Statistics

Don’t be so sure

A number of mixed items today, mostly with the theme that Experts are often too sure of themselves.

  • The organization GRASP, among many others, until yesterday warned of the “imminent extinction faced by gorillas” and other primates (not humans). NASA, an organization of experts, has a page called “Gorillas in the Midst of Extinction.” They used sophisticated, powerful, high technology satellites to count gorillas “giving scientists and conservationists” a way to count gorillas. The phrase “scientists and conservationists” must mean there is a difference between the two types of creatures. Anyway, the previously (?) communist magazine New Scientist recently had an article called “Ebola pushes gorillas towards extinction” (in the late 1990s there were several books published warning of the same fate for homo sapiens sapiens).

    And then yesterday came a report by a group that unexpectedly came upon a troop of about 125,000 gorillas in the Congo, which more than doubled the previous estimate of the number of gorillas alive. Jillian Miller, the director of the conservation group Gorilla Organization, shockingly admitted (quoted in today’s New York Post), “I think the lesson for conservationists today is that, yes, the world is full of surprises. There’s a lot of uncharted territory.” I wonder if she’ll still feel the same way during the next round of fund raising.

  • “Bubble fusion” researcher Rusi Taleyarkhan‘s research was burst at Purdue this past week. This is the guy who claimed in 2002 he could induce fusion using the force of collapsing tiny bubbles (the learned word for bursting bubbles is cavitation). The claim was always silly, which is fine, because there are more than enough silly ideas that pass for “research” in academia. The press and others originally bought the idea, however, and surely there will be some people who will always believe, just like there are still some who tout cold fusion. But the claim was too silly for some, who were angered by Taleyarkhan, and they sought to punish him.

    This week’s Science magazine has an article (subscription required) on how Purdue is castigating Taleyarkhan. They suspected he fudged his data, but couldn’t prove it, so like the feds with Al Capone, they got him on a technicality, a move that I hope they are not proud of. Turns out that Taleyarkhan wanted a second author on a paper so that the paper would appear stronger: supposedly, more authors means less likelihood of cheating. So he showed the paper to a graduate student who made changes and recommendations, and then Taleyarkhan put the grad student’s name on the paper. Bingo! Research misconduct! cried the judges. Well, maybe, but if so, then roughly 98.3% of all academics are guilty of the same crime. People often, for a host of reasons, politics, fear, friendship, tit for tat, habit, and on and on, put names of people on papers even though those people had little or nothing to do with the work. Ah well. Poor Taleyarkhan.

  • For fun, we have a list of the Top 30 Failed Technology Predictions from the List Universe. Here’s #2, from Mr Bill Gates, a well known rich person who lives near Seattle: “We will never make a 32 bit operating system.” And #8 from Lord Kelvin, who was a mathematician and physicist, and president of the British Royal Society, 1895: “Heavier-than-air flying machines are impossible.”

    Ho ho ho, we say to ourselves when we read these prognostications. How stupid can they be! We experience mirth. But that is exactly the wrong emotion. You might despise Bill Gates, but he is an incredibly bright person, an expert among experts in his field. Kelvin, who you probably haven’t heard of, was one of the smartest people who ever lived (not at the top of the list, to be sure, but ahead of all of us). These, and the other people with quotes on the List Universe page, were masters, yet they made remarkably huge mistakes.

    You must also remember that when these men, superior in perception to their peers, made these predictions, there were not hosts of others saying the opposite. Most people believed the predictions, and with good reason. These experts had often been right before. What we should take away from this list is an increased skepticism, a belief that experts are not nearly right as often as they’d like us to think they are. Doubt, therefore, is the proper emotion.

Categories: Statistics

21 replies »

  1. Matt:
    Can we equate “Experts” to major league batters? Hitting major league pitchers at .300 is very good. Most of us would be far less than .100 with many .000, don’t you think?

  2. My favorite prediction was when ATT forecast a small market for cell phones and decieded not to go into that business because it wouldn’t be profitable.

  3. Matt:
    There must be empirical studies of experts involved in making multiple predictions: Stock market pundits come immediately to mind. Is there any pattern of % correct predictions among such experts? Is there a reason why they would be different from other experts?

  4. I agree with the platitude at the end. But is there really any evidence that Bill Gates is particularly smart or highly regarded as an expert on technical matters?

  5. James,

    Before I answer, let me tell you that I am solely an open-source guy. I’ve been using Linux for years, for example. I believe software should be transparent.

    Saying that, oh my yes, there is more than enough evidence to show Bill Gates is a smart man. Go and listen to TWIT.tv’s “Bill Gates- Not Evil” broadcast for a fascinating panel discussion.

  6. Check out Phil Tetlock’s “Expert Political Judgement.” In a word, there isn’t any, but some nonexperts are worse than other nonexperts.

  7. Can you name one major software innovation that Bill Gates created?

    He bought MSDOS because he needed something for IBM and his company only was selling a Basic compiler(interpreter). That, his ability to keep the rights for non-IBM machines, and a ruthless destruction of any competition is what made his billions.

    Word, Excel, powerpoint, Windows are all all copies of other peoples work.
    (Windows came from Xerox, and later Apple, Excel from Visical/123, Word from Wordperfect)

    Networking? Appletalk.

    He’s been caught stealing outright (DiskDoubler won a suit against Microsoft)

    Granted he’s a great businessman (like Rockefeller, or Vanderbilt) and incredibly smart but innovator?

  8. What equipment were the gorillas using? Do you think they used the same tactic on the scientists and the conservationists?
    Maybe this was the folly of the humans but I like to think the gorillas got one up for once. Never underestimate animal cunning.

  9. Doubt is the only emotion allowed in scientific method. I’ve listened to PZ Meyers’ interview at a local radio station today, and when asked why should people have “faith” on scientists, he answered that they shouldn’t, that people should always audit what other people are saying, that skepticism was the foundation of science. Well, he didn’t say exactly on those words (I’ve a terrible memory for details) but that was the gist of it.

  10. The story about the endangered gorillas was published in the Journal of Montreal today.

    Anyone want to bet that they’re will no mention of the 125,000 gorillas found in congo.

    And of course, they are endangered because of man activities, nature has nothing to do with it.

  11. Silvain:
    No mention because the gorillas have it covered. They’ve gone to ground again.

  12. If, when presented in front of the emperor, you notice that he’s not wearing any clothes, would you dare make mention of it, since no one else was, or keep
    Quiet and go with the crowd. This is the difference between someone who questions the authority and someone who thinks this is wrong or unacceptable.

    Take the analogy of the valley people or the hill people. Some are content to stay where they are and accept that there’s nothing beyond their own world. Others like to go to the top of the hill and see for themselves. Maybe there’s nothing wrong with accepting the givensituation at face value but I simply can’t relate to it.

    In my field, not to “doubt” would be dangerous, as it’s one of the fail- safes in the system of medical care. Did you ever wonder, if you’ve been to a hospital to different departments why everyone seems to ask the same questions over and over? I know many have because people complain about it frequently. I’ve known situations where a patient has not been ask the pertinent question from the start and so is receiving the wrong or no intervention as a result. So whether the “expert” is higher in authority or a peer it’s simply good practice to ask the obvious or question someone’s reasoning. Anyone worth his salt would not take offence

    That is the culture that should exist in any scientific field where the stakes are high. There ought to be no place for gurus. Otherwise science is fallen to the level of modern art, where none shall question its merit; or religion where none shall question the doctrine.

    Nullius in verba, I believe, should be the order of the day. This is the motto of the royal society, lets hope they will wake up some time soon and remember.

    Who will be the one to topple the house of cards that is anthropogenic global warming? The stakes are high and so is the state of the science. My money’s on the gorillas.

    Hope springs eternal!

  13. May I add that one way to resolve a doubt in science is to have an open mind? Unfortunately, the phrase open-minded (Republican) skeptic seems to be an oxymoron sometimes… Sorry, not funny at all, how do I strike out the word “Republican”? ^_^

  14. Bill Gates is a technical expert. From 1000s of ideas about technology direction he directed the selection of those that hit it pretty big. There are millions of technical ideas floating through corporate head space. The companies that select the right ones and apply business savvy succeed. Additionally, it is a well known technology sector fact that 15% (pick a number) of your projects will fail; if you don’t have enough failures, you aren’t generating the output needed to find the key successes. That is how you manage complexity instead of fearing it! You must make mistakes.

  15. jhon, what does politics have to do with science? You could try the delete key before pressing submit, that would work.

  16. Good post. On the listing of authors on papers, a few years back a global warming paper appeared, the lead author of which is a very well known global warming alarmist. The paper was just another attempt to promote the anthropogenic global warming theory. I downloaded the paper and was surprised to see the name of a well known global warming skeptic listed as a co-author (along with many other co-authors). I contacted the “skeptic” to ask him about his involvement with the paper. He responded that he had never seen the paper and had no idea how his name ended up on it. I’m not sure what, if anything, was ever done about it, but I found it astounding that a scientist would list another scientists name on a paper without the latters knowledge.

  17. A formal paper has recently been published proving that our kidneys simply cannot handle the recommended dose of 1.5 litres of water per day. This dosage has of course been standard doctrine in every health pamphlet for a long time. The original study was flawed by bad calcs, bad assumptions, poor data and bad stats but it still managed to dominate because nobody questioned it. Crucially though, virtually nobody ever drank 1.5 litres per day, most certainly not the health professionals who parroted this line. The few who did actually try it ended up in hospital or the morgue. But compare the gross herding of the intelligentsia, who are happy to promote a message without apparently even thinking about it’s rather obvious absurdity, with the general population who instinctively recognize BS even when coming from a trusted source. We just assumed nature knew best and so we just drank when we were thirsty.

    I await a paper telling us that eating 5 fruits/vegetables a day in nonsense too. But I also await a paper which finally questions the practice of putting 8 diseases in a single child’s injection: if this practice turns out to be ok then it’ll be nature correcting for man’s cost-cutting stupidity once again – like for example with the BSE debacle.

  18. JamesG,

    1 litre (liter) of water is 1.056688 US fluid quarts, or 33.8 fluid ounces. 1.5 litres is about 50.7 ounces.

    The drinking glasses in my cupboard we use most frequently hold about 13 ounces when filled to convenient level. 4 glasses x 13 ounces = 52 ounces. I drink somewhat more water than that every day, and certainly even more if I am doing chores around the outside of the house on a hot day.

    Those who work outside in hot and especially dry climates can drink gallons per day. While a lot is lost to perspiration, if you stay well hydrated a substantial amount is processed by the kidneys.

    I have not seen the paper you mention – are you saying that a person cannot urinate 1.5 litres per day without ending up dead as a result?

  19. The Vaccines do not, in the main, contain live microbes. I am not sure about eight in one, but I had five in one arm and three in the other when I travelled to India. It is not the same as subjecting an individual to eight diseases at the same time. Some vaccines cause temporary effects for a few hours but any interactions of these affects are rigorously tested prior to their inclusion in the same vaccine. The effect is also not the same as having a mild case of the disease, but is usually a generic immune reaction.
    Just imagine how many live bacteria and viruses the human body fights every day. It can handle a response to eight new ones that are dead or broken up before injection. It’s a bit like being allowed to see your enemy’s weapons and action plan so you can be ready with a bullet-proof response without actually being subjected to the forces of your enemy.
    It’s the only moral form of cheating.

Leave a Reply

Your email address will not be published. Required fields are marked *