales are big cuddly sea-philosophers, rather than enormous wild animals who probably think that you’re at best an ignorable nuisance and could at any moment go into “I live in a state of nature and will kill the shit out of something that’s near me” animal mode. And large sea animals can potentially get much, much, much closer to you while you are totally unable to see or hear them than can land animals.

    • Baeraad says:

      I hear you. People say, “but whales don’t eat people!” I say, “yes, I know that, but do they know that? What if just one gets it into its head to experiment with alternative food sources?”

      And anyway, the size they are? They wouldn’t even have to want to do anything to hurt me. I don’t go around plotting to eat flies, but I’ve still gotten one in my throat once or twice because I inhaled at the wrong time.

  2. Two McMillion says:

    Has anyone here read Island by Aldous Huxley? What did you think of it?

    For those who don’t know, Island is basically Huxley’s attempt to describe an ideal society after deciding that Brave New World was incomplete by itself. Essentially, he’s describing a society that gets right everything that the society in Brave New World gets wrong. From a literary standpoint, it’s not particularly well-written, but a lot of the ideas discussed seem like they’d appeal to the Less Wrong/rationalist crowd.

    • Standing in the Shadows says:

      “Island” is creepy as hell, even more so than “BNW”.

      • powerfuller says:

        Oh, I actually read this recently! I agree it’s creepy. The society is as controlling as BNW in some ways, and it feels the only difference at times is the sense I got of Huxley’s saying “Isn’t this great?” instead of “Isn’t this terrible?” The mutual adoption society thing in particular struck me as a bizarre, unworkable idea. Like most Utopias, his solutions always work for the characters, so therefore they should work in the real world; there’s never a sense that a lot of people in a society might dislike his proposed methods or find them ineffective for themselves. Also, as aforesaid it’s pretty lacking as a novel, but IMO the same is true of BWN.

  3. IrishDude says:

    An SSC shout out popped up on the most recent EconTalk podcast with Megan McArdle about internet shaming and online mobs. Megan talked about the Motte and Bailey post in reference to claims of sexism and misogyny. Then Russ Roberts mentions the Out-Group post as a great essay on not dehumanizing people who disagree with you. Both had high praise for SSC. A transcript to the podcast is in the link.

  4. Thegnskald says:

    Has anyone come across data on correlations (or lack thereof) between infant testosterone levels and parental interactions?

    (I recently came across interesting data suggesting infant testosterone levels have surprisingly low genetic correlations, based on analysis comparing twins, and given that interacting with children lowers adult testosterone levels, I am curious to know whether the reverse is true.)

  5. johan_larson says:

    A discussion in another forum gave me reason to dig up stats on police shootings, and it turns out the rates vary dramatically, even between quite similar countries.

    Fatal police shootings/law enforcement homicides per million people

    United States 2.9
    Alberta 1.4
    Canada 0.7
    Ontario 0.6
    Nfld & Lab 0.3
    Australia 0.2
    Germany 0.1
    United Kingdom 0.04
    Japan 0.0

    Homicide rates vary too:

    Number of homicides per 100,000 people per year

    United States 4.7
    Alberta 2.3
    Canada 1.6
    Ontario 1.3
    United Kingdom 1.0
    Australia 0.9
    Germany 0.8
    Nfld & Lab 0.7
    Japan 0.3

    It makes sense that these rates vary across societies. Canada isn’t Somalia. But Canada is rather a lot like Australia, and they have a whole lot less killing than we do. (source)

    • S_J says:

      You’re mixing sub-Canadian regions (Alberta, Ontario) with various national comparisons. I don’t know if that helps or hurts, but it does remind me of something.

      In this table, Manitoba Province in Canada has a homicide rate of 4.1 per 100,000 people per year. That table references the year 2012.

      The Northwest Territories and Nunavut had rates above 10 per 100,000.

      The American States of North Dakota (adjacent to Manitoba!) had a homicide rate of 2.0 per 100,000 in 2016, with variance from 1.1 to 3.0 per 100,000 over the past 20 years.

      Manitoba should try to achieve a homicide rate more like North Dakota. (Or the State of Washington, or Minnesota, or New Hampshire…)

      • S_J says:

        The data I cited above for homicide rates inside various American States comes from here:

        https://en.wikipedia.org/wiki/List_of_U.S._states_by_homicide_rate

      • Eltargrim says:

        So I was at one time Manitoban, so I can comment on some of this. Apologies for being blunt.

        The reason Manitoba’s murder rate is so high is because of its large aboriginal population, and in particular its large urban aboriginal population.

        Aboriginal people make up 66% of the victims of homicide in Manitoba, despite being only 11% of the population. A substantial number of homicides in Winnipeg (the only city of note in Manitoba [suck it Brandon]) are gang-related, and to my memory most of the gangs in Winnipeg are aboriginal.

        Aboriginal poverty is endemic in Canada, and it’s no different in Manitoba. It’s an extremely hard problem to solve, both practically and politically. However, attempting to reduce the murder rate of Manitoba to that of North Dakota without solving the issue of aboriginal poverty is a non-starter. The same goes for the territories.

        • Thegnskald says:

          This is veering close to culture war territory. Let’s not have this conversation here.

          • Eltargrim says:

            Sorry, I forgot that it was the off-week. If anyone wants to discuss what I posted above I’ll keep an eye out for it in the next thread.

    • Standing in the Shadows says:

      Those numbers and metrics are much more illuminating and actionable if, instead of looking at states and nations, one looked at the micro-geographic level (in rural areas: counties, and urban areas: neighborhoods), and also factored in demographics, ethnicity, and “participation in the drug trade”. But nobody wants to go there…

  6. keranih says:

    For reasons that don’t bear going into at this time, I’m trying to track down a particular SF work that made use of ‘gate’ portals between Earth and other worlds. By ‘SF’ I mean “not obviously to a magical land, which leaves out Thomas Covenant and the like”, by “portals” I mean repeatable transportation methods of some mystery and trepidation (think ‘Star Gate’) but not a mechanical tech like Star Trek transporters and by “between Earth” I mean not primarily space warp based (like wormholes) but on a planetary surface.

    And by “particular” I mean ‘remind me of this example that I thought of but have forgot.’ Which points at being somewhat popular but not necessarily so – it’s not Milestone’s Shadow Cabinet, nor Narnia, nor Terminator, nor the Morgaine novels by CJ Cherryh, nor Andre Norton’s Witch World. Not Portal (too tech). Very likely to be pre 2005, certainly to be pre 2012.

    Shan’t mock any suggestions, promise.

    • eyeballfrog says:

      Do you remember what medium this was? Book, TV show, movie, video game, etc?

      • keranih says:

        Not a clue. Almost certainly not video game, because that’s not my scene. More likely book or tv series, because the emotional sense of this is more long term than one-off. (yes, yes, I remember how I felt about this, but not the thing itself. Sorry.)

    • Jaskologist says:

      Hyperion had the Farcaster network, but I’m not sure if that is “too tech” to count.

    • Nornagest says:

      Hyperion was the first thing that came to mind for me, too, and Glasshouse was the second, but those are both pretty techy. And all the ones we see in Glasshouse are between habitats, not planetary surfaces, though the space aspect is not very important in the novel.

      Neil Gaiman’s Neverwhere and Stephen King’s Dark Tower books both feature portals, but may be too fantastic.

    • The Nybbler says:

      The Hell’s Gate series by David Weber and Linda Evans is on planetary (Earth-analog) surfaces. It’s about a clash of two civilizations, one of which uses magic in place of technology, the other is more technological but also includes psychic powers.

      Probably not the only one to fit the criteria by far.

    • AnonYEmous says:

      did it involve mages able to make portals ?

    • Thegnskald says:

      I think Piers Anthony had a book series that revolved around that concept, but I didn’t read it, so I am uncertain.

      There was the TV show Sliders.

      Can you recall any other information? Quantum Leap almost sort of fits.

    • dodrian says:

      Robert Charles Wilson’s Spin trilogy (2005) used this in a limited way (there was only one on Earth, and it only connected to one other planet).

    • Urstoff says:

      Pretty sure it was Timecop

    • Well... says:

      Anathem (sci-fi novel) by Neal Stephenson. TL;DR: people traveled between dimensions by groaning.

    • Conrad Honcho says:

      Peter F. Hamilton’s Commonwealth saga?

      Note that while yes, the initial form of the wormholes is Stargate-ish, eventually Ozzie’s just walkin’ the path, man.

      Even if that’s not it, if you haven’t read it, you should. Good books.

    • Standing in the Shadows says:

      Was it the one where the first gate was opened by some grad students to Mars, and they beat the multihundredbilliondollar manned mission by a few hours, stealing the thunder from the guy who was picked to be the “First Man on Mars”?

      Then later when the gates were in intersteller production use, people and goods traveled between planets via train.

      • sandoratthezoo says:

        Er, sorry, thought you were replying to the post above you.

        The “beats the spaceship to Mars and then later you take a train to another planet” one is Hamilton’s Commonwealth series (Pandora’s Star and sequels).

    • Charles F says:

      I would recommend cross-posting to the sci-fi/fantasy stack exchange. They’re generally pretty good at this. This is a little light on details, so you might try looking at this to jog your memory first.

    • herbert herberson says:

      Fredrick Pohl’s Gateway? (probably not, since those are pre-programed spaceships rather than portals, but it certainly hits the mystery and trepidation marks)

    • Nick says:

      Have you tried looking through some tvtropes compilations?

    • littskad says:

      Larry Niven’s Known Space books (Ringworld, etc.) had the “displacement booths”. Alfred Bester’s The Stars My Destination had “jaunting”. Robert Heinlein’s Tunnel in the Sky had “Ramsbotham jumping”.

  7. Argos says:

    Since it seems that the ban on topics related to culture war leads to a real dearth of posts, let’s discuss another topic close to the heart of the commentariat: social ineptitude!

    Firstly, I find it bewildering just how bad mainstream advice is when it comes to skills like holding conversations and making acquaintances (although it is a little bit better than for dating). The by far most frequent piece is “Ask questions about your interlocutor, everybody likes talking about themselves”. Following this advice literally produces very dissatisfying results. For example, when getting to know other college students at student events, I thought that “What made you choose your major” would be an interesting question, because it allows people to talk about themselves and something they enjoy, however most people just felt visibly weirded out and gave short nondescript answers. In addition, as a friend of mine remarked after eavesdropping on one of my conversational attempts, it produces very unnatural, interview like conversations.

    Another gripe is the recommendation to small talk about FORD topics (Family, Occupation, Recreation, Dreams), which just makes me wonder if that is an “Emperor’s new clothes” phenomenon. How does nobody laugh at the idea that two minutes into a conversation I should start asking about a stranger’s dreams? Is this one of those things, where there are obviously implied and (by the majority of people) understood other rules? Like for example how nerdy people misinterpret the advice of being polite to women (“Well, yes you should be polite, but obviously don’t address your dates as m’lady!”)?

    Anyways, to be a little constructive, I did once come across a very neat and logical conversational technique for getting to know people at parties, on the street or networking events. As an aside, I found it in one of the unlikeliest places: It was one of those typical scammy self help websites where there is just one really long landing page with both the outlandish claims like “This course will make you the most sociable guy at any party”, “Get rid of your social anxiety forever”. However, the author had given out some free coupons on reddit, so I gave it a try.

    First, that guy explained that “just asking questions” does not work because you need a certain amount of situational and conversational context for asking a question, so the other person is not left wondering “why is this person asking me this”. And the process for creating this right context are basically these four steps:
    – The introduction step where you just ask a boring question, like “how do you know the host”, “Is this the first time you are at this event”. This works better in situations where it’s normal that strangers start conversations with each other.
    – Now the second step, which is the neat one, involves the “zooming out” technique. Basically, you zoom the conversation “out” by gradually and smoothly skipping to topics that are n+1 for the current topic. For example, it works on a geographical basis: You can ask the stranger at the party which part of the city they live in, talk a bit about that area and then ask where they are originally from. When you have exhausted that topic, you could start talking on the country level, by discussing recent travels or something. Going straight to “which countries have you been to?” would have been weird, however. This zooming out also works well on a thematic level. It’s important to make sure that you also say something about the topics and about yourself. You can make the switch to topics more smooth by offering an opinion, or talking about something you recently read or heard.
    – The third step is to ask for advice after a bit of zooming out. Once you have established a bit of a connection, you can ask for help in an area your interlocutor is knowledgeable about. If they mentioned that they can speak Italian, say that you always wanted to learn Italian, and how you could get started and what the difficulties are. It’s best that you offer a reason why you are interested (always wanted to learn it) because this way the other person feels like they are really helping you. Ideally, you should ask advice about something you are generally interested in.
    -The best case scenario is if you can help them with any problem they might have, but this may obviously not always be possible.

    While this still requires some basic social skills, like topics which are to be avoided and the ability to say something about a topic that is at least somewhat interesting to the other person, I like that it enables you to steer the conversation towards a topic that you do enjoy talking or learning about, thereby making it a more pleasant experience for both people involved.

  8. Wrong Species says:

    Let’s say you tell me something. I mishear you and think you’re coming up with an idea when it wasn’t. It turns out to be brilliant. Is it my idea? Should I get credit for it?

    • I actually have one article I’m fond of that started as my misunderstanding of someone else’s article.

    • rlms says:

      It depends, but I think the mishearer should usually get most of the credit.

    • James says:

      I get this quite often with i) jokes and ii) song lyrics. I’m inclined to say the mishearer gets what credit there is. One possible metric: would the same mishearing spark the same idea in other people’s brains? If not, it’s probably some unique or unusual combination of ideas present in the hearer’s brain, so give them the credit.

    • John Schilling says:

      I mishear you and think you’re coming up with an idea when it wasn’t. It turns out to be brilliant. Is it my idea? Should I get credit for it?

      Who cares? The proper credit for an idea is almost always properly described by the phrase, “two cents’ worth”. See also, inspiration/perspiration percentage asymmetry.

      The proper question is, should you get credit for the implementation? That one usually isn’t too difficult to figure out.

  9. johan_larson says:

    What was the best decade of the twentieth century?

    Counting the years 1901-1910 as the first decade and 1991-2000 as the tenth decade, we can rule out the second decade (1911-1920) for WWII, and the fifth (1941-1950) for WWII. But that still leaves eight more (or seven, if we let WWII ruin the fourth decade too.)

    In my field, computing, the eight decade is remembered as a particularly fertile time. It gave us C, the Xerox Alto, VAX-11/780, Apple II, and Ethernet. The IBM PC was a little later in ’81, and structured programming, UNIX, the PDP-11, and relational algebra came a bit earlier.

    I’m thinking the tenth decade may take it. The Cold War was over and we had the internet boom.

      • Vamair says:

        For the US it’s probably the ninth decade (1990-2000). For the Soviet bloc it’s probably 1960-1980 (00-s were bad, 10-s were Civil War, 20-30-s were the height of repressions, 40-s were the worst, 50-s were the post-war, 80-s were when everything started to go down, but a bit more peaceful. 90-s were awful). I don’t know that much about other countries to say anything.

    • bean says:

      I think the Great Depression ruins the fourth decade on its own.
      Personally, I’m split between the seventh and the ninth. The seventh saw us going to the Moon. The ninth saw the defeat of the Soviet Union without the need to have a hot war. Also, lots of cool military changes. And the last return of the battleship. (OK, that’s probably taking my biases a bit too far. But I could have said the first or second, for the fleets built then.)

    • Well... says:

      There was something (well, a lot) about the 90s (i.e. the 10th decade) I really liked, but I don’t feel anywhere near confident enough to say it even might have been the best decade. (I’m not from Bosnia or Rwanda, for example…)

      Maybe I could say it was the best for popular culture?

    • SamChevre says:

      I’ll go for the first decade: very little war worldwide, rapidly rising standards of living. World War 1 was an incredible disaster, and the world has not yet recovered in my opinion.

    • Paul Brinkley says:

      I’m trying to figure out how one would be systematic about this. Having no system as yet, I shall brainstorm:

      D01 had Teddy, John Muir, national parks, E=mc2, airplanes, and the Model T; but it also had Taft and Leon Czolgosz.

      D02 had WWI, Wilson, Leninism, the Titanic, and Prohibition. (Might be contender for worst.) To be fair, it also had general relativity, feature films, and jazz.

      D03 had Coolidge, electric refrigerators, and Roaring. But it also had the Great Depression.

      D04 had WWII.

      D05 had WWII in the Pacific, with the bomb. But it also had Los Alamos, ENIAC, the transistor, cryptography, the microwave, and television.

      D06 had rock and roll and Adenauer. But it had Stalin (though not for long).

      D07 had drugs, the Vietnam War, and Castro. But it also had the Beatles, MLK, and the Moon.

      D08 had more drugs, and disco. And Watergate and Iran. But it also had Star Wars.

      D09 had the Atari 2600, Stallone, Schwarzenegger, Jackson, and Madonna. But it had Iran-Contra.

      D10 had the fall of the Wall, and then the USSR. And then the Web. But it had the Starr Report, and Titanic.

      Any patterns?

      • Well... says:

        D02 had prohibitions! 1914 was the Harrison Act, 1919 was the Volstead Act. (Marijuana Tax Act was D04.)

        BTW if prohibition is a mark against D02, then civil asset forfeiture and creation of the position of Drug Czar ought to be a mark against D08 or 09. Or you could just write “Joe Biden”.

        If you’re using Jackson or Madonna to prop up D09, you could at the very least use Rogers & Astaire to prop up D04.

    • Wrong Species says:

      The 90’s has massive advantages that the other decades, to varying extents, don’t have if you look at it from the Rawlsian Veil of Ignorance.

      The world was much more equal in the 90’s compared to other decades. You would have it to restrict your options to first world countries to avoid the overhwhelming probability of absolute poverty.

      Even then, there’s still the question of race and sex. Most minorities and women would say that whatever advantages earlier decades had, it doesn’t make up for the lack of rights. And even with the advent of legal rights, culture changes at a gradual pace. So to make a fair fight, you have to avoid the possibility of being anything but a white male.

      One disadvantage that the latter 20th century had compared to the earlier was the existence of nuclear weapons. For over 40 years, the average person had the knowledge that they and most of the people they know could be killed in an instance. But with the end of the Cold War, it becomes much less of an issue. So the 90’s were the only decade in the latter half of the 20th century that didn’t have to worry about it.

      And finally, wealth and technology was just much more advanced. Even as a white American male who didn’t think much of nuclear war, you are so much richer with vastly more possible uses of tech the later you go.

      Knowing all that, how could any other decade compare?

      • Conrad Honcho says:

        the average person had the knowledge that they and most of the people they know could be killed in an instance

        Well, of course, but you just have to run back to the entrance portal from the nearest graveyard.

    • John Schilling says:

      The seventh decade had the Apollo program and everything leading up to it, it had enough computers to keep the nerds happy, it had enough in the way of feminism, civil rights, free love, and awesome drugs to fuel the belief that the Dark Ages were finally going to be swept away in one generation of enlightened activism, and it had the distinct advantage of not knowing how badly most of that would turn out.

      Well, OK, it also had the worst parts of Vietnam. Still, it’s looking pretty good to me.

      • johan_larson says:

        Was the seventh decade the best of times for aeronautical engineers, or was the fifth decade better?

        • John Schilling says:

          I might actually go with the fourth. Even if you’re only interested in warbirds, that’s when Kelly Johnson, Ed Heinemann, Sydney Camm, R. J. Mitchell, Barnes Wallace, Willy Messerschmitt, Kurt Tank, Ernst Heinkel, Alexander Kartveli, Jiro Horikoshi, Andrei Tupolev, Mikoyan and Gurevich all designed the airplanes that made them famous. You also had the chance to basically invent the airliner from scratch, twice – first in romantic form with the great flying boats and then with the more utilitarian Electra and DC-3, you had unlimited air racing as a sport with broad public interest, you had Wacos and Stearmans and so forth for fun, and if airplanes generally weren’t your thing you still had the last great zeppelins and the first helicopters. And people had stopped laughing at Robert Goddard, so he and Von Braun, Oberth, Korolev, Glushko and the rest could get on with building spaceships for the future.

          Wait, wasn’t there a Great Depression going on then? Something doesn’t fit here.

        • bean says:

          I’m really tempted to go with the sixth, actually, although John makes a good case for the fourth. You had all sorts of stuff going on. ICBMs. Spaceflight. Jet airliners. The U-2. A dizzying profusion of really interesting military jets.

    • cassander says:

      It’s hard to beat the 90s. The 20s might have it, but prohibition casts a shadow.

      • John Schilling says:

        Ineffective prohibition just adds to the fun, at least for certain definitions of fun.

        • johan_larson says:

          Jail-time in the early twentieth century is not for the faint of heart.

        • cassander says:

          Eh, I’ve bought weed, it wasn’t particularly fun, mostly just annoying. And even if getting it was fun, the quality level fell of precipitously and the expense rose considerably.

        • Paul Zrimsek says:

          I suspect the definition of fun being alluded to here is something like “spraying rival suppliers with machine-gun fire”.

          • John Schilling says:

            Or driving fast cars down winding country roads at midnight with no lights, or hanging out in a speakeasy with all the cool people, or spending all the bribe money you get on top of your meager policeman’s salary. It’s all good clean fun, from a certain point of view.

    • Baeraad says:

      I want to say the 90s, but that might be because I grew up during them and they shaped my idea for How Things Are Meant To Be.

      But I’ll still bite the bullet and go with that answer. If nothing else, you had the coolest technology and the most social equality. And the best tabletop roleplaying games, which might not matter to most people but which I personally happen to feel strongly about. Oh, and the world was neither engaged in, recovering from, or threatening to at any moment fall into a world war, which must have made for a nice change of pace.

  10. Noobish question about WordPress: How can I receive comment notifications that are 1) not through e-mail and 2) only for replies to my comment? I can live without (1), so I checked “Notify me of follow-up comments by email.” But that gave me notifications about all comments on the post, failing (2), which I’m far less willing to concede.

    • CatCube says:

      WordPress can’t do it, so far as I know, but @the verbiage ecstatic did a notification widget that you can sign up for here: https://slatestarcodex.com/2017/05/25/open-thread-76-25/#comment-504989

      Break

      Scott, can the commenting widgets be added to the comments page? This one and the autocollapsing threads one from Bakkot here are really useful, and I have to track them down every time I need to reactivate them for some reason (cookies on a new browser, for example). That second one took like a damn half hour to find this time, and ended up with me looking for the open thread where you talked about it to link me back to the open thread where Bakkot posted it.

      A word of warning to anybody using that second one: Bakkot’s link is to a regular https://slatestarcodex.com link, and that will not activate the widget on https://slatestarcodex.com; so if you use the second one, you’ll have to manually add the “s” to turn it on during your normal browsing experience. (Same thing as the new comment highlighting, really.)

  11. Robert Liguori says:

    So, two random sets of comments on two of my hobbies: tabletop gaming and Minecraft.

    A new chunk of playtest for Eclipse Phase v2 just came out recently. This game…ah, this game. On one hand, I want to recommend it to the people here, since the main themes of the game are about transhumanism and deal with lots of interesting questions which people like to wrestle with here.

    On the other two hands, the rules are still very poor, and the game doesn’t really understand the questions it’s asking. The rules are ridiculously fiddly when it comes to things like gear and body modifications. There are huge gains to be had by carefully examining the gear lists and picking just the right combination of equipment, which seems at odds with the stated intent of play (to be ready to be fast and loose with everything you own, including your body.)

    But worse than the fiddly bits is the way the game falls completely flat on its face when you engage with its premise even a little bit. The game features, among many other things, make-anything-nanotech-assemblers (including more nanotech assemblers), full brain-reading, and easy uploading, downloading, and copying of consciousness, and even some editing (although that and reintegration of copied minds have notable limits.)

    One of the game’s taglines is “Your mind is software; hack it.” But the game has huge chunks of the system which fall over when you even try to do that. Take that whole mind reintegration thing. Eclipse Phase is a horror game. There are rules for taking mental HP damage, getting traumatized, and going insane. Messing up on a mental editing check or waiting too long to reintegrate with one of your mental copies causes mental stress.

    But your brain is software. You can just wipe the copy of the mind that you botched and try again. You can copy the mind dozens or hundreds of times over and make the attempt dozens and hundreds of times. If the copies of you are too similar, you can copy yourself dozens or hundreds of times, put all of yourselves through enough simulated time and randomized simulated experiences so that you diverge, then have all of your divergent copies attempt, then wipe all of the copies and all but the best result of the psychosurgery. And all this takes is a little time, a little bit of computing space, and the willingness to actually engage with the setting’s mechanics.

    Does anyone else here enjoy tabletop games, and have any of you looked at Eclipse Phase, old or new? If so, what do you think?

    The other thing I’ve been doing lately is attempting to build my own Minecraft mod pack for 1.12.2. I’ve run into something which feels like there should be a solution out there, but if there is, I haven’t found it. Most Minecraft mods had dependencies; many mod writers put their core functionality into an API mod, then have that be a dependency for all of their main mods. Some mods use other entire mods as dependencies; for example, the Botania mod depends on Quark, which in turn depends on AutoRegLib.

    There’s a handy website called Curseforge, which hosts mods and tracks dependencies. What I want is to be able to put together a list of mods, hit a button, and get the latest 1.12.2 version of each mod and each mod’s dependencies. I’ve come most of the way to writing a utility in Python which scrapes the curseforge pages and does the downloading, but I’ve held off on building the recursive dependency tree builder, because I really hate dealing with recursion.

    Anyone here a fan of modded Minecraft? If so, is this a solved problem where I’m just not aware of the solution?

    • toastengineer says:

      (Epistemic status: haven’t touched this stuff seriously in like six years… but I checked in a few months ago and very little seems to have changed)

      The Technic Launcher auto-installs modpacks but I don’t think it works like the apt-like solution you’re asking for; it doesn’t do dependency management and needs lots of provided configuration.

      Minecraft mods don’t… really work that way, installing any more than a few mods usually requires a lot of really fiddly configuration and testing for conflicts. Block and item ID conflicts alone mean you can’t just drop things in to the Forge mod folder and expect it to work with more than like three mods.

      There’s a reason the pre-built packs became so popular so fast and why the wars over them became so white-hot. Unfortunately the lady with the skills to fix all these problems got toxicity-ed out of the Minecraft community and went off to build her own thing from scratch. As far as I know they still haven’t fully recovered from Eloraam leaving, she’s the reason you don’t still have to un- and re-zip your Minecraft JAR all the time. 😛

      Why do you want to build your own from scratch anyway?

      but I’ve held off on building the recursive dependency tree builder, because I really hate dealing with recursion.

      Enh? Why?

      • Robert Liguori says:

        Oh, I know I’ll have a bunch of secondary work handling conflicts, tweaking recipe overlaps, and doing a bunch of other stuff, but being able to get a current picture of the mods I’m using seems like a starting point I want, especially when I’m using mods which update semi-frequently.

        Why do you want to build your own from scratch anyway?

        I keep getting into modpacks and finding that they’re super-bloated with resources I don’t use, or that the designer left in a recipe conflict which renders an item I really need uncraftable, or a block I really like from one modpack ends up crashing the world because the designer never tested it and didn’t even think to exclude it.

        As for the recursion, well, I just don’t like dealing with it. Based on my performance with it in college and out, dealing with non-trivial recursive solutions requires of me either a great deal of focus and concentration which makes programming Not Fun, or “Fuck it, I’ll get drunk and throw code at the screen until I see what sticks.”

        Hell, this specific problem sounds like something that I don’t even need a full recursive solution for. I think I can just read my list of mods, and for each item in the list, check if its direct dependencies are in the list, and if not, add them to the bottom. But since I’ve only got a few mods in my modpack at present (since I only recently started playing with 1.12 from 1.10.2), I don’t yet need the full rebuild functionality.

      • Nick says:

        Did Eloraam ever get much progress on that other project she was working on? I don’t think I ever saw anything beyond a few pictures on her redpower blog.

    • Nornagest says:

      I haven’t seen the v2 rules, but v1 Eclipse Phase was a frustrating game. Fantastic premise, and some parts of the rules show a lot of promise, but they’re full of weird, awkward issues that have to be ignored or houseruled around. Kinda reminds me of D&D 2E that way, except that this is supposed to be semi-serious SF so it’s a lot harder to fall back on “it’s just a stupid game, we should really just relax”.

      I also think I’d have been happier if it figured out whether it wanted to be Altered Carbon or early Charles Stross and stuck with one. I don’t care which, they’re both good. But their approaches to the whole transhuman issue are incompatible in subtle ways, which tend to get a lot less subtle once you let a half-dozen half-drunk, half-psychotic players gnaw at them for a couple of hours.

      To be fair, though, I expect there are maybe a hundred people in the world who’re capable of writing good transhumanist SF, and most of them are in fields less fickle than RPG publishing. And I can’t condemn any game too much if it lets me play as an uplifted octopus.

      • Robert Liguori says:

        In fairness, you can do a lot with weird, awkward rules. Hell, let me talk about another game with great promise and massive mechanical issue that I happen to really like; Exalted 2E.

        Exalted 2E has you playing a mythic hero in the vein of Hercules. The world itself is fairly grim and gritty. The source book has extensive sections of rules for bleeding out from wounds in combat, or getting a lethal disease and dying before you ever make it to the battlefield, and so on, and at the end of nearly all of these sections are the rules for why this is never a problem for Exalts. The rules are weird and fiddly, but they’re there not to be used; they’re there to express, in the language of the world’s physics, how different you actually are from the common man.

        There, the rules were written with a coherent vision of not only a world, but how each mechanic should support and reinforce that world. EP was not.

        The thing is, though, there are so many things you can do with just small tweaks to make the setting really interesting. Like, one thing I want to throw out in a game is an aggressively self-forking sophont who has been programmed to get a deep and abiding sense of satisfaction out of grunt work. The personality farcasts to remote asteroid bases, cheerfully sleeves into the clankiest cases available, and immediately charges out to use its very specifically specialized skill-set to run nanocable and change nano-oil, then uses all of its profits to upgrade its tools and buy more super-clanky cases, and so on.

        What happens to the actual clanking masses when one personality has single-handedly made them uneconomical to deploy? What does it mean for the lower classes in general when automation not only makes them redundant, but the platform the automation runs best on is their own bodies?

        Eclipse Phase shies away hard from the questions it claims to raise. Being uninterested in toppling the status quo is fine; there are loads of stories where dragon-knights clash in the skies and no one asks too many questions about their impact on mercantile shipping and cross-cultural exchange across geographic barriers. But when you set yourself up as sci-fi, and act like you’re asking questions like “What does it really mean to be human?”, being completely unable to engage with potential answers is a glaring, unoverlookable weakness.

        • Nornagest says:

          See, I wouldn’t call that awkward in the context of Exalted. Weird, maybe, but there’s a thematic reason for those rules to be there, it’s pretty easy to figure out what that reason is, and it doesn’t normally get in the way of gameplay. When it does, it’s because you’ve got a squishy mortal hanging out with your pack of Exalts for some reason, and then we’re right back to thematics.

          What I would call awkward in the context of Exalted is how combat between high-level Exalts turns into a boring attritional slog until someone runs out of Essence, which is the opposite of what you want in a game about mythic heroes punching gods in the face.

          • sandoratthezoo says:

            Roleplaying games have a general problem with non-attritional combat.

          • dndnrsn says:

            The opposite of non-attritional combat can be bad too – games where one shot can kill often end up with a situation where PC deaths are more commonly “eventually a mook gets lucky” than “PC is killed by worthy opponent.”

          • Nornagest says:

            Granted. But Exalted’s particularly bad — where building a combat-effective character in D&D (older versions, at least, I haven’t played 5E and barely played 4E) is mostly about striking a good balance between giving yourself more tactical options and maximizing your ability to damage or incapacitate enemies, building one in Exalted is all about Essence efficiency. There are so many unbeatable defenses available that you’re actively disincentivized from using your good attacks until they’re gone, and then the fight’s pretty much over anyway.

            That’s not to say there aren’t things I like about the game — I’m a big fan of stunt rules, for example, and Limit and its friends make a good marriage of mechanics and theme.

          • sandoratthezoo says:

            I mean, Exalted 2e is legendarily bad at combat mechanics. I think that the overall view of it is that it’s incredibly complex, that there’s only one viable approach to combat (Combo a perfect defense and a flurry attack every action), that tracking NPCs is an extraordinary pain in the ass, etc.

            But even take all that away, and you’ll be in the case that the only viable approach RPGs have ever found to combat is attrition, because all the other approaches result in killing PCs too often.

          • Robert Liguori says:

            For those who aren’t in on Exalted 2E: one of the big concepts of Exalted combat is the idea of Perfect defenses. They’re maneuvers which, when used, literally trump every attack, including attacks which say they trump everything, and explicitly including attacks which say they trump perfect defenses. (One of the meta-principles of the game’s rules is that in a fight between an irresistible force and an immovable object, the object, or defender, wins.)

            The problem is that it’s also really easy to carry a big fuck-off sword, get stupidly accurate with it, and deliver, with no further cost, deadly strikes. As such, the optimal form of Exalted combat is basically getting to the point where you can attack with a deadly enough attack to force your opponent to spend resources on a Perfect defense without costing yourself any energy points, reserving your own entire point pool for your own Perfect defenses, and dealing with anything less optimized than you by running it out of magic points by attrition and nuking it thereafter.

            It is, as mentioned, not an ideal system for what Exalted wants to emulate.

            But it’s actually worse; because this is an emergent property of the rules and the charmset, the 2E book encourages you to buy basically useless defensive charms. If you happen to stumble into this tactic, or a close cousin, you will be wildly more effective than a combatant who doesn’t. And due to the way lethality works in Exalted, there is almost no margin between “Untouched because I Perfected the last attack.” and “Dead because I didn’t Perfect that last attack.” once the dice pools get to even a moderate size.

            As for me, I actually like what the rules were trying to convey, I just found them a bit hinkey if you actually tried to apply them and make them notable features of the world.

          • sandoratthezoo says:

            One notes that Exalted 3e did away with Perfect Defenses.

          • Unsaintly says:

            Exalted 3e has its own set of messy issues, but is a significant improvement over 2e, especially when it comes to combat. In terms of mechanics, Exalted 3e’s withering/decisive system is my favorite way to depict heroic combat.

            In short: it’s Dissidia
            In longer: Most attacks don’t damage Health directly, they instead steal Initiative, with a big bonus for reducing an enemy to 0 or lower. Then you can cash in on your initiative to deliver strikes to the enemy’s Health. Also, perfect defenses – the few that still exist – are very much not spammable and are instead desperation tools. This results in a combat the feels like the sort of back-and-forth maneuvering that heroic duels are supposed to be, while still offering plenty of chances to come back/escape once you start losing. In fact, in most battles the players won’t even lose health at all, meaning that they don’t suffer much long-term cost for engaging with the system.

    • Said Achmiz says:

      Based on my experience with Eclipse Phase, my view is basically the same as Nornagest’s and yours. (Background: I have been playing D&D and similar for many years.)

    • sandoratthezoo says:

      Eclipse Phase! Man, I’ve had fun with that setting.

      Not that game, mind you. The game is… well, it’s probably fine, but my group has no patience for that kind of fiddly long-term thing. But I ran it systemless a bunch of times and have spent a lot of time trying to grind the setting down into something that makes sense.

      So the first thing I’d suggest you do with Eclipse Phase is, regardless of what system you use, take morph ownership away from the players. Put ’em in the morphs that you want to put them in, not the ones they want — especially not the ones that they lovingly hand-craft. So first of all, that prevents them from spending time obsessing over the gear list and finding ways to create overpowered badasses. Second of all, it’s what the game is actually about: your body is gear, it’s not something you invest in.

      Most of my games have been one-shots in which I couldn’t assume player familiarity with the setting, so I’ve come up with a lot of excuses to have the players be people from outside the setting reanimated in random bodies. Here are a few types of minds that you might use:

      * People from considerably in the past who got their minds recorded (perhaps experimentally, long before there was a way to reinstantiate them!) and got mixed in with other infugees from the fall of Earth.

      * People from the Jovian Republic who grew up with restricted technology and restricted access to information about the solar system as a whole.

      * People from weird creepy cults either out in the Pandora Gate network or in isolated habitats, insulated from transhumanity.

      * AIs created to fill out virtual characters in some kind of fantasy roleplaying game, probably without knowing that they are AIs.

      And then here are some excuses why some of the random minds above might get plunked into weird bodies:

      * They are collateral damage from some kind of Firewall containment play, and someone feels bad and reinstantiates them.

      * They were found on an iPod (-equivalent) in somebody’s closet after decades of neglect and someone hastily shunts them into bodies to try to jolly them around like, “What? No. You weren’t incorrectly filed and ignored and not instantiated for years. Look at this awesome (pod) body!”

      * They were rescued from a cult/VR fantasy scenario, their captor’s finances were drained, and they were given bodies hastily locally purchased using their captor’s finances in reparations.

      * They won a lottery and get given a cool body and then immediately someone tries to con them into indentured servitude.

      * Some semi-trapped TITAN-descended AI is making some kind of long play and they’re part of its scheme.

      Okay, so, economy and making the world basically make sense:

      1. Assume that nanofabs are wildly energy-inefficient and that it’s not actually possible to support the great masses of people living in inner-system traditional economies with just “nano-fabs” as your answer to everything.

      2. Inner-system traditional fabrication is more materials-efficient, but not as labor-efficient. Basically, the inner system’s “hypercapitalism” are the blue-collar workers of the solar system.

      3. Outer-system autonomists are actually dependent on the labor/material goods of the inner system, and pay for it by being the specialists in software, nano-design, and other high-tech disciplines. Extropians and others essentially broker between the two economic systems, providing material inputs for the autonomists and intellectual property inputs for the hypercapitalists.

      4. You can’t nano-fab a living body because nano-fabbing is too slow, the body while die while it’s half-completed. So getting a living body is still a result of growing one from an embryo, which is why biomorphs are an aspirational good for most people.

      5. Synth morphs are fundamentally unpleasant for most people to sleeve long term.

  12. Deiseach says:

    Still alive and well and not relocated to the North Pole after Hurricane Tropical Storm Ex-Hurricane Feck It The Big Wind Ophelia!

    Got off fairly lightly down here, not as apocalyptic as the forecasts made out – very windy, the worst were the gusts of wind which were severe, and some rain but honestly not as much as I expected. More like a November winter storm than the kind of disaster well-meant American advice about preparing for a hurricane on social media had led me to expect. Power was out mid-day for about three hours but the national power company did very well to get it back (round here, anyway; there are still outages all over the country).

    Though nationwide there were three deaths, one in the locality, involving trees blown down – two landed on cars, one was a guy trying to clear a fallen tree on the road and another tree/part of a tree landed on him. But that was the worst of it and really we got off lightly, apart from some roofs being blown off in various parts of the country. We didn’t even get the yellow/orange/red skies the English are now getting!

    And there were some eejits who went windsurfing, surfing and swimming even in the gale force winds. You couldn’t be up to them!

    • Loquat says:

      I believe this video from Hurricane Sandy is relevant to that last part.

    • Thegnskald says:

      Hurricanes are usually no big deal. Usually.

      Just be cautious, if it happens again, not to assume it will be like this time. The deadliest hurricanes tend to be relatively shortly after a relatively harmless hurricane was predicted to be deadly, because people get complacent and assume the forecasters always overstate the dangers.

  13. actinide meta says:

    Slate Star Codex has a serious information leak which can compromise commenters’ e-mail addresses. I don’t want to go into more detail because the leak has not been patched, but I have reason to think it was discovered a long time ago. If you care about your pseudonymity, I recommend changing your registered e-mail to something you don’t mind being disclosed before posting any more comments.

    Scott, I’ve attempted to contact you by e-mail, to get the problem fixed before disclosing it, but I’ve received no response, and at this point I feel like this is better than taking no action. Please check your shireroth e-mail, which should already have a more detailed explanation of the problem.

    • Edward Scizorhands says:

      I’m following up to increase visibility. I’ve made two separate attempts to reach out on private channels with no response. The number of people who know about this is large but doesn’t include the people who can fix it.

      Everyone else: change your email address. Go to https://slatestarcodex.com/wp-admin/profile.php. Make it garbage if you want.

    • Standing in the Shadows says:

      Is this “leak” the “the URL to your user icon contains the MD5 hash of your email address” one?

      Because that is literally how gravitar works, and how it is supposed to work.

    • Brad says:

      Can confirm. This is bad.

    • Nornagest says:

      I see it too. Yeah, that’s bad.

      Garbled my email. Hopefully we can get this fixed soon.

    • rlms says:

      Fourthed. Luckily I don’t care about my pseudonymity (see the link in my username).

    • Bakkot says:

      This was fixed a while ago and got un-fixed at some point; it should now be fixed again, I think. Sorry it took so long; someone reached out to me a while ago, but it got lost in the noise. (This isn’t really my area – I’m just in charge of the client-side JS stuff – but I happened to have looked into it previously.)

      That said, as pointed out below, the icon system on the blog exposes an MD5 hash of your email address, which is only barely more secure than plaintext. So you should not rely on the email address you use here being secret, even now.

      (You shouldn’t anywhere else, for that matter.)

      • Wrong Species says:

        If I change email addresses now, will it still be a problem?

      • Aapje says:

        @Bakkot

        someone reached out to me a while ago, but it got lost in the noise.

        That was me, probably. I sent you a Reddit private message some time ago.

        @Wrong Species

        I’m think that your MD5 changes if you change your email. It certainly messed up my icon until I fixed it, which suggests that it did.

      • rlms says:

        I think there is a pretty big difference between the security of MD5 hashes of emails and the problem under discussion. This guy managed to work out 70% of email addresses on some French forums from their gravatars by using GPUs and custom cracking software. The leak here allowed people with no technical expertise to find out the email address of any recent commenter without doing any computing.

  14. sandoratthezoo says:

    We should reduce the length of patent protection to 7 years (from the current 20), but, for any patentable item that requires FDA approval or similar government approval in order to legally sell it, we should “refund” the time it takes to get approval (or final denial, in the rare case where there may still be value to a patent that has been denied).

    Tell me how wrong I am below.

    • Lillian says:

      Patent protection is fine as is, twenty years is a decent and reasonable term. It’s copyright that has gone off the deep end. Lifetime royalties plus 70 years after the death of the author is ridiculous and moronic. There is no reason why J. K. Rowling’s great-grandchildren should continue to profit from Harry Potter.

    • Douglas Knight says:

      What problem are you trying to solve at the FDA?

      What do you mean by time to approval?

      1. Time from submission to response?
      Are you trying to give them an incentive to move faster? Long approval time is no skin off of their nose. And under your system, the drug company might slow it down, if doing so could maximize the odds (or, more likely, optimize the wording).

      2. Or do you mean 7 years from approval? That sounds pretty bad to me. The drug company will try to get lots of data before submission and the clock won’t be ticking on other people building on it.

      • sandoratthezoo says:

        I’m not trying to solve a problem at the FDA. I’m trying to solve the problem that spending 20 years profiting from your patent is way too long. But that if you just make it let’s say 10 years (or 7, or 5), then there will be lots of drugs that only can actually be sold for a few years — or less, maybe zero — because FDA approval takes many years, especially in the “moderately bad” case.

        I don’t know why we’d care if a company tries to get lots of data before submission, and it seems like there’s already plenty of incentive to get your drug on the market as early as possible even if you have a guaranteed 7 year interval on the market no matter when it starts. It is surely better to profit from your drug from 2020-2027 than to profit from your drug from 2025-2032, all else being equal.

    • Brad says:

      There already exists provisions in law to extend patents of drugs to take into account the FDA approval process. See https://www.fda.gov/drugs/developmentapprovalprocess/smallbusinessassistance/ucm069959.htm

      In addition there is Hatch-Waxman exclusivity which is completely separate from the patent system: https://www.fda.gov/downloads/drugs/developmentapprovalprocess/smallbusinessassistance/ucm447307.pdf

    • pontifex says:

      Is there anything we actually need patents for? We have publicly funded research. We have the internet to distribute research worldwide. We even have tax breaks for corporate R&D. Most of the economic growth in the last few years has come in sectors that have limited to no patent protection, like software.

      Maybe we need some kind of patent-like system to incentivize drug research. But it doesn’t have to actually be patents. It could just be a 20 or 30 year period after the FDA approval of a drug, wherein only the company that pushed it through the approval process has FDA approval to market and distribute it. We might actually want to provide drug discovery bounties too, since the current system seems to be failing to properly incentivize certain lines of research– like new antibiotics.

      • CatCube says:

        It could just be a 20 or 30 year period after the FDA approval of a drug, wherein only the company that pushed it through the approval process has FDA approval to market and distribute it.

        It’s…not clear to me what the distinction between this and a patent is. Except that a patent is shorter.

        • Douglas Knight says:

          We already have this (of short duration), as Brad mentioned.

          Being administered by the FDA rather than the Patent Office is a very big deal. The point is to reward good drugs. This is two steps, invention and verification. The patent office decides what counts as invention and provides most of the reward in the form of monopoly. But you still have to get through the FDA. If the FDA judges that you did a lot of extra invention to convince them, they should reward you with a monopoly.

          It’s not clear what pontifex means, but proposals like this often have the problem that they result in races, where two drug companies attempt to do the same studies at the same time, the first to finish winning all the rewards. Whereas, in the current system, there’s usually no race to patent the chemical, because it isn’t worth much, and then the monopolist is the only one to put in work on that chemical. The problem is that if no one puts in the work before the patent runs out, no one will. (Except that the FDA can now provide ~2 years monopoly in return. more for special classes of drugs.)

          (The current system does result in races between “me-too” drugs that all use the same mechanism, but aren’t covered by the patent. The first to market doesn’t keep the others out, but usually gets most of the market share.)

  15. ADifferentAnonymous says:

    Have liberal prison reform advocates ever considered starting a non-profit to operate a private prison? This could probably work a lot faster than trying to convince legislatures to give the progressive ideals a shot, and if it could achieve impressive recidivism reduction, that would probably go a long way towards getting its approach adopted more widely.

    • HeelBearCub says:

      Without buy in from the government, this is highly unlikely to work well.

      If you are operating alongside for profit prisons that are squeezing costs and juicing retention rates, your ability to operate as a high-service non-profit will necessarily be limited. Absent some cooperation from the DOC, ensuring approval of non-standard programs and providing comparable prison populations to provide proof of lower recidivism, showing results will be difficult.

      And then there is simply the long time lines to consider. Results on recidivism will not be available for years.

      • ADifferentAnonymous says:

        My thought was that donations would subsidize costs to allow competition with for-profits. As for timeline length, I’d expect meaningful results well within a decade, which doesn’t seem prohibitive.

        The other two concerns are the big questions for me: what are the legal barriers to a progressive private prison? And how close to randomized is prison assignment? I’m sure both of these issues would be big headaches for anyone trying this idea, but bearing in mind that one could shop around for the friendliest state, I do wonder if anyone has seriously considered it.

    • Nabil ad Dajjal says:

      Putting aside practical concerns of the kind HeelBearCub mentioned, generally speaking these are exactly the people who are least likely to find “then start your own firm!” convincing.

      In my experience, those on the left who support prison reform are also deeply skeptical about markets. Most find the very concept of a private prison immoral and support nationalization of the prison industry. Asking them to run a private prison, even a non-profit one, is asking them to be part of a system which they view as fundamentally corrupt.

      Beyond that, I would imagine that prison reform advocates believe that they already possess the evidence that this experiment would provide. If you don’t notice the differences in demographic composition between Scandinavia and the US, then it makes sense that adopting their extraordinarily lenient prison policies would produce a similar result. Noticing those demographic differences is racist.

      • HeelBearCub says:

        I don’t think this is really true.

        As a “for instance”, my sister works for a private foundation that advocates for the food justice movement. They run a working farm where they provide agricultural goods, primarily to “farm to table” restaurants (they also run or did run a restaurant). This is an adjunct to their grant work. They first incubate their ideas on the farm and then use the grant making process to help private, for profit farmers enact those ideas.

        They completely recognize that making the ideas profitable in the broad marketplace is essential to these ideas coming to fruition.

        The market in “prison services” is going to remarkably distorted for a lack of customers.

        • AnonYEmous says:

          Isn’t there some complicating factor though, insofar as operating a farm is natural and so on and so forth? You can find lefty communes and co-ops farming the earth, all across the world even (what else is a kibbutz meant to be?). What I guess I’m trying to get across is that a lot of left types are attracted to farming, which might be why something like this goes against what you would expect.

          I will say that your sister and their foundation sound a lot more willing to work within the system than some extreme types I’d expect and maybe those people are a lot more prevalent / prevalent within ‘the left’ than I personally expect.

          • Aapje says:

            @AnonYEmous

            Lots of lefties are in caring professions, so it shouldn’t be too hard to find people who want help to reform prisoners. You may have difficulty finding those who want to carry shotguns around and/or merely patrol, but isn’t that the sort of thing that the liberal prison would do away with? It would be a low security prison anyway.

            You could also set up a prison like a kibbutz…

            It would also be very interesting to try to run a prison as a partial democracy, where the budget is transparent and prisoners can decide to some extent where and how the money is spent. Do the prisoners want better meals? They can vote to buy better products. Unhappy with the meals? They can vote to have more curries (or whatever). Do they want more pay for work? They can vote for that. Although…they might not necessarily desire the things that those running the prison like, so this may appeal a lot more to libertarians than non-libertarian lefties.

          • Paul Brinkley says:

            I can see a few possible failure modes that liberals might be worried about:

            1. A food co-op failing means a few thousand servings of food goes to waste. A prison co-op failing means a few violent prisoners escape and someone later gets hurt. The latter is very visible; success is largely unnoticed.

            2. If the prison is nice enough, there may end up being people who commit crime deliberately to get into such a prison, as it would be an improvement over their current lifestyle. Conservatives would probably fuss about this.

            3. Prisons are supposed to be the government’s job. That’s probably not going to change any time soon(?). As long as it’s up to the government, it’s going to be a monopsony (as HBC hinted above). As such, any co-op will be competing with interests that aren’t necessarily better at running humane prisons, but rather better at persuading the government that they will make the elected official appear tough on crime.

            4. Liberals are caring, but this care is primarily toward the downtrodden – this sympathy is harder to acquire when violent felons are the subject. A co-op will consequently have too much trouble finding people willing to work the prison the way they want. (Not a slam on them; most people don’t want to work with prisoners.)

            If this idea took off at all, I would expect the first one to be for non-violent offenders, especially if they appear amenable to rehabilitation. Possibly a women’s prison.

    • [Thing] says:

      Have liberal prison reform advocates ever considered starting a non-profit to operate a private prison?

      Yes, apparently. This was the first result when I googled “nonprofit private prison”: Nonprofit Floats Unusual Alternative To Private Prison | HuffPost. It’s a nonprofit prison-reform activist group proposing pretty much exactly what you are. It’s from 2014, though, and more recent news suggests the privately run prison they were hoping to take over will revert to government control instead.

  16. dodrian says:

    I’m thinking of planning a ~1 week vacation in late Jan / early Feb. Ideally it would combine some light-to-moderate outdoor activities (hiking or canoeing or similar) with some cultural ones (a unique museum, an area of historical interest, a regional art, etc). My initial thought is to stay within the US, but I’m open to suggestions that aren’t too expensive or difficult to get to from Texas.

    Some cultural places I’ve always wanted to visit are New Orleans, Nashville, and DC. Some parks I’ve wanted to visit are Crater Lake (probably not a winter destination), Yosemite, and Zion.

    Any suggestions of other places to consider, or city/outdoors trip that would incorporate the above?

    • Chalid says:

      San Francisco would have pleasant weather, lots of culture, and very nice hiking nearby.

      • dodrian says:

        That does sound like a good option, thanks!

      • Chalid says:

        Another idea – Hawaii. Most people go there to lie around on the beach all day, and it’s undeniably great for that, but there’s also great hiking, kayaking, snorkeling, etc in an absolutely beautiful setting. For culture there’s a lot of Native Hawaiian stuff that will be very different from what you’d have seen before if you haven’t been out of the continental US. There’s also Pearl Harbor.

        • dodrian says:

          Went to Maui February last year, it was phenomenal. Sadly my budget this year is much smaller, so I don’t think I’ll be going back.

    • onyomi says:

      New Orleans would be Mardi Gras around that time. New Orleans Mardi Gras is a bit crazy (and hotel rooms expensive; stay out of the French Quarter for less crazy+more family-friendly, less expensive), but really worth seeing, imo.

    • bean says:

      I went to DC in January once. Probably not a place I’d choose to take a vacation to at that time of year, particularly if I wanted to hike.
      California might be a good option. I’m most familiar with LA. In terms of things to do, there’s a battleship (I couldn’t resist, sorry), and a bunch of museums for almost anything. Joshua Tree National Park is 2-4 hours away, depending on traffic, and the highs should be in the 60s or 70s. Might not want to camp unless you’re OK with it getting a lot colder, though. There’s hiking closer, but I have no experience with it.
      Weren’t you the one who was looking at coming to visit Iowa a couple months back, but went to Griffith instead?

      • dodrian says:

        Yes, I’m afraid that was me! We actually went to Joshua Tree as well, it was great to see in the car (but 110F+, so no hiking).

        The battleship is top on my list of things to do next time I visit LA, I promise 😉

      • Chalid says:

        Death Valley isn’t your classic outdoors trip but it’s got some incredible, alien landscapes and it is a pleasant temperature in January/February. Four hours from LA is a bit far though.

        • dodrian says:

          Four hours being far is an opinion. It’s how far we travel to visit my mother in law for the weekend every other month or so!

          • bean says:

            There’s a huge difference between four hours as practiced in SoCal and four hours as practiced on open highways. Part of the trip to Death Valley won’t be so bad, but unless you happen to be right on the relevant edge of town, you’re going to have an hour or more of LA freeway, which, even when it’s not rush hour, gets old really quickly. The worst part of my drive from LA to OKC was the two hours it took me to get to Palm Springs. After that, driving was actually pleasant. (Except in Albuquerque, where they just don’t know how to drive.)

      • Nornagest says:

        DC in January beats the hell out of DC in August, I can tell you that. Not a good time for hiking, though.

        • The Nybbler says:

          There is no good time for hiking in DC; if it’s not hot and humid it’s cold or rainy. The Catoctin Mountains in Western Maryland aren’t far, though. Best in spring or fall and tolerable in the summer. I’ve never been there midwinter, but it’s probably fine if the weather is clear and there hasn’t been a huge snowstorm recently.

    • professorgerm says:

      That time of year, I’d lean towards Nashville. Good music scene, not known for terrible weather.

      Zion is one of the most beautiful places I’ve been to, but I do not recommend going in winter (really, more like November-late March or so). The roads will likely be impassable and most of the little towns nearby shut down right after Thanksgiving until the thaw. Plus, there’s nothing within several hours drive of Zion except a couple other national parks (Arches, Bryce Canyon, etc). Salt Lake City is around 5 hours away, and I found it to be a bit of a hidden gem. The Natural History Museum of Utah has (as I recall) the largest collection of triceratops anywhere, and the food scene is surprisingly nice.

      • dodrian says:

        Weather-wise I don’t mind a bit of cold or snow, so long as the roads are clear and I don’t need special equipment other than a few layers and some decent hiking boots. I hadn’t imagined Utah as particularly snowy, though it looks like I may have been wrong there!

        • gbdub says:

          Lots of Utah is desert, but it’s a pretty high altitude desert, so it definitely gets cold / snowy.

          We did a driving tour of the “Utah 5” national parks (highly recommended) over the Thanksgiving week last year. Had some very nasty snow on the way to Bryce from Zion, and it was definitely chilly for most mornings. And proffesorgerm is right that a lot of the supporting towns (not as bad for Zion, but definitely Moab and the North Rim) shut down all or a lot of services either after Oct. 31 or after Thanksgiving.

        • professorgerm says:

          That’s fair! The interstates should be clear if you aren’t caught in the middle of a storm (driving in a whiteout to Zion was… less than fun). Overall they did a good job of keeping them clear in my experience.

          And as gbdub mentioned, it’s the altitude. The first week of June 2016 it was in the 60s/70s in SLC; up the mountain to Snowbird Resort (30 minute drive and maybe 3000 feet higher) there was still a foot or so of snow and ice. Was fun stomping around in the snow in shorts, though.

    • gbdub says:

      San Diego. A ton of museums (including great naval ones) and parks (and not just the famous zoo), weather will be cool but pleasant, it’s probably too cold to swim/surf so finding accommodations should be easier/cheaper. More breweries than you could reasonably visit, if you’re into that sort of thing. Much smaller than LA, so not nearly so annoying to get around. Definite hiking / natural scenery opportunities if you go up the coast, or a couple hours inland are the Imperial Dunes (where they filmed the Dune Sea portions of ROTJ).

    • Bryce Canyon is pretty impressive.

      For a very different sort of thing, the Tucson Gem and Mineral show in early February, the largest such event in the world, is lots of fun if you like markets, especially if you have any interest in gems, minerals, or jewelry. The main show in the convention center isn’t bad, but the fun part is wandering through motels converted for the week into markets. A nice lady from Africa with a pile of ruby rough in front of her. Ethiopians with opal. Finns with spectrolite. People selling tools.

    • Nornagest says:

      Yosemite is definitely worth seeing, but it’s at about four thousand feet on the wet side of the Sierras, so January or February probably isn’t the time for it. You’ll likely be able to get around on the valley floor unless there’s been a recent storm and the roads are closed, but all the trails will be covered in snow (and unhikeable without snowshoes) until at least March, maybe April. Camping might be difficult for the same reason, and while there are hotel rooms in the valley and they probably won’t be sold out that time of year, they’ll still be expensive.

      • Paul Brinkley says:

        As someone who grew up in Texas and its road system, I will advise against Yosemite unless you’re okay with driving on narrow snow-covered roads with multi-hundred-foot dropoffs all along one side and not much in the way of guard rails. I drove in and out of Yosemite Valley twice last year, during a clear, dry September, and I found the experience harrowing. And I usually drive faster than average on the interstate. But maybe you drive around west Austin all the time and are used to that, in which case, you’ll have a blast. (But bring layers. And boot spikes.)

        If you do go, hopefully you’ll also be within an hour or so of the Bay area, in which case, you might enjoy the Winchester House. Also, the Pinnacles is close by – the US’ newest national park.

        • Nornagest says:

          If you go to the Winchester House, do the flashlight tour. I found the daylight one kinda underwhelming for the price.

          • I live a mile or two from the Winchester house and have never visited it. Should I?

          • Nornagest says:

            You don’t strike me as the type to enjoy it much. You might like the Winchester rifle museum they’ve got on the premises, though; they have a pretty comprehensive collection.

    • SamChevre says:

      I went with a friend to Pensacola FL in February a couple years ago. Very interesting history (Forts Pickens and Barrancas), pretty beach/Gulf hiking at the National Seashore.

      Nashville in January/February is iffy for hiking, but there’s plenty to see and do indoors. If the weather is good, I’d probably go to the Stones River battlefield for a mixture of history and outdoors.

    • Paul Brinkley says:

      DC is arguably better to visit during the spring, when the cherry blossoms are in progress. Trouble is, that’s a one-month window, and hard to predict more than a week out. If you’re not set on cherry blossoms, though, it should be fine. No inaugurations or fireworks, so you’ll be okay traffic-wise (relative to DC, that is; it’s never good).

      IMO, the best hiking is in Shenandoah and environs. Old Rag is the usual thing to do. I often go with a DC Metropolitan Hikers group on Meetup; they usually have something nearly every weekend, and it’s worth my driving down from MD. (And sometimes I join them for Sugarloaf Mountain in MD, followed by a delightful stop at the winery at the base.)

      If you decide on DC, hopefully it’s around the time of one of our meetups. Reliably at least a dozen or two of us.

      • dodrian says:

        What would hiking in Shenandoah in February be like? Are good hiking boots enough, or would there be more than an inch or two of snow on the trails?

        • Paul Brinkley says:

          It will of course vary a lot by trail, by weather at the time, and by your experience. If you’re only hiking twice a year, say, then any grade might give you enough trouble that you’ll want one or two walking sticks. Most of the AT (Appalachian Trail) is so flat, though, that we have septagenarians walking it and just being slower about it. …except maybe for the part just south of Harper’s Ferry, and surrounding areas; those were steep enough that I was slipping and sliding my way down a couple of the trails, enough so that I stopped by REI the next weekend and bought a pair of slip-on chains… that I’ve yet to use after five years, because the weather’s been on the warm side. (Last winter, it barely snowed.)

          If you’re the type who prepares for the 10% case, I’d advise you to be ready for drifts of up to six inches, but only in flat areas; inclines will have less. I typically hike in a pair of waterproof boots sold by Keen; they come up just above my ankles, and have been great, combined with any pair of jeans.

          A few treks include water crossings (e.g. Devil’s Staircase). I strongly recommend walking sticks for those. (I had none, and it was… exciting.) The AT has nothing like that though.

          Plan for snow or rain, just in case. And again: layers.

    • keranih says:

      Supposedly in winter in Yellowstone you get wolves, which I would count as super cool.

      Try St Augustine, FL. Oldest inhabited town in the Americans, or close to it. If you get further down the FL east coast, there’s Cape Canaveral, and all the space stuff there. But I think St Augustine is closer to what you want.

      New Orleans is…well. It’s an over crowded urban town. I would expect the tourist load would be lower in January. I’ve been wanting to go to the WWII museum, but haven’t yet.

      • Nornagest says:

        The oldest continuously inhabited town in the States is probably Taos, New Mexico — its pueblo, which is still inhabited, was most likely built sometime in the 13th or 14th centuries but could conceivably date as far back as 1000 AD. Spanish contact came in 1540, permanent Spanish settlement sometime before 1620.

        Some cities in Mexico could give it some competition, though — Mexico City for example was founded (as Tenochtitlan) in 1325. I expect there are older ones but I’ve no idea which.

        • keranih says:

          I slouch corrected!

          (Perhaps St Augustine’s the oldest one in the eastern USA. Or maybe the oldest Euro settlement.)

        • CatCube says:

          There was evidence of continuous habitation at Celilo Village at the foot of Celilo Falls on the Columbia River from about 13,000 BC until 1957, when the falls were submerged by the closure of The Dalles Dam. There was a new, current village built above the forebay, but I don’t know if you’d consider this “continuously occupied.”

  17. Amelia Preston says:

    Hi. Hear you like motte and baileys.

    That’s a video of Matt Easton looking around the ruins of Bramber castle, which is of the motte and bailey kind. It is not well preserved, but I think it’s a good illustration of the general idea.

    • Aapje says:

      But…but…we actually dislike motte and baileys.

      • [Thing] says:

        Sure, sure … SSC fans love to proclaim their disdain for motte and bailey tactics to anyone who will listen, and they’re all too happy to then take advantage of the benefits that naturally accrue to those who signal intellectual superiority via contrarian opinions about the esoterica of Medieval fortification design. But the moment they receive the slightest pushback, they fall right back to claiming they were merely objecting to “motte and bailey” the rhetorical tactic, and what sort fool would you have to be to disagree with that? Honestly, you see that sort of underhanded, disingenuous doublespeak so often, there ought to be a name for it …

    • aNeopuritan says:

      The thing of concern here should have stayed called “strategic equivocation”, as per the post where Scott presents it. I do like Matt Easton’s work.

  18. bean says:

    In Naval Gazing news, Said Achmiz has kindly agreed to host me. I’ve gotten a couple of semi-prototype posts written up, and I’m frankly astonished at how much better it reads now that I can incorporate images. The current plan is to put up a pair of posts I have written on mine warfare in 86.75 and 87, and then suspend until I get all the loose ends tied up and start posting there. I’ll gradually update and move the stuff posted here, as well as writing up new content.

    • bean says:

      John, I found some really interesting data on our long-ago discussion of dual-purpose batteries in WWI. The US found that at 6,000 yards, the danger space for the 5″/51 was 96 yds, the 4″ (I think the 4″/50 used on US destroyers in that era) was 71 yds and the 5″/25 was 36 yds. So the low-velocity gun was about a third as accurate as the high-velocity gun against destroyers. (I’m not sure if these numbers are before or after the USN started including ship beams in danger spaces. If they aren’t, add 10 yards to each, the beam of a contemporary destroyer.)

      • John Schilling says:

        Interesting. The 5″/25 had about twice the rate of fire, leaving the 5″/51 with a net 50% edge in anti-ship performance. Obvious question, given the importance of muzzle velocity in AA fire, is why not just make a high-angle mount for the existing 5″/51 gun?

        • bean says:

          One of the main drivers for the 5″/25 was a short, handy weapon. The 5″/51 was not such a weapon, you’d have trouble with loading at high angles, and you probably would have needed power ramming, which was not something they favored at the time. I doubt you could maintain half the ROF at high angles. Also, under some doctrines for AA fire in the 20s (there were a lot, believe me), they were mostly trying to create barrages the planes flew through. In that case, ROF dominates muzzle velocity. The fire control probably couldn’t react fast enough for the shorter flight time to be a big factor before about 1930, which is when the US began developing what eventually became the 5″/38.

      • Andrew Hunter says:

        For those who had never heard the term before, this seems like a good explanation of “danger space.”

        • bean says:

          Hey! I defined it in the OT86 post. (And put an image in in the revised version, which helps.)
          Also, how did I find a new naval gunnery textbook (from 1917, but still) in that article’s cites?

          Re USN definitions of danger space, Friedman says they’d switched to including deck by 1917. My usual heuristic is to trust him over other sources unless I have good reason not to. I’m actually rather baffled by that post, as the author explicitly cites Naval Firepower, which is the book I just checked. And danger space is defined at the beginning of the first chapter. Did he just skip to the USN chapter? The RN kept using side-only for a much longer time, which probably had a lot to do with the USN’s greater focus on long-range combat in that era. I’d guess they moved over around the time the Nevadas were designed. The RN didn’t get there until 1916/1917, and they were too busy having a war to play definition games then.
          (And just to be clear, the numbers I gave, and the 5″/25 itself, are post-1917, so I’d assume the deck is included. Probably.)

      • gbdub says:

        So NavWeaps defines danger space as: “Danger Space – That distance in front of the target, measured parallel to the line of fire, that the target could be moved toward the firing point, so that a shot striking the base (waterline) of the target in its original position would strike the top of the target in its new position. The flatter the trajectory, the greater the danger space.”

        Is it really fair to call that a measure of accuracy? It sounds more like a measurement of your margin of error, or “how accurate you need to be”. That is, you could mis-estimate the range by up to the danger space (actually half the danger space, if you’re aiming at the “center” of the ship) and still score a hit. But it doesn’t measure the accuracy of the gun itself (which I would naively describe as the likelihood that a given shot goes where you aim it).

        • bean says:

          To unpack my thinking slightly, I was basically assuming that the range scatter was greater than the danger space by a substantial amount, so that the percentage of hits would be proportional to the danger space. Some research shows this to be a good assumption, as the 5″/51 batteries were patterning 329 yds at 10,258 yards. If we assume that the pattern size is proportional to range, then you have 192 yds at the 6,000 of these tests, or twice the danger space. If they aim at the right spot, they’ll get 50% hits, while a 5″/25 aimed at the right spot (assuming the pattern is the same size, which is slightly conservative) will only get 19%.
          “Accuracy” may not be precisely the right term to use, but the numbers do show that all else equal, the 5″/51 is going to give a lot higher hitting rate, and the system as a whole can’t put in a good enough performance to remove this advantage. The 5″/38 was a more accurate gun, but even for it, a full 6-gun pattern was ~84 yds at 6,000 yards range. The gun scatter was important in making up for the inherent inaccuracy of the system.

          • gbdub says:

            Makes sense, and the assumption probably works here. But as you note, you’re implicitly assuming that the scatter is the same for both guns. If they were very different, that would need to be included in the comparison.

            In the event, how big was the danger space (and/or the gun scatter) relative to the accuracy of rangefinding? If your initial range estimate were very poor, and you had to “walk in” your fire, then increased ROF might give you some added advantage in getting you more iterations per unit time as you adjust your target point.

          • bean says:

            Makes sense, and the assumption probably works here. But as you note, you’re implicitly assuming that the scatter is the same for both guns. If they were very different, that would need to be included in the comparison.

            This is data I don’t have. In this particular case, most of the scatter is probably pointing error. The weapons weren’t power-operated, and definitely didn’t have RPC, so you’re going to see guns pointed slightly wrong. The 5″/25 might have a slight advantage, but not a big one, due to being lighter and easier to point. And its shell is flying for a longer period of time, which usually leads to increased scatter. Given the magnitude of the difference, it’s hard to see the 5″/25 coming out ahead. Both guns were generally well-regarded.

            In the event, how big was the danger space (and/or the gun scatter) relative to the accuracy of rangefinding?

            I don’t have hard numbers on hand, but control of the secondary batteries before the 5″/38 was pretty poor.

            If your initial range estimate were very poor, and you had to “walk in” your fire, then increased ROF might give you some added advantage in getting you more iterations per unit time as you adjust your target point.

            Not really, although the idea is good. Time of flight for the 5″/51 to 5000 yds was 6.2 sec. RoF for the 5″/51 was 8-9 rpm, so at 6,000 yds, waiting for the previous salvo to fall would set their ROF, to say nothing of the 5″/25 (which Navweps doesn’t give TOF for). There are some games you can play with patterns of fire to make this a bit better, but you’re still stuck waiting for all the shells to fall.

    • Andrew Hunter says:

      I’m going to miss the discussion here, I have to say. Am I wrong to think that not all of the SSC commentariat who likes to kibitz on Naval Gazing is going to make the transition to commenting on your posts directly?

      • bean says:

        Probably not. I was planning to cross-post here until I saw how much better it works in that format, particularly the pictures. What I may do, at least for a bit, is post links in the OTs, and run two discussion threads.

      • Said Achmiz says:

        Here’s a question/request/challenge, for you or for anyone else with an interest in this:

        Do you have or know of a way, in PHP, to retrieve part of a remote page—specifically, an element of a specified ID, and all of its descendants?

        If someone were to provide me with a way to do this (or point me to a sufficiently specific / complete implementation, or instructions for one, elsewhere), then I could construct a way for bean to embed comments posted here (in a specified thread) in the comments section of the corresponding post on his blog. That would provide a way to unify the discussions (in one direction, at least).

        I unfortunately can’t devote the time that developing this myself from scratch would require, but if someone else provided me with the hard part already done, then adding this capability to bean’s blog would not be too burdensome. (Bonus points if someone were to build this wholesale, as a PmWiki recipe—perhaps building on this one?)

        • Brad says:

          I don’t think the http protocol allows for that. Your only choice is to do GET on the URL get the whole page and throw away the parts you don’t want.

          • Said Achmiz says:

            Yes, I’m aware; the solution would, presumably, involve parsing the HTML response to get the relevant chunk, which is what I’m asking for. This really does not seem hard, it’s just not the sort of thing I have the time to implement right now, hence my question/request.

        • Aapje says:

          I’m not a PHP developer, but a quick google turns this up:

          http://simplehtmldom.sourceforge.net/

          Then you can do something like:

          $html = file_get_html('https://slatestarcodex.com/2017/10/15/open-thread-86-5/');

          $ret = $html->find('div[id=comment-557164]');

          Then you can loop over elements and/or traverse the DOM.

        • rlms says:

          I can give you my Python code for scraping the SSC comments if that would be helpful. Maybe at some point I’ll put it online and expose an API, but that probably won’t happen any time soon.

          • Said Achmiz says:

            Hmm… probably not? I mean, unless there are some non-obvious bits of technique that would be relevant for a PHP implementation—are there? Otherwise, yeah, it’s got to be PHP, I’m afraid.

    • bean says:

      A related request. When thinking about things I’m going to need to do to update old posts, the biggest gap I have is probably in mapping. The Jutland posts are most prominent here. Personally, I dislike maps where they put a whole battle on a single sheet with a bunch of numbers for times, and making maps with shorter timescales would work well in my format. What I don’t have is a particularly good way of constructing them. Does anyone have suggestions? Google maps doesn’t seem quite right, and doing it in GIMP or something like that seems like a lot of work.

      • hlynkacg says:

        I wrote some custom tools that I might be able to hook you up with, but I’ll have to double-check and that there’s nothing proprietary in there.

      • orihara says:

        There’s a javascript tool called Leaflet. It’s designed for mapping applications. Some plugins let you do things like time-based rendering (have a time slider, and stuff moves around appropriately). I’ve been meaning to play around with the time based plugins, but haven’t yet gotten around to it myself yet.

        • bean says:

          While that sounds amazing, and something I dearly wished someone had done for Jutland when I was writing that series, it’s a step above what I need right now. I’m basically planning to trace existing maps, simplify them somewhat for web resolution use, and do the time slicing myself.

          • cassander says:

            If I ever have a team of grad students and a free summer, I want build something like this for the pacific war at the strategic level, getting the movements between the great battles.

      • Matt C says:

        I suspect this is comparable to whatever Google Maps offers, but you might take a peek at Mapbox Studio. https://www.mapbox.com/mapbox-studio/

  19. bean says:

    This is the last post for a while on air travel, unless my sister (who commented on this post) chooses to write some. I’m focusing on Naval Gazing for now.

    Also, it was recently noted by someone on the Discord that commercial air travel is safer per mile than being fired into the sun. I wanted to signal-boost this, as it’s an amusing look at how well the aviation safety people do their jobs.

    One topic that has come up several times in recent discussions of airlines is boarding methods, and the supposed inefficiency of the methods most airlines use. As such, I thought a column on the subject was merited.

    This question almost inevitably starts with someone asking why they don’t line everyone up by seat in the gate area and send them in by column (left window-right window-left middle etc). In theory, this should give the absolute minimum time, as each person has enough space to work, and doesn’t have to worry about someone being in the way.

    There are a couple of reasons this is impractical. [sister bean: yes, bean is saying that the method Mythbusters found when they tried to come up with the most efficient boarding method is actually quite terrible. I wholeheartedly agree with him.] First, while it works great in the computer, and when tried out by volunteers, these are not the best representations of what goes on during an actual boarding. Volunteers and computer simulations alike are alert, interested in the process, traveling alone, and aren’t on late connections. How well does this describe the last flight you were on? A good boarding method needs to be robust against all of the various problems people can have while trying to get on the airplane. Parents traveling with small children. People who are exhausted, scared, distracted or otherwise impaired. People who don’t speak the local language very well. People who are flying for the first time and don’t know what to do. People who have mobility issues. People who had late connections and should have already boarded, but are just now running up to the gate. How many of these problems each flight has will obviously vary widely. The described method would probably work very well on a flight from JFK to ORD at 5 PM on a Friday, when everyone onboard is a businessman. It would not work nearly as well on a flight during spring break from JFK to MIA. Airline customers value consistency, so a method is chosen that works in all cases the airline has to deal with.

    Before you dismiss this, I usually fly Southwest, the only US airline that routinely lines people up at the gate. And on a fair number of flights (I’d estimate 25%) I see someone mess up badly. Most common is someone holding a boarding pass in B or C getting in the line for an earlier group. There are also people who just don’t understand how the system works, and it’s not at all uncommon for the line to be slightly out of order. Because they don’t assign seats, it doesn’t matter much if someone who should be two places behind me is two ahead, but that would mess up the proposed system badly.

    The second problem is one of money. Early boarding positions are a fairly substantial source of ancillary revenue for the airlines, which this scheme would eliminate. They’re also a valuable elite benefit, and there are good reasons most airlines work hard to keep their elites happy.[sister bean: even Southwest charges for early boarding (A1-A15) with their “business class” fares, and they also charge to be automatically checked in.]

    Yes, I know what you’re going to say now. “But if we just got rid of checked bag fees, then there would be plenty of overhead bin space, and they wouldn’t need to sell priority boarding.” Leaving aside the financial problems with that approach, it isn’t true. Southwest does not charge for the first two checked bags, and they seem to only have marginally fewer carry-ons than other airlines I’ve flown. [sister bean: Southwest also uses the whole “bags fly free” tagline as a marketing scheme, especially with the rise of basic economy fares where there’s no carry-ons allowed.] More than that, there are several reasons to not check bags beyond saving money. The bag is with you and under your control at all times. Frequent fliers often say that there are two types of bags, carry-on and lost. You don’t have to wait at the baggage claim, either. Elite travelers, who have free checked bags as part of their status, are less likely than average to check bags.

    There are a couple of different options besides the system proposed at the start. I’ve already mentioned Southwest, but seat assignments are another thing airlines charge for, and I don’t see anyone following Southwest’s lead on giving up that revenue. The traditional system is to board in zones back-to-front after priority boarding. This is not a good system, as it puts most people in close proximity during the boarding process, although the increasing number of priority boarding groups has made it less bad recently. American uses a different system, basing non-priority boarding groups on method of check-in, but boarding throughout the cabin for each group. Variations on outside-in with random lineup order are in use by a few airlines

    • Jiro says:

      Also, it was recently noted by someone on the Discord that commercial air travel is safer per mile than being fired into the sun. I wanted to signal-boost this, as it’s an amusing look at how well the aviation safety people do their jobs.

      You can’t get the death rate per mile by dividing the death rate for the whole trip by the number of miles.

      Death rate for the trip = 1 – (1 – death rate per mile) ^ (number of miles).

      Death rate for the trip is *not* equal to (death rate per mile) * (number of miles).

      However, the latter is a good approximation of the former, as long as the overall death rate is much less than 1. So you can do that division for airplanes, and cars, and anywhere else you’d normally use it–just not for being fired into the sun.

      Furthermore, this sort of death rate assumes an equal likelihood of dying at each portion of the trip. Being shot into the sun has a lumpy distribution with all the death rate at the end. A trip which had an equal chance of dying at each point, and an overall death rate equivalent to the overall death rate of being fired into the sun, would have a 100% chance of death at every point on the trip.

      So if you do a proper comparison, airplanes and cars will always have fewer deaths per mile than being fired into the sun, and *not* because the death rate of being fired into the sun is 1 / (miles to the sun).

      • bean says:

        You can’t get the death rate per mile by dividing the death rate for the whole trip by the number of miles.

        Obviously, I can. How else would you get that number?
        It was an attempt at humor, and at putting the numbers for air travel deaths in a form that’s easier to grasp. The sun is very far away, but my chance of death flying that distance is less than 1.

        That said, the death rate per mile isn’t exactly flat for air travel, either. The majority of crashes happen near takeoff or landing, although I can’t remember the exact percentage. The airplane is near the ground (obviously), which means the crew has less time to deal with the problem, and the problems are more likely to show up because you’re using more systems closer to their limits.

      • ADifferentAnonymous says:

        Eh, I think that statement conveys the right intuition.

        Alternate formulations:

        If you could magically risk-swap a mile of your commercial flight with a random mile of a trip to the Sun, it’d be a bad idea (even if the latter trip only includes one fatal mile).

        You’d have to fly over an AU of total distance in order to expect to die in an accident (not quite equivalent, but it’s similar and also true).

        • Jiro says:

          That has the same problem: It would lead you to conclude that a 3 mile trip with a 50% chance of death per mile is worse than a 3 mile trip with a 100% chance of death at the end. Swapping a mile (in an otherwise safe trip) with the first one gives you a 50% chance of death and swapping a mile with the second one only gives you a 33.3% chance of death. Yet in fact the second trip is the more dangerous one.

    • bean says:

      Some interesting news in this field. Boeing has been pursuing an illegal subsidy case against Bombardier’s CSeries of smaller jets. The US was going to put a 220% tariff on the planes for ‘illegal dumping’. This was a terrible decision by all involved. Boeing doesn’t even make planes in that market any more, leaving aside general free trade considerations.
      Airbus just announced that they were partnering with Bombardier to bring an assembly line to Mobile, Alabama. There’s a lot of red faces in Puget Sound tonight. The Airbus marketing machine will now be available to bribe prospective buyers (only half joking), and this may well lead to greater sales of the airplane worldwide. Of course, the chief beneficiary is Delta, who has been the leader in the crusade against Emirates, Eithiad, and Qatar, while buying their airplanes overseas. I’ll give them credit for consistency, although I rather wish they’d show a speck of principle at some point. Maybe I should just buy their stock.
      At this point, I’m just going to say ‘A Pox on Everyone’s House’ and be done with it.

    • massivefocusedinaction says:

      I’m always surprised that airlines didn’t swap the bag fees, charging a premium for a guaranteed overhead bin space and allowing baggage to fly free.

      Even before the baggage fees, whenever possible I packed to carry everything on, both to greatly reduce the risk of a late/lost item as well as to speed my departure from the airport. I suspect an enormous number of other travelers have learned to do the same.

      • bean says:

        Spirit actually charges more for carry-ons than they do for checked bags, although a cynic might think this was a way to get more revenue out of people trying to avoid the awful bag-check lines they usually have.

        Charging for guaranteed overhead bin space is called ‘priority boarding’, and the legacy carriers have recently started to charge for carry-ons in the form of Basic Economy. (You have to pay a premium to not be in Basic Economy and get a carry-on.)

        That said, Ryanair seems to be doing something quite like this. They’ve banned all carry-ons for people who don’t pay for priority boarding, and lowered their bag fees, in an attempt to improve operational performance.

        • massivefocusedinaction says:

          Interesting, I haven’t yet flown on an airline that didn’t allow carryons on their base ticket, I’ll have to start watching for that.

          Thanks for the Ryanair story.

          • bean says:

            Assuming you’re American and have avoided the likes of Spirit and Allegiant, it’s a very recent change. Delta’s Basic Economy allows carry-ons. United and American rolled theirs out in the past 6 months and don’t.

        • Brad says:

          One problem with this model seems to be that no one wants to enforce baggage size limits. It’s been years since I’ve seen someone forced to check a bag that’s obviously bigger than the airlines say is the maximum carry-on bag size. And in terms of carriers that have tickets without a carry-on allowance they seem to have a lot of trouble enforcing the personal item size rules — both in terms of size and in terms of not sneaking them into the overheads.

          • Nornagest says:

            I’ve seen it, but mostly on small regional planes with smaller overhead compartments.

          • bean says:

            The problem there is, I think, one of work for the gate agents. They’re responsible for getting the flight out, and the last thing they want is to delay the flight by having a bunch of people parade through the bag sizer. Also, annoying customers with the bag sizer might be seen as worse than making them gate-check for whatever reason.

          • Brad says:

            That makes sense. But the upshot is that boarding order is really the only reliable lever there is (which I think has been your position all along). If basic economy gets mixed in with full fare economy then they are going to have their oversized personal items (and coats) in the bins and the last guy on is going to have to check his roller bag. Doesn’t matter what the flight attendants announce over the loudspeaker.

      • JayT says:

        I flew on Allegiant for the first time last weekend and their pricing structure was $40 for a carry on, and $50 for a checked bag. Since I was flying with my wife, it was obviously cheaper to have one large checked bag rather than two small carry on bags. I’ve never had my baggage lost, and I don’t check anything of real value, so I’m fine with the checked bag. Also, I like not having to worry about dragging a bag all around the airport.

        Because of this pricing structure, it was pretty obvious that everyone else on the plane had basically the same idea as me, because the overhead bins were mostly empty. Even though it was a full flight, I have never gotten on and off an airplane so fast before. I’d say it was at least twice as fast as normal. It was great.

        Everything else about my experience with Allegiant was quite nice too. I would fly with them again.

        • bean says:

          This is rather interesting, as Allegiant is the one airline that doesn’t need to do quick turns. Their aircraft utilization is something like half that of most airlines.

          • JayT says:

            I suspect their motivation is more that they wanted to charge more, and that the easy on and off of the plane was a happy coincidence. Their base rate was almost $300 less than the next cheapest airline, but by the time I paid for all the extras I only saved about $50. They make you pay extra for everything.

          • John Schilling says:

            Have they reached the point of selling the flight itself for fifty pence?

    • keranih says:

      A few bits of antidata:

      – Airlines have slowly been increasing the use of boarding groups – American and Delta both use them and I’ve seen people attempting to board early or boarding with oversize or over number bags (carry on plus purse plus backpack) asked to step aside and wait.

      – Nobody (domestic) serves free food to cattle car class anymore. On the flip side, the boxed snacks are cheaper than before and less likely to make you sick than the meals. Of note: Southwest booze is super cheap.

      – I would push back against the idea that checked bags are lost bags. It has been over a decade since I got a bag lost, and that was via an international flight (and the bag eventually showed up.)

      – Airplane folks tend to be fairly competent and decent. TSA is a whole nuther ball of wax, but there is a sharp variation from airport to airport, and large hubs are the worst. (Really, flying from small airport to hub to small airport seems to be the best way, if one can make it work.) Outside of the DMV, I don’t know another place where I have been exposed to more people with poor social skills and slow wits. I despair of a nation where DC beltway types and the DMV types run the place between them.

      – TSA equipment sucks. The scanner machines recognize neither braided hair nor bra snaps. The xray machines miss a multitool on the way out but catch it on the way back, or else confuse a metal clipboard with…god, I dunno. Also, stack a kindle on top of a laptop and the system blows a fuse every time.

      – Airports have exerted themselves to providing places to charge phones and get online. Airport food is still criminally overpriced, but that’s only to be expected.

      – I would welcome an effort post on airport security, +/- comparing TSA to Israel, but I completely understand if those in a position to know refuse to explain.

      • bean says:

        No post on security coming from me, sorry. I can’t defend the US system, and I don’t want to do any research, either.

  20. Loquat says:

    Effort post: Medicare (US verion)

    With yesterday being the start of the Annual Enrollment period for Medicare, it seemed like a good time to make a post explaining Medicare and the associated private plans one can get. As an insurance agent specializing in said private plans, I often hear from people new to Medicare who have no idea how it all works, and below is more or less the explanation I give them.

    For historical reasons, Medicare is divided into multiple parts. Essentially, Part A covers inpatient services like hospitalization, Part B covers outpatient services likes doctor visits, and Part D covers prescription medications. If you’re wondering what happened to Part C, it got assigned to Medicare Advantage plans, which will get their own section.

    As US residents may already know, Medicare usually doesn’t cover you 100%, but instead leaves a deductible and/or coinsurance for you to pay. Part A has a large deductible on hospitalization, currently $1316, and it’s per hospitalization not per year so it’s entirely possible to incur it more than once in the same year. Part B has a small annual deductible, currently $183, after which Medicare pays 80% and you pay 20% (of their approved amount, not the actual billed amount, but most providers that take Medicare accept the approved amount as payment in full, and if they don’t they’re supposed to tell you that up front). There are also various other out-of-pocket costs for long-term hospitalization, etc, which I’m not going to get into unless someone asks.

    A Medicare Supplement plan, also known as a Medigap, can pay part or all of these part A and B expenses for you. Typically these plans simply pay the same amount you would have had to pay, so there’s no need for the doctors to have a contract with the plan, be in a network, or anything like that. These plans are also standardized by law to make it easier to shop around, and 47 out of 50 states signed on to the same standardization scheme. (WI, MN, and MA each felt the need to have a unique scheme.) Confusingly, that 47-state scheme also uses letters, so that there are Medigap Plans A, B, etc, which are totally not the same thing as Medicare Parts A, B, etc. Another common pitfall is the assumption that because these plans “pay what Medicare doesn’t” that they’ll pay for services Medicare denies coverage for, which is very definitely not the case unless the plan materials specifically said otherwise. Medigaps are also not allowed to include prescription drug coverage, so you usually have to get a separate part D plan to handle that. Medigaps tend to be relatively expensive, with premiums usually over $100/month for all but the lowest coverage options.

    A Medicare Advantage plan, by contrast, replaces Medicare altogether with its own plan, which by law must cover all the same services as parts A and B, but can have totally different cost-sharing. They also frequently have networks, and can be set up as HMOs where you have to pick a primary care doctor and get referrals to see specialists. On the plus side, they can include part D drug coverage and extras like vision, dental, and gym memberships. Their premiums are also typically much lower than Medigap premiums, and can even be as low as $0 since they get subsidized by the government.

    Part D, added in 2006, is quite honestly an overcomplicated mess. There’s no government option; instead there are several private companies which can each pick their own list of drugs to cover and pharmacies to work with, though by law they have to cover a certain minimum number of drugs in each of several treatment categories. The basic concept of part D was that each year the plans would pay 75% of the user’s drug costs until they hit the Coverage Gap (aka the Donut Hole) at which point the user pays 100% until they hit the threshold to go into Catastrophic coverage in which they pay either 5% or a nominal copay (costs never go to 0, there’s no out-of-pocket cap) until the end of the year. Of course, most people don’t know how much their drugs cost and prefer set copays, so D plans usually divide their drugs into 4 or 5 price tiers and assign flat copays to the tiers where most of the commonly used drugs are. Note, however, that many plans also divide their pharmacies into “preferred” and “standard” and give you lower copays at preferred pharmacies. Plans can also choose whether or not to have a deductible before they start paying anything.

    The Coverage Gap gets its own special section as well, because getting in and getting out count different things. In 2018, you go into the gap if your total drug costs reach $3750, and come out into Catastrophic if your out-of-pocket costs reach $5000. That means that if you have a drug with a $100 price tag, and a $10 initial copay, the full $100 counts towards getting you in but only the $10 goes towards getting you out. Also, the Affordable Care Act included a plan to gradually close the Coverage Gap, so in 2018 if you go in the gap you’ll pay 35% of full price for brand drugs, 44% for generics, and both of those percentages will decrease again next year and hit 25% in 2020 at which point the gap is declared closed. And just to be extra confusing, 50% of the price of brand name drugs in the gap, which you’re not paying, is actually a mandated discount from the pharmaceutical company and gets counted towards your out-of-pocket cost to get out of the gap.

    So with all of that, if you take more than 1 or 2 meds and are trying to pick a drug plan, trying to find the best coverage manually is going to get real complicated real fast. Luckily, medicare.gov has an online estimator where you can plug in your zip code and your meds and it’ll give you a list of every plan available in your area with an estimate of how much you’ll pay and whether your meds will be subject to any restrictions.

    Everyone on Medicare can freely change part D and Advantage plans from now through Dec. 7, so if anyone here is on it or needs to help a loved one who’s on it, now’s the time to start looking at options. If anyone wants to know more, just ask!

    • Brad says:

      Thanks, good write up. One part I didn’t understand when I was looking into this for my parents last year was the underwriting cohorts for medigap plans. I came across some suggestions that it was a bad idea to get a low deductible F plan (I think, it’s been almost a year) because even if the premiums were affordable in the beginning they would get worse over time for some reason having to do with adverse selection–but that reason eluded me and I moved on. Any idea what that might have been about?

      • Thegnskald says:

        Speaking out of ignorance of the specifics: Assuming economically rational actors, I imagine people who don’t end up spending much will gradually move to higher deductible plans, while those who stay will be disproportionately those who spend more. This will become a self-reinforcing cycle, as each time a group of people leaves to save money on a lower premium plan, average premiums will rise, putting more people in a position where it is cheaper to get the higher deductible plan.

        • Brad says:

          But there are new people constantly coming in as they hit 65 and people dying. The part I don’t know the details of is how the underwriting groups are set up. Are their cohorts by age (and state?) within each insurer for each medigap plan (i.e. A, B, C, F-low, F-high) or does it work some other way?

      • SamChevre says:

        I know this from the insurer side, slightly. (I’ve never worked on MediGap, but I have talked with colleagues who do or did.)

        MediGap underwriting and pricing works like this*. Anyone newly enrolled in Medicare can enroll in Medigap at the stated premium, without underwriting. Anyone buying it later (including switching plans) can be required to pass underwriting.

        For a given plan and issuer, the premiums can increase for the entire plan IFF the plan costs are higher than the the expected costs.

        So what issuers do is have multiple legal entities; company A issues new plans, then eventually stops selling new plans and company B starts selling new plans, and so on. Once a company stops selling new plans, as costs on the pool rise, the rates increase as the costs increase. People who can pass underwriting buy a new plan from someone who is actively selling; people who can’t pass underwriting keep their coverage in the plan they have, but the premiums tend to increase substantially as the pool shrinks to be more and more heavily people who have high costs and can’t re-underwrite into a cheaper plan.

        I am not sure why low-deductible plans would have higher likely increases; I don’t know the distributions of underlying costs well enough.

        *So far as I know and remember. I’m an actuary, but this is not an actuarial opinion and etc.

        • Brad says:

          That makes a lot of sense. AFAIK F-low deductible is the most comprehensive medigap plan you can buy. If you can choose it without underwriting then all other things being equal the people that choose it are likely to be disproportionately the sickest. But that in turn means likely higher premium costs down the road because you’ve selected yourself into a pool likely to experience an adverse selection spiral. Sounds like an interesting game theory question.

      • Loquat says:

        No-deductible F (and C, for that matter, which is almost as comprehensive) is actually in an unusual situation right now, because Congress passed a law declaring that both of them were to stop enrolling new patients in 2020. So unless enough people yell at their Congresscritters to persuade them to repeal that, the pool for F and C will be locked down in just a few years.

        Why, you ask, was this law passed? Because both F and C leave you with basically zero out-of-pocket cost, so the theory is that this then enables F and C policyholders to use more medical care than they would have if they had to pay something out of pocket, thus costing Medicare more money, therefore taking away those plans would save Medicare enough money to be worth the trouble. They can’t just force the people already on F/C onto lower-coverage plans, though, because by law Medigaps are guaranteed renewable barring something drastic like the insurance company going out of business, and changing that would require much more legislative work than simply closing F and C to new enrollments. Some other plan letters (E, H, I, and J) were discontinued for other reasons a few years back, and it was basically the same deal where existing policyholders could keep them but nobody new could join. My company still has some of those plans on the books, even; some policyholders have those at rates comparable to currently available plans, others have seen drastic increases, and I haven’t seen enough of them to nail down a pattern.

  21. johan_larson says:

    I’ve been watching episodes of David Mamet’s 2006-2009 series “The Unit” about a special forces unit patterned on the Army’s Delta Force. The series depicts a small group of soldiers (“operators”) and their wives. The operators travel around the work, kill people, and blow stuff up. They live in an almost entirely male world. They are all men, and virtually everyone they encounter in the course of their work is a man too. Meanwhile, all the spouses are women, and their world is the base housing and their children. Some of them manage to work, but it’s a bit tricky because their husbands’ jobs come first and soldiers often have to move. And they don’t know much about their husbands’ work, because all the details are classified.

    It’s striking how the show is almost two completely different shows running in parallel. You could edit them that way and air them separately, aiming at very different audiences, of course.

    • Thegnskald says:

      If you like the format, Manhattan used something similar, at least in the first few episodes; didn’t watch very far into it, so cannot say if it continued.

      • johan_larson says:

        Mad Men had some of that going on too, but there the work lives and home lives were more connected since some of the female characters worked at Anderson Cooper.

        • Don P. says:

          Sterling, not Anderson. Although it took me a second to be centain.

          • johan_larson says:

            Thanks for the correction. I thought it sounded a bit hinky. I googled it and got pictures of a vaguely familiar white-haired man who looked just right for the cast of Mad Men, so I went with it.

            The world is a strange place.

        • Le Maistre Chat says:

          I’d watch a drama about the home and work lives of women who work at Anderson Cooper.

    • MrApophenia says:

      Is it good? I was always curious based on how much I like “Ronin” and “Spartan” but I never went back and watched it.

      • johan_larson says:

        I enjoy the male half of the show: men of action actioning all over the place. The female side is rather tedious.

      • Trofim_Lysenko says:

        If you liked Spartan and Ronin, you’ll like The Unit for at least part of it’s run. Those are two of my favorite movies and I managed to make it through the whole series, though I was pretty disappointed by the end.

        It starts out pretty decent, but goes off the rails as they exhaust the sorts of stories that will interest your average TV audience and have to start dredging up increasingly silly military narratives on the unit side and soap opera drama on the home side. It’s hard for me to delineate a hard “Stop watching after Episode X” point because there are some very good episodes even fairly late in the show’s run, but the problems appear well before the last episode I consider strong.

        Also, the limited budget shows in some really grating ways at times if you have some military knowledge. Like a bunch of Tier 1 hardcases tooling around in Waziristan (which looks suspiciously like the hills outside LA but whatever, that’s actually not too bad) looking exactly like this. I mean, ok, on a weekly TV budget you probably can’t afford this, but -I- could knock up more convincing costuming with a couple hundred bucks at surplus stores and airsoft outlets.

        • Randy M says:

          It’s interesting that, besides the accessories and headgear, one difference is that the ratio of beards/clean shaven is basically inverted in those pictures.

          I remember watching a couple seasons of the unit when it came out. Pretty good.

          • hlynkacg says:

            There aren’t many opportunities to shower or shave outside the wire.

            It was a common vector of ribbing between the spec-ops guys and expeditionary/support guys like me. They’d be rolling in with the machismo and we’d counter that we may not have the Gucci-gear or the luxuriant facial hair but we didn’t have to walk either. Have fun storming the castle boys, we’re going to Balad to get pancakes and a shower! 😉

          • keranih says:

            There aren't many opportunities to shower or shave outside the wire.

            Not arguing that, but it is my impression that the association of SF & facial hair is fairly recent, while procuring hot water and a decent razor anywhere on the planet has never been easier. While fads come and go, I do wonder if the heavy involvement in Afghanistan and the Mideast is at least in part driving this.

          • Another Throw says:

            @keranih

            My impression is that being clean shaven was adopted as a (perceived?) bona fida military requirement (needing to wear primitive gas masks, and sanitation when field conditions were expected to allow for shaving but not much more) when it was generally accepted for men to wear facial hair. Since this rubbed against the cultural grain, it (and a really shitty haircut) became a marker of military distinction at a time when the military was highly regarded.

            About the time the military justification had evaporated (improvements in gas mask design, etc) culture had shifted and professional men were expected to be clean shaven. Since the military was poorly regarded at the time, the justification was retooled to be that of a marker of the “professionalism” of the force. Facial hair on special forces was an unfortunate, necessary exception for operational considerations.

            The military has jettisoned the rest of the markers of “professionalism” by, for example, switching from the standard dress code of a suit and tie analog to wearing a combat uniform in an office thousands of miles away from the closest place that could reasonable require it. The military is, for the time being, still clinging by the nails to the shaving issue.

            Meanwhile, facial hair has recently grown from an unfortunate exception for operational necessity to a mark of distinction. [Wild speculation alert!] Since our operations in Afghanistan and Iraq have lacked(*) substantial, safe rear areas, combat experience has penetrated much further into the rear echelons than previous conflicts. Even those pansies in the maintenance company are at routine, substantial risk! (The assholes in the headquarters are still assholes, though.) Operating alone, surrounded by a hostile populace, with nothing but your weapon, your squad, and the seat of your pants. This is the stuff that for the last 50 years only the SF community has even been close to, and it is now just a part of military life. (Unless you’re in the headquarters. Asshole.) Of course SF does it harder faster better stronger than anybody else, of which their are justifiably proud, but facial hair has grown [sic] to be a status marker in a military where their accomplishments are less unique(*).

            (*) Relatively

          • johan_larson says:

            I suspect part of the special forces’ preference for beards is simply them taking advantage of being special, in that they can get away with violating some minor rules that others are expected to obey, and using that to mark their specialness in a way that is visible to all.

            If the military were to absolutely insist that the special forces be clean shaven, they could make it stick, at least on base.

          • CatCube says:

            The primary driver for SF and facial hair has been operating in countries where being clean-shaven is seen as effete. The Taliban used to jail men who didn’t have enough of a beard, for example. They’re primary purpose is to embed in with and train local troops, and they have to fit in to a certain extent to be effective in this.

            That’s not to say that the SF guys also like to keep it as a sign of being above the clean-shaven rules imposed on regular troops, as well, but the primary driver has always been effectiveness in the Middle East. AFAIK they don’t wear big facial hair when operating in South America or Asia, because those areas tend towards clean shaven, and they’d stand out.

            They do occasionally get into fights with senior officers over it. A while back there was a minor scandal with some of them growing straight-up Mountain Man Grizzly Adams hair and beards. While the ME *does* culturally require beards, the locals *do* actually groom and shape them. These guys were well outside of what would fit in.

            Finally, I recall a blog a while back (long enough ago that I can’t remember where) where an SF guy posted about some arguments they had with a senior (SF) officer–IIRC, assigned from a different Group with responsibility for a different area of the world. The senior officer demanded that they be clean-shaven, as required by regulation (there was no written policy exempting SF from those rules–I don’t know if there is one even to this day). The guys on the ground protested, of course, because they really can’t do their jobs if they’re not respected by the locals they’re working with. The senior officer started insisting they send in a group photo every week with a newspaper to show that the photo is current. They ended up taking one group photo and photoshopping in the front page for the week.

        • johan_larson says:

          I noticed they use colored filters to signal the location. Yellow for Africa/hot and blue for Russia/cold.

    • Well... says:

      The HBO miniseries From the Earth to the Moon was like that. Great show.

    • johan_larson says:

      “The Brave” is a current TV show on NBC somewhat like “The Unit”, but they don’t deal with the characters’ home lives. It’s all job all the time.

      Interestingly, they chose to include a woman in the main team. “The Unit” was all male, and the somewhat similar current show “SEAL Team” only has women among the support staff.

  22. Universal Set says:

    Is the open thread a good place to get career advice? Let’s find out!

    My situation:
    I’m tenure-track math professor at a nonselective SLAC in the midwest. I ended up at my institution in part because I burned out on research in grad school (I still like working on hard problems; it’s the part where I have to produce publishable material on a deadline and be judged on it that’s hard on me; I had some bad experiences early in grad school that kind of messed me up). My pay is low even for academia (about $45k as a 9-month salary, with a few thousand more available if I teach a summer class online). I have very few math majors (currently a grand total of 6; that’s 6 total, not per year) and most of the other students on campus are math-phobic. I’m concerned for the future of the major as well as of my institution, as undergraduate enrollment has been in a decline. Low enrollments in the major mean small classes, which means that I’m required to teach a heavy load in terms of total credit hours (currently averaging 14 per semester), and I’m feeling a bit exhausted as being early-career means lots of preparation time. On top of that, there’s no guarantee I’ll get tenure anyway. I also don’t feel like I’m using my talents well in my current role.

    On the other hand, I like my location: I’m in a small town, live a 5 minute walk from my job, have a vegetable garden in my yard, and own a cozy little house with my wife. I do like sharing mathematics with other people, so there are parts of teaching that I very much enjoy. I also like my colleagues; many of the professors at my institution are great people. I’m not eager to move again right now. If I do end up leaving eventually, I don’t think I’d want to stay in academia. The low pay is an issue, but not an immediate problem; my town is low-cost-of-living and we’re pretty frugal. So I’m not sold on trying to leave my current job yet.

    What I’m looking for:
    I want to find an activity/occupation with which I can (a) use my talents and problem-solving skills in a way separate from teaching, so I can stretch myself mentally and not feel boxed in and like I’m an underachiever wasting my potential, (b) make some substantial extra money over the summers to shore up my low salary, and (c) develop skills for a second career in case I do decide to (or am forced to) make an exit from my current position. Also, I’d need to be able to do this without moving. This is a pretty limiting set of requirements, so I’m willing to call it a win even if requirement (b) takes some time to get off the ground.

    My skills/talents/credentials:
    I have a PhD in mathematics. My field was in pure mathematics (computability theory, though I started out as a number theorist so I know a little bit of that too), not applied. I have some reasonably impressive contest-math accomplishments from high school and undergrad (USAMO top 12, Putnam HM) but nobody cares about that once you’ve finished undergrad. I’ve also done some programming – ACM ICPC team in undergrad, and over 300 project Euler problems in my last year of grad school while procrastinating on my PhD thesis – but don’t have any experience with actual software development. I’m a decent writer. Beyond that, I’m not sure what counts as relevant skills; I’m sure there are some, but I don’t have a good gauge of what sorts of things are actually valuable.

    I’m looking for (1) good ideas for where to start looking, (2) sober thoughts on my prospects in this endeavor or outside academia in general, and/or (3) some actionable first steps. I figure that SSC readers include a bunch of people with similar talents, but who have made more effective use of them than I have, so maybe some of you will have some good ideas.

    • johan_larson says:

      If you are looking to change careers, software development would be a good choice, particularly since you say you can already program. Assuming you are willing to move for work, you might hope to triple (!) your income if you can snag a job with a major software company in a major center (San Francisco area, New York, Seattle). As a PhD in math, you have a very impressive credential, and the big dogs (Google, Amazon, Facebook, Apple, Microsoft) would all be interested in interviewing you.

      You do need to prepare. All the majors put their candidates through grueling interview routines requiring coding on whiteboards, and you need to be ready for that. These interviews aren’t particularly like actual software development, but you need to pass them to get hired in the first place.

      Working your way through an algorithms and datastructures book would be good preparation. There’s a lot of such books. This one isn’t one of the classics, but looks sound:

      https://www.amazon.ca/Data-Structures-Algorithms-Michael-Goodrich/dp/0470383275

      If you want to go a step farther, some familiarity with concurrent programming would be useful. This book looks decent:

      https://www.amazon.ca/C-Concurrency-Action-Anthony-Williams/dp/1617294691

      For more about how these interviews work, try this book:

      https://www.amazon.ca/Cracking-Coding-Interview-Programming-Questions/dp/0984782850

      If you know anyone at these companies, it would be useful to get them to nominate you through their internal HR system. Things will go faster. If not, just apply directly to an entry-level software development job and wait. It’s possible that HR will steer you towards mathier, more analysis-oriented jobs because of your background in math. If so, I advise you not to fight it; they are steering you towards your strengths.

      If pure software development isn’t quite your thing there is a related profession called data science. Data scientists are mad hybrids of business analysts, computer programmers, statisticians and AI experts who focus on finding patterns in large data sets. But I expect someone else in this forum can give you better advice about how to join their ranks.

      • James says:

        Seconded, but I would just emphasise that there’s life outside the Big Four (or Five, or whatever), too, and their interview processes tend to be less gruelling than what johan_larson describes.

        Alternatively, you may be able to pick up freelance programming work in your summer months. Web development is reasonably well-suited to freelancing. One route to that: learn a web framework (probably either Rails (with Ruby) or Django (with Python)), make a website or two to show potential clients that you can do what they’d be paying you to do, and then hustle hard until you find clients.

        Worth considering, but I can’t guarantee that it would work for you or that you’ll necessarily be able to find work that suits your particular life situation!

        • johan_larson says:

          The problem with part time work in the OP’s situation is that the dash for tenure is already at least a full-time job in itself. Taking on additional responsibilities — such as a summer job doing software development — lowers his (?) chance of making it. I think he should commit fully one way or the other, either fighting for tenure or getting out of academia entirely.

          My post assumes he commits to getting out.

          • Universal Set says:

            This is a good point about the “dash for tenure,” but that’s actually less of an issue at my institution. The requirements for research are not terribly high, and as far as I know are mostly a binary “are you still doing publishable research” rather than scaling with number and quality of papers. Much more of the decision is based on teaching (including the dreaded student evaluations), reflecting the high teaching load. This is why I feel uncertainty about tenure, but think that doing something else over the summer is plausible. (I suspect that I could even sell my department chair and dean on the idea that doing this sort of outside work over the summer would help me to advise students and better prepare them for the workforce, and my chair is on board with the idea of my spending the next summer mostly developing skills related to programming or data science.)

      • Drew says:

        I agree. I left academics for tech. I’m very happy with the decision.

        That’s an excellent list of recommended books. I’d add https://leetcode.com/ to the mix. It collects questions from the big firms.

        The preparation is kind of obnoxious, but you eventually realize that the interview questions are variations on a handful of themes. I’d guess it would take about 120 hours of prep to pass the coding exams.

        If you go this way, I’d recommend posting an occasional followup. Quite a few tech-employees read SlateStarCodex and I suspect you could get help with interview prep or even recommendations.

        • Universal Set says:

          I’d love to hear more about your experience leaving academia for tech. Would you care to write some details about your story? (If you don’t want to share in a public place, that’s OK, or I can give you my email if you’d like.)

      • Universal Set says:

        Thanks for the information! I’m honestly not keen on moving to a major city, which I know cuts down my options quite a bit. Madison, WI was fine (I did grad school there) and I wouldn’t mind living in a suburb in e.g. research triangle, but NYC / SFBA are really not my idea of a good time. If you happen to know what prospects look like outside of the major centers, I’d be especially interested.

        I should mention that I’m not ready to jump ship entirely just yet. I’m more in the explore-my-options mode at the moment; my current plan is to spend time this coming summer on developing skills and exploring my options, and re-evaluate in Spring 2019. But advice like “It would be a good idea to work through a book on concurrent programming if you want to switch to software” *is* the sort of thing I’m looking for, so thanks.

        • johan_larson says:

          You can find software work in many places, but it does tend to be a big-city phenomenon. If you really want to live in a small town, it would be best to work in a bigger city in your first job, and then look for remote from-anywhere work for your second job.

          This article does an analysis, and comes up with a list of reasonable places I’ve heard good things about.

          https://techbeacon.com/10-best-cities-for-software-engineers

          Denver and Austin are particularly promising, and Austin is quite small.

          • Universal Set says:

            It’s encouraging to see Raleigh on those lists; I have relatives who live in Wake Forest and work in Raleigh, and it’s quite possible to have a more small-town living environment there (unless things have changed more than I think in the last 10 years).

        • professorgerm says:

          In relation to the below comment, you’re still correct ten years later: Raleigh still provides a small-town feel within a reasonable commute. I moved here last year and plan on staying for some time; it’s a good balance of city amenities and being not-too-far from both the beach and the mountains. However, I’m in biotech and not computers/academia and can’t say much for those job markets.

    • rlms says:

      On the same topic: does anyone know which Silicon Valley companies sponsor visas for interns?

      • Brad says:

        For students in American colleges on F-1 visas they can work both during summers and after graduation on optional practical training for a total of 36 months (for those working on STEM degrees).

        The visa most appropriate for other types of interns would probably be a j visa. The state department maintains a list of sponsoring organizations here: https://j1visa.state.gov/participants/how-to-apply/sponsor-search/

        Keep in mind that some employers may go through one of the listed sponsors rather than sponsoring aliens directly. Also keep in mind that some J visas have a two year home‐country return requirement before you could reenter the US on a different visa.

        • rlms says:

          Thanks! Yes, I would need a J (presumably -1?) visa: I’m a UK citizen/resident.

          • Brad says:

            Yes, J-1 for the primary applicant. J-2 visas are for spouses and minor children of J-1 visa holders. Unlike some other visa categories, J-2s can generally get employment authorization.

            The UK does not appear to be on the Exchange Visitor Skills List so the two year requirement would only apply if the US government, the UK government, or certain international organizations funded the internship, which it doesn’t sound like the type of program you are looking for.

            Googling around a little bit it seems like a lot of the big SV employers work with these guys: https://culturalvistas.org/

    • SamChevre says:

      I’m an actuary. I’ve had numerous colleagues who had been academic mathematicians.

      The qualification process is painful, but not as difficult as getting a PhD. The work can be very different depending on your role–it can be anything from “basically software development” to “basically explaining things clearly”. If you email me–this username at gmail–I can try to connect you with colleagues who have made that switch and can provide more actionable information.

      • Universal Set says:

        That’s interesting to hear – I wouldn’t have expected that many academic mathematicians to make the switch to being an actuary. You mention that the qualification process is painful; is there more to do to get one’s foot in the door than passing the first few exams?

        • SamChevre says:

          Not really, but it has been a long time since I was at entry-level (and the foot-in-the-door requirements have gone up). You might look at the data science field within insurance; with a couple exams and a math PhD, I would imagine that you’d be a very strong candidate.

  23. Scott Alexander says:

    Does anyone who attended the Berkeley meetup have any comments or suggestions for the next one?

  24. Emerald says:

    Part 2 of “does anyone want to meet up in Boulder, CO?”: if you’re interested, please send an email to slatestarcodex.boulder [at] gmail [dot] com and/or join this Facebook group. We could have regular meetings on, say, the third Tuesday of every month (to complement the Denver group), but I’d also be up for climbing/hiking/biking/walking/skiing/etc. with other SSCers.

    • ProfessorQuirrell says:

      As a graduate student with very little free time, I am tentatively interested in this.

  25. JRM says:

    You’re not allowed to argue quantification of “beyond a reasonable doubt,” in criminal trials for the very good reason that we have human jurors who are bad at it.

    What should the percentage be, though? (This is a different question than the N Guilty Men question of how many innocent people we can stand convicting for each guilty person convicted.) Should it vary by crime? The standard does not vary by crime now.

    • James says:

      (This is a different question than the N Guilty Men question of how many innocent people we can stand convicting for each guilty person convicted.)

      They’re different numbers, but I think we can derive one from the other, at least if we know (or are willing to guess) what fraction of the population is guilty.

    • Drew says:

      Base-rates make the problem really, really weird.

      The DA gets a report that a doctor is having patients die of increasingly improbable drug interactions. People come in with a cold. They’re dead within a week. It keeps happening.

      The DA contacts a statistician. The statistician proves that “P < 0.00001" these aren't random chance. Prosecutor accuses the doctor of murder. Do we convict?

      The problem is that the doctor is claiming to have a 1-in-100,000 run of bad luck. That's unreasonable. Until we realize that there are ~1 Million doctors in the US, so about 10 of them should have bad luck on that level, per year.

      If we'd discovered this stat after we'd started suspecting the doctor (perhaps doctor was accessing odd or inexplicable drugs), then it would be damning evidence. But the bad-looking-number is a big reason that the charges were brought in the first place.

      So, our threshold can't be, "P(innocent | evidence)" but needs to incorporate the chance that someone was charged.

      But this itself gets weird. It implies that, if prosecutors are selective enough about charging people, the mere fact that someone was charged should be solid evidence for their guilt.

      • Aapje says:

        But the bad-looking-number is a big reason that the charges were brought in the first place.

        So, our threshold can’t be, “P(innocent | evidence)” but needs to incorporate the chance that someone was charged.

        Exactly: the statistics we use to calculate the likelihood that a person is guilty must not be the filter we (directly or indirectly) use to decide to charge that person.

        Or to put it another way: we can’t use a base rate that is valid for group A if our filter results in a subset of group A with a different base rate.

        Another example makes this much more clear. Oncologists have many more patients die on them than ophthalmologists. If we use the base rate for the average patient who seeks healthcare, then we can statistically ‘prove’ that oncologists are a bunch of murderers, while ophthalmologist appear to be superb at curing patients. So then we could choose to replace all oncologists with ophthalmologist and have the latter try to cure cancer.

      • AlphaGamma says:

        There was an infamous case (or series of cases) in the United Kingdom where several women were wrongly convicted of murdering their children based largely on the testimony of a single expert witness, Professor Sir Roy Meadow, who had written a book stating that “one sudden infant death is a tragedy, two is suspicious, and three is murder until proved otherwise”.

        In the highest-profile case, that of a solicitor named Sally Clark whose two sons had both died before the age of three months, Meadow took the probability of a sudden infant death in an affluent non-smoking family like the Clarks and simply squared it to give the probability of two such deaths in the same family as 1 in 73 million. Obviously, this fails to take into account the possibility of a common environmental or genetic factor.

        Clark spent three years in prison for the murder of her sons before being freed on appeal. She died a few years later, of alcohol poisoning caused by psychiatric problems brought on by her experiences.

        Meadow lost his licence to practise medicine, though it was reinstated on appeal. Shortly after that, he retired.

        • rlms says:

          I think the more relevant issue in the Clark case is that Meadow didn’t take into account the fact that the prior probability for a (double) infant murder in a general family is also very low. Even if he’d properly adjusted for non-independence, it would still have been wrong to take the adjusted figure as the probability Clark was innocent. He should have used Bayes’ rule: P(innocent | deaths) = (P(deaths | innocent)P(innocent))/(P(deaths | innocent)P(innocent) + P(deaths | ~innocent)P(~innocent)) which is (P(deaths from SIDS)*P(not a murderer))/(P(deaths from SIDS)*P(not a murderer) + P(murderer)) not just said that P(innocent | deaths) was P(deaths). With that method, even taking the incorrect estimate of 1/73 million for P(deaths from SIDS) and using the UK murder rate as P(murderer) (an overestimate, since there are fewer murderers than murders, women are less likely to be murderers, and children are less likely to be victims) gives a vastly different probability of innocence of 1/670.

  26. AutisticThinker says:

    The Moral Economy: Why Good Incentives Are No Substitute for Good Citizens by Samuel Bowles

    I would like to thank Wrong Species again for recommending this book. I read it and it partly changed my mind. It seems that a society of sociopaths can never work well regardless of what kind of laws that exist.

    Harmful forms of ethics (e.g. ISIS, Nazis, Communists, etc) remain a problem. However having no ethics and solely relying on self-interest does not seem to result in a good society either.

    Moral indoctrination is not a good solution either because it hampers reason and promotes conformism in an already very conformist species.

    Is there any way out?

    • gph says:

      Genetically modify all future humans to be so pro-social that causing harm to another human would be unthinkable.

      Of course a mutant sociopath might spring up and have the world for their taking

      • Gobbobobble says:

        Genetically modify all future humans to be so pro-social that causing harm to another human would be unthinkable.

        In From the New World, they did so and then some. Causing harm to another human can literally kill you.

        Highly recommend the show. Your 2nd paragraph may or may not be relevant 😉

        • Heinlein has a throwaway line, I think in Starship Troopers, about a past attempt to breed aggression out of the human race. The small minority who didn’t agree to go along were exiled to Madagascar. The result was obvious–the wolves ate the sheep. The Madagascar confederation conquered the world.

          (By memory, so not verbatim)

          • Nancy Lebovitz says:

            Also, fictional evidence at best.

          • Thegnskald says:

            Historically, civilized people tend to do better at war.

            To substitute one just-so story with another, pro-social behavior makes military units more cohesive and effective.

          • Aapje says:

            Perhaps even more importantly, pro-social behavior tends to result in better logistics, so armies actually get food delivered to them, rather than to have to forage for it (which limits the size of the army to the food they can forage from their immediate surroundings).

            Then again, the Mongols were extremely effective, while not being that civilized.

          • Thegnskald says:

            The Mongols get a bit of an unfair reputation, historically – a bit like the Vikings, actually.

            They were extremely effective at government, and their society was, for the era, quite progressive. Their major fault was the absence of a counterbalancing force to the government, which meant that when succession was in question, the only way to answer it was war.

          • hlynkacg says:

            Quibble.

            Civilized != unaggressive.

          • Aapje says:

            @Thegnskald

            They thought that they had a right to rule the world and just kill everyone who objected. They looked down upon both farmers and city folk. Their governance was extremely tribal. One of their war campaigns ended because a Khan died and all tribal leaders had to meet up to vote for a new one. That is the model that many tribal people used or used, but that simply doesn’t scale to large societies or a large empire.

            Also, most of their wealth seems to have been built around simply demanding tribute, not actual governance of their conquered territories.

            None of that seems very progressive or civilized to me.

          • Gobbobobble says:

            Also, most of their wealth seems to have been built around simply demanding tribute, not actual governance of their conquered territories.

            Ancap equivocates this to taxation in 3… 2…

          • Thegnskald says:

            Aapje –

            Which empire didn’t think it had the right to rule the world?

            And yes, the Mongols had taxation. They had codified tax laws for peasants, which was particularly important in the region which became Russia, where previously taxes were in fact tribute (whenever local rulers needed funds, they demanded them). This was a Big Deal.

            Trade laws were likewise universalized, coinage (as well as paper currency) was minted, military defense was provided for, police forces were patrolled for, roads were constructed, canals were dug, academies were established, trade routes were created and defended…

            What exactly is the metric of civilization that the Mongols don’t meet?

          • Aapje says:

            @Thegnskald

            Which empire didn’t think it had the right to rule the world?

            The second part of my sentence was where the meat is. The usual modus operandi was that the Mongols would demand tribute and if not granted right away, they would massacre the people, burn farmland, burning cities, etc. In Iran and Iraq, they destroyed the extensive irrigation systems, setting those regions back centuries in agricultural productivity.

            If they did take prisoners, rather than kill people right away, they often used them as human shields in battles.

            Fundamentally, their war strategy was built on terror: submit or die; in a much, much more brutal fashion than pretty much all empire builders. The Romans were not that brutal, aside from the destruction of Carthage, which the Romans saw as an existential threat when Rome was almost defeated in the 2nd Punic War.

            And yes, the Mongols had taxation.

            They preferred tribute though, which means that they didn’t take real administrative control over many regions. That is again different from the Romans, who set up administrative centers in their conquered territories, exporting their engineering, culture, etc.

            The Mongols made use of a lot of expertise of the regions they conquered, but I don’t think that they civilized other regions like Rome did. In fact, the Mongols were very wary of becoming too civilized and losing their military edge.

            Anyway, the tribute demanded from Koryo (part of what is now Korea) included many child and artisan slaves, which is not very progressive. The Romans in principle only forcibly enslaved those who resisted. So I consider the Romans a lot more progressive when it comes to slavery.

            Romans: submit and pay taxes or be defeated and be made slaves
            Mongols: submit and have many of your people be made slaves or be defeated and die

            What exactly is the metric of civilization that the Mongols don’t meet?

            They were not so civilized in the same way that mafioso whose income mostly consists of taking protection money are less civilized than Mark Zuckerberg who actually built up something of his own by making management decisions to defeat others on a fairly level playing field, rather than building his business on violence.

            Of course, Rome and other empires were a mix of mafioso and Zuckerberg, but Rome is a lot more Zuckerberg, while the Mongols were way more mafia-like.