Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Google Businesses The Internet

Google Putting Crowd Wisdom to Work 190

Posted by CmdrTaco from the just-don't-start-a-stampede dept.
daveperry writes "The Google Blog has a post about their use of prediction markets to forecast certain events that are relevant to their business. From the article: "Our search engine works well because it aggregates information dispersed across the web, and our internal predictive markets are based on the same principle: Googlers from across the company contribute knowledge and opinions which are aggregated into a forecast by the market. Sometimes, just feeling lucky isn't enough, and these tools can help." In related news, some software was recently open sourced that enables people to set up their own prediction markets."
This discussion has been archived. No new comments can be posted.

Google Putting Crowd Wisdom to Work More

Google Putting Crowd Wisdom to Work

Comments Filter:

  • See Here (Score:3, Funny)

    by XFilesFMDS1013 ( 830724 ) on Thursday September 22, 2005 @08:36AM (#13620891)
    use of prediction markets to forecast certain events that are relevant to their business.

    Nothing to see here....hmmmm...bet they didn't see that one coming.
  • I just recently started participating in it. Doesn't seem too complicated, though I'm still a bit unclear as to the relevancy of the predicitions it generates. It seems to me that for this sort of thing to work right, you'd need a much larger sampling (which maybe Google is hoping to get), but then, maybe I just don't know what I'm talking about?

    • Re:Sounds Like Owise.com (Score:5, Interesting)

      by drangundsturm ( 909630 ) on Thursday September 22, 2005 @09:42AM (#13621475)
      Owise.com is actually a lot different in underlying mechanism, making predictions by blending opinions like a nerual network does.

      Our philosophy is: why bother with all the trappings of the "market". It just confuses people, and leads to all kinds of gamesmanship that has nothing to do with what you're trying to predict. We simply ask people what they think the outcome will be, and we use a mathmatically correct way of ranking their accuracy (called a "proper scoring rule" for those of you who like math).

      People who are right get higher weights in our system, can win prizes like Amazon gift certificates, and gain "titles" such as "Senior Political Analyst, Level 3". Bloggers can use our data graphs free, just don't nuke our watermarks. The system has already been more accurate than experts at some predictions, although we are just starting out and need a lot more people.

      • People who are right get higher weights in our system...

        Out of curiosity -- do you have empirical evidence that weighting people's guesses on their track record gives better accuracy than equal weighting does? It's not obvious to me that that would be so.

      • I just joined, as I've designed something very similar in my head and it is my policy to join/buy/participate in such things when I see them.

        Per the notes in my other comment [slashdot.org], you should probably shut off feedback amoung the players before they choose a ranking. In this case, you should consider not automatically showing the current consensus of the site to a person making a prediction. Show it after the person makes one, or allow them to click through to it if they are interested with no opinion.

        Much bette
      • Owise looks interesting... at least as an entertaining way to kill time, until I can find out more. Unfortunately it looks like only the front page of the site is alive and everything else is down.
      • Owise.com looks very cool! I had a similar idea, which I called MemeTrade.com [memetrade.com]. (get it? rhymes with E*TRADE :) I wanted to create a "fantasy stock market" community similar to BlogShares, where people could busy and sell "memes", i.e. word frequencies from the many news and blog sites that my memetrade web crawler would crawl every 15 minutes. For example, the meme "RITA" might have a rising value today as more news and blog stories are written about hurricane Rita, while other words would be losing value a

  • Yay Delphi. (Score:3, Informative)

    by Anonymous Coward on Thursday September 22, 2005 @08:38AM (#13620907)
    And once again John Brunner wins. Can't believe the Shockwave Rider was written in 1975... compensated abstinence zone in New Orleans anyone?

  • Awesome Mindpower! (Score:3, Funny)

    by mfh ( 56 ) on Thursday September 22, 2005 @08:38AM (#13620909) Homepage Journal
    We also found that the market prices gave decisive, informative predictions in the sense that their predictive power increased as time passed and uncertainty was resolved.

    This is why I like Google. The use of intelligence to develop accurate results in a predicative system, and keep it all flowing -- it shows not only wisdom -- it shows an early level of omnipotence, which has to be the key ingredient to success today! If God is supposed to be omnipotent, why not try it? (haha I'm not a religious nut, FYI... just like to use the data available)

    Consider the alternative solution to success [slashdot.org] and you really must put your investment dollars where you have the most faith. Google stock can only go up, thus breaking the law of Gravity. Only a supra-genius (Wyle E. Coyote) knows how to bend the laws that govern market economy to their favour. Are you really going to bet against that kind of mindpower?

    Being geeks, we naturally used information theory to measure the entropy of our probability distributions:

    This is a demonstration of wisdom. Knowing the rate of market decay is a HUGE BENEFIT for all Google stockholders. Google keeps proving that time and time again: being a geek puts you in the best position for the continuation of the species, even if you're not getting laid *today*, you will get laid PLENTY when you get lots of money, and therefore sire many children and overwhelm the market with clones of yourself. Then improve the genome of your clones in modular functionality, so that parts become interchangeable. You can now patent genes, so you don't have to patent actual clones to profit.

    Seriously, at this rate, I see this coming as the next logical step for Google. Bioharvesting of nerds for fun and profit. Especially nerdy gamer chix!

    • Re:Awesome Mindpower! (Score:2, Insightful)

      by Otter ( 3800 )
      Seriously, at this rate, I see this coming as the next logical step for Google. Bioharvesting of nerds for fun and profit. Especially nerdy gamer chix!

      I like how you put this in bold and still managed to catch an Informative...

    • Re:Awesome Mindpower! (Score:4, Interesting)

      by dsginter ( 104154 ) on Thursday September 22, 2005 @09:06AM (#13621113)
      The use of intelligence to develop accurate results in a predicative system

      You don't even need intelligence, in many cases. Just the "wisdom of the crowd". Read this [cnn.com] for more info. A quote:

      In an early example, Surowiecki refers to a study conducted by the British scientist Francis Galton. Galton was a believer in the power of the elite, noting "the stupidity and wrong-headedness of many men and women being so great as to be scarcely credible." But at a fair, he noticed a wagering competition in which people bet on the weight of an ox. Eight hundred people participated; some were butchers and farmers, others just idle guessers.

      When Galton averaged the estimates, he expected the result to be way off. Instead, the crowd had come within one pound of the ox's weight.

      • noting "the stupidity and wrong-headedness of many men and women being so great as to be scarcely credible."

        Yeah, we know. Just look at the vast majority of people that you have to interact with on a daily basis.

        No, I'm not trolling. I'm being serious. Take a good look at those around you and you'll see the truth behind Galtons comments.

        The difference between why a crowd can give you better results than asking a single person is because each of us has our biases and preconceptions. However,

        • However, put us in a situation where others can give input and the groupthink mentality starts to take over.

          You should be more precise. "Groupthink" is usually considered a perjoritive, but it doesn't quite look like you're using in that sense.

          If members of the group can provide feedback to each other directly, then groupthink can take over and I'd expect a group to start acting more like an individual. This is the perjorative sense.

          However, if the members of the group don't feed back to each other (or do s
      • This is the basis of a theorem in math called the Central limit theorem. The central limit theorem says that data which are influenced by many small and unrelated random effects are approximately normally distributed.
    • Google decided to pose the question to it's search engine. "Is there a GOD". After feeding in all internet information available they typed it in and waited. After a lot of hard disk searching and the checking of all drives the computer went into an eerie silence for a few hours and then started typing.

      "Not enough resources to compute answer."

      This time they were going to get an answer to an age old problem and nothing would stop them. After months of negotiations with government

      • Re:Awesome Mindpower! (Score:2, Informative)

        by Anonymous Coward
        For those not in the know, this story was first written by Isaac Asimov, and he called it "The Last Question." Iirc, in Isaac's version the computer's last statement was "Let there be light"
    • I see this coming as the next logical step for Google. Bioharvesting of nerds for fun and profit. Especially nerdy gamer chix!

      Wouldn't that violate the Endangered Species Act?

  • Gmail is the ultimate prediction market (Score:5, Interesting)

    by nokilli ( 759129 ) on Thursday September 22, 2005 @08:40AM (#13620918)
    Want a tip on when a stock is going to move? Monitor the number of times your users send email to one another containing the stock's symbol in the message. When the number goes up, activity is sure to follow.

    But they wouldn't do that, right? Because...

    Because...

    Exactly.
    --
    You didn't know. [tinyurl.com]
    • If they're tracking anonymous data, then I'm fine with that. They're providing the service for free, so it's not unreasonable for them to try and benefit out of it, and as far as ways to get something back from the subscribers, it's not a bad one.

      I'd love to see what kind of stuff they come up with.

      Anyone know of any legal issues concerning google doing something like this?
      • If they're tracking anonymous data,

        Rather than USENET, I believe the O.P. was referring to gmail. Web based or not, people (perhaps wrongly) have an expectation of privacy in their personal communications. For Google to be snarfing this out of gmailed data would be extremely evil (violating one of their core principles.)

        While the suggestion that Google is monitoring gmail for stock activity is really funny, in real life it wouldn't happen for several reasons. First, they'd be violating their user's

    • That's only useful if those emails pick up before anyone else's stock predictions. In order to make money in the stock market, you not only have to predict the future, but you have to do it better than your competitors. If your forecast comes after everyone else's, then the price will have inflated or deflated by the time you can buy or sell, meaning you cannot profit.
    • Eh, who cares if they do?

      What matters is if they start connecting info to private profiles.
    • Because knowing that there'll be activity tells you nothing you can turn into cash. "Up or down?" is the question you want answered not "is it moving or not?".
    • Easier to monitor searches for the company, and for stock investment information relating to the company.

      Most people research the buy a little first, and since Google is pretty much the dominant research tool, they'd likely get a good sense of what was going on from looking at those numbers rather than the notional idea of a bunch of businessman types talking about stock on their gmail accounts.
    • You don't necessarily need to be Google to do this; for a popular enough company anyone could just use AdWords and monitor the stats. Of course just because someone's talking about a stock doesn't mean they're saying nice things about it!

  • Result of this (Score:1, Insightful)

    by Anonymous Coward
    1. Predict Market
    2. Act on Prediction
    3. Cause Market Impact
    4. Need New Prediction!

    World is quickening

  • putting their users to work (Score:3, Insightful)

    by garat ( 899448 ) on Thursday September 22, 2005 @08:41AM (#13620927) Homepage
    Hmm... prediction markets... Ahh: http://en.wikipedia.org/wiki/Prediction_market [wikipedia.org]
    Wonderful, Google realizes that many people think they're absolutely wonderful and finds a way to put those people to work for them. As for the possible MSN/AOL deal about to occur and "kill" Google, with ideas like this and a willing user base Google isn't threatened at all.
    • Wonderful, Google realizes that many people think they're absolutely wonderful and finds a way to put those people to work for them.

      Users believe that Google is absolutely wonderful because Google appears to be working for the users! Now, who is receiving the best benefit? Obviously Google but does that matter to the users?

      That's not something we can answer but it appears that the users are happy. That may change depending on what Google does.

  • Soon available in Beta (Score:4, Funny)

    by WormholeFiend ( 674934 ) on Thursday September 22, 2005 @08:43AM (#13620949)
    GoogleAstrology
  • The markets were designed to forecast product launch dates, new office openings, and many other things of strategic importance to Google. So far, more than a thousand Googlers have bid on 146 events in 43 different subject areas.

    So the prediction for a new office opening gets more accurate when you discover that there are less workspaces than people in the office. A production launch date more accurate when the people working on the project are occupying those workspaces all the time and really start to

  • Sheep (Score:5, Insightful)

    by scottennis ( 225462 ) on Thursday September 22, 2005 @08:47AM (#13620979) Homepage
    "Group wisdom" is a misnomer for "herd mentality."

    I predict that when you herd sheep from field A to field B, they will eat whatever is in field B rather than return to field A.
    • That's why you usually have a small group of leaders at the top of companies, so that coherent consistent decisions get made. However, sometimes the people under them all firmly agree that they're making a bad decision, or missing a business opportunity, etc. Usually all the employees just start gossipping and complaining at the water cooler. An internal prediction market will harness this internal dissent, and allow the rare ocassions where mob/democratic rule is better than oligarchy, to influence the
      • In other words, if enough sheep get stomach aches from eating the fodder in field B and start bleating loudly, the "leaders" may decide it's a good idea to move tham all back to field A (or maybe even to field C).
        • So you're saying that the sheep should never bleat?

          Guess what "freedom of the press" is for? To increase the loudness of sheep bleating. So that, let's just say, maybe there's a hypothetical situation where a large hurricane hits land, and the leaders are perhaps a little too slow in their emergency management jobs. The sheep should bleat. And then when the next disaster comes, look at that, the leaders act more quickly.

          The sheep aren't remotely smart enough as a mob to predict the path of the hurr

          • Re:Sheep (Score:3, Funny)

            by mysticgoat ( 582871 )

            This can be summed up by the plaque I once saw on the wall of a manager who had a reputation for making good decisions:

            I must hurry and catch up with the others, for I am their leader!

    • Re:Sheep (Score:4, Informative)

      by SilentReallySilentUs ( 908879 ) on Thursday September 22, 2005 @09:24AM (#13621260) Homepage
      You should read "Wisdom of Crowds" by James Soriecki. It is a very nice text on how, where, and why "collective wisdom" of a few average individuals is better than the wisdom of a few experts.

      • Re:Sheep (Score:3, Informative)

        by SparafucileMan ( 544171 )
        Well you should read "The Madness of Crowds". amazon [amazon.com]

        the crowds don't know shit when it comes to financial markets.

        • Re:Sheep (Score:2, Interesting)

          by dubl-u ( 51156 ) *
          the crowds don't know shit when it comes to financial markets.

          Most of the time, they do. "The Madness of Crowds" is about a relatively rare phenomenon, the speculative bubble. Most of the time markets work very well.

          It's also worthn nothing that bubbles are much less spectacular than they used to be, thanks mainly to the crowd learning about bubbles and crashes. If you've read "The Madness of Crowds", you'll know that some of the early ones caused widespread financial disaster. Our most recent ones, the Int

          • Re:Sheep (Score:2, Interesting)

            by SparafucileMan ( 544171 )
            um. the internet bubble was the greatest bubble since the great depression. there are many markets in the world. when you consider them all, speculative bubbles happen quite frequently, and they are always nasty.
            • um. the internet bubble was the greatest bubble since the great depression

              Care to quickly compare and contrast the economic effects of the Internet bubble with the Great Depression? I note [bea.gov] a 45.6% drop in nominal GDP during the great depression versus no nonminal GDP drop at all after the Internet bubble popped.

              Really, if the Internet bubble is the worst that happened in 70 years, you're making my point for me. Compared with the value we get from markets, bubbles are a small problem.
    • It may be true that most people are followers and not leaders, but in most prediction markets (not Google's yet, as I understand it), people have to back up their predictions with cash. Those that are informed in their predictions win money and come back. Those that make ignorant predictions lose money and go away. This is what makes prediction markets work. It's a way to consistently sift the wisdom from the "herd mentality," without resorting to populism or democracy.

      Prediction markets are a great inv

    • Re:Sheep (Score:2)

      by X ( 1235 )
      Actually, you got the idea behind group wisdom totally wrong. In fact group wisdom tends to be poorer when the group is influenced by "leaders" or "experts".

      Groups/information markets aren't always "wise". The circumstances where they perform best is when there are no experts because the relevant information is spread diffusely amongst the members of the group.
  • that this story will be duped by the end of the week

  • The honest weatherman (Score:2, Interesting)

    by Anonymous Coward
    Reminds me of E.T. Jaynes's "Honest Weatherman" example in his book, Probability Theory: The Logic of Science (section 13.5). By making a weatherman's salary proportional to a certain function (logarithm of the probability he predicts, or entropy, or something, I forget), the weatherman has an incentive to make the best possible predictions: his salary will be directly proportional to the quality of his predictions, being maximized when his rain forecast probabilities match the actual probabilities of rai
    • [being marginally off-topic ramblings]

      when his rain forecast probabilities match the actual probabilities of rain.

      Does anyone know what meteorologists mean when they say "x% chance of weather event"? Does this just mean "given current circumstances, we have seen x% chance of said event occurring in the past" or is it some other silliness? For instance, I love the NHTSA star-rating system which says "this vehicle has an x% chance of a rollover in an accident" which is meaningless without ancillary informa

    • By making a weatherman's salary proportional to a certain function (logarithm of the probability he predicts, or entropy, or something, I forget), the weatherman has an incentive to make the best possible predictions: his salary will be directly proportional to the quality of his predictions, being maximized when his rain forecast probabilities match the actual probabilities of rain. You could set up a "free market betting system" this way, rewarding the quality of people's guesses regarding the likelihood

  • Isn't this exactly what the Pentagon tried to do.. (Score:4, Insightful)

    by no_nicks_available ( 463299 ) on Thursday September 22, 2005 @08:53AM (#13621022)
    with terrorism? It's too bad people got all pissy since this method of prediction works fairly well.
    • The Pentagon was ahead of its time I think.
    • Not just the Pentagon. Strategypage [strategypage.com] runs a prediction market with this type (terrorism, and national security/defence-related in general) of events.
    • Yes, this is similar if they are using common futures market methods. That event illustrated more than most what a collection of barely-single-celled organisms humanity is. People here on Slashdot we're foaming at the mouth thinking there was going to be actual money wagered of terrorist events, even though that should fail to get past even rudimentary BS detectors. Even mainstream news media was incorrectly reporting that. I remember it made me want to start lobbing Ebola bombs or start stabbing people at
    • Why is this modded as funny? Is that just some poor sarcasm on the part of the local audience, or are that many slashdotters that ignorant of the fact that the Pentagon actually DID put together exactly such a system, and that statisticians and other analysts thought it was a great idea. That the Stupid Media(tm) could only bring themselves to report on it in terms of "people betting on next terroris targets" (thus leading shrill lefty bloggers to screech about how craven the DoD is, how much they think of
      • I have mod points, and I was going to mod your post as funny, but I'm sure you don't get the irony of what you're saying.

        On one hand you say that groups are so wise that they would be able to predict security threats, but at the saem time you're complaining about groups of people being so stupid because they don't understand.

        So would the Stupid Media(tm) and shrill lefty bloggers be able to predict a terrorist attack? Or would they be excluded from participating in this system? How do you decide which g

        • On one hand you say that groups are so wise that they would be able to predict security threats, but at the saem time you're complaining about groups of people being so stupid because they don't understand.

          Not at all. I'm saying that people who spend their lives dealing with security related issues, intelligence analysis, foreign affairs, law enforcement, diplomacy, and international economics are the components of a group that certainly will have something useful to say about terrorism hotspots.

          I'm a
          • Well less ironic, maybe, but definitely poorly thought out. There's the old saying: "when all you have is a hammer, all of your problems start looking like nails." When you're a statistician, then all problems look like they can be solved by statistics.

            I can see some value to such a system, but all value is lost if you make it public. If an enemy can read that you're predicting an attack next week, he'll probably just postpone his attack until the week after next.

            Also it seems to me that statistics will

    • Yup. It was called the Policy Analysis Market [wikipedia.org], which was essentially a prediction market [wikipedia.org] for forecasting the future of the middle east and how possible policies might impact the region.

      Slashdot had a story on it back in 2003, titled Pentagon lets you bid on terrorism? [slashdot.org]

      It's quite interesting to see how your typical slashdotter reacted to it back then. Some quotes (all from comments moderated +4 or higher):

      "In any event, while I support innovative ways of fighting terrorism (as opposed to wiretapping everyone

  • No Rocket science here folks (Score:5, Interesting)

    by OlivierB ( 709839 ) on Thursday September 22, 2005 @08:54AM (#13621029)
    Predicting involves extrapolating from a current ** sample** of data to predict the future based on one own interpretation /recognition of patterns.

    The better your size and quality of your sample combined with a finely tuned pattern recognition the better your forecast (I won't go into exceptional events which by definition are exceptional).

    So what do we have here? A larger sample of better quality for starters.

    Also do not underestimate the power of the masses. If your sample=population size this is no longer forecasting (i.e. extrapolation) but the writing on the wall! (as long as people do as they say they are going to do).

    So if your sample size increases dramatically, with better quality (smart employees) things will tend to happen as per the survey!

    I know I am oversimplifying but but these are the basics of neural networks (or in this case a neural network of neural networks). Look it up.

    Of course, predicting is not extrapolating if it is purely a random guess.

    Have fun!
    • "Also do not underestimate the power of the masses. If your sample=population size this is no longer forecasting (i.e. extrapolation) but the writing on the wall! (as long as people do as they say they are going to do). "

      If sample size = population size, all you have is certainty about what people believe (assuming they answer honestly). What people believe != what is going to happen. The idea here is that as time approaches the time of the event, people can guess more accurately about the event.

    • Predicting involves extrapolating from a current ** sample** of data to predict the future based on one own interpretation /recognition of patterns.

      The better your size and quality of your sample combined with a finely tuned pattern recognition the better your forecast

      I think it is quite a bit more complex than that, with the complexity arising from a really simple mechanism: the metadata that is shared among members of the crowd.

      For instance, I expect that a "guess the weight of an ox" game played b

  • I predict... using my wisdom (Score:3, Informative)

    by Gopal.V ( 532678 ) on Thursday September 22, 2005 @08:55AM (#13621040) Homepage Journal
    I predict that google stock will rise ...

    I've started playing Yahoo! Buzz games for a few months now (I dropped a virtual packet on FireFox). It's called Yahoo! Buzz game [yahoo.com]. These folks seem to gather their data from Yahoo ! search queries and from the logs on who clicked what. Which is why it seems to really follow the non-geek popularity levels too well - Google is what the geeks use (which is why I almost always seem to lose there).

    This is just Google calling Market surveys by another name. I'm sure I can dig up a couple of WalMart papers about the same thing in the real world. Sort of vote with your money approach vs tell us approach (demand vs desire).

  • Non-ideal behavior: bubbles & manipulators (Score:4, Insightful)

    by G4from128k ( 686170 ) on Thursday September 22, 2005 @08:58AM (#13621056)
    This is really cool, but I hope they control for two of the less-than-ideal behaviors of markets.

    1) There are two ways to "be right" in a market. First, I can make the right choice as to the actual ending outcome. This a buy-and-hold strategy. Or, second, I can make the right choice as to the direction of price movements in the market. This is the speculator's buy-and-sell strategy. The first strategy means the the market converges on the "true" expected value. The second strategy leads to bubbles and crashes that don't provide as much useful data on the actually variable being modeled by the market. Google wants to encourage the first type of trading, but not totally suppress the second type of trading because speculators provide liquidity in markets.

    2) Manipulators can "cause" events to occur in the way that maximizes their return, but suboptimizes Googles performance. If I bet that a given project will be done in October and it looks like its getting done early, what stops me from causing a small delay? Of if the project is being delayed too much, what stops me from descoping the project or doing a fast, low-quality job to complete the project within my chosen time frame. In either case, I can manipulate the outcome to win in the market, but hurt Google.

    Note that I don't include insider trading in the list problems. Google doesn't really care if the market is fair, only that it provides accurate predictions. In fact, Google might encourage insider trading as a way to encourage communication inside the company. The more people that share their "inside" information on upcoming strategic, the better.
  • The Wisdom of Crowds by James Surowieki. [randomhouse.com]

    I'd recommend it, unlike a lot of popular science books it does actually cover the material reasonably accurately and its quite engagingly written as well.
    • the madness [amazon.com] of crowds...

  • I'll buy it if... (Score:3, Funny)

    by AnonymousYellowBelly ( 913452 ) on Thursday September 22, 2005 @09:00AM (#13621073)
    ... they do correctly predict:
    - launch date of MS Vista;
    - launch date of Ballmer's furniture;
    - which 'feature/bug' will be dropped from the Vista release.

  • (1) Create a predictive survey with quantitative questions
    (2) Graph the responses, and calculate entropy using information theory (or, just calculate variance)
    (3) Measure the change in variance/entropy over time
    (4) Conclude that as a product gets closer to release date, people have more consistent ideas of the price.

    The difference between this and traditional quantitative market prediction surveys? Google gets its sample from "Teh Internets".

    Nothing to see here, please move along.

  • Google! (Score:4, Funny)

    by Mr. Underbridge ( 666784 ) on Thursday September 22, 2005 @09:04AM (#13621102)
    Hey, I wonder what Google's doing today, I haven't heard anything about them lately.
    • Hey, I wonder what Google's doing today, I haven't heard anything about them lately.

      Exactly, Slashdot has been sucking on Google's proverbial teet for far to long now. It's getting tiresome.

  • will it finally be able to tell us when Duke Nukem Forever will be release?

  • Central limit theorem in action (Score:5, Interesting)

    by call -151 ( 230520 ) * on Thursday September 22, 2005 @09:27AM (#13621297) Homepage
    The classic example of "crowd wisdom" is the jellybeans-in-the-jar experiment, often used in introductory MBA classes to convince people that open markets value securities (basically) fairly. The experiment goes like this: the professor brings a jar of jellybeans and asks everyone to guess how many there are in the jar. The individual estimates may vary quite a lot, but the average of the estimates in the class is usually close, in fact often closer than the closest estimate of any of the students depending upon the size of the class. That is in a situation where the students have very little information about the jar and perhaps no experience with such estimates. If there were greater experience and/or they were allowed more information then presumably the individual and average estimates would be even closer. Basically, this can be described as the "Central Limit Theorem" in action- that the standard deviation of averages is smaller than the standard deviation of the individuals by a factor of the square root of the sample size, as illustrated in this applet [web.uvic.ca] or in this Mathworld description [wolfram.com]. The CLT actually says more- that as the sample size increases, the distibution of averages approaches that of the normal ("bell curve") distribution, so the distribution of avergaes is roughly normal, and then techniques designed to analyze the normal distribution can be applied with greater certainty.

    • very misleading (Score:2, Interesting)

      by Main Gauche ( 881147 )

      This actually has nothing to do with the CLT. The CLT is a statement about the distribution of a sample average in relation to the population average.

      OTOH, TFA is about information markets, which concerns comparisions between the average of people's forecasts of events, and the "true" (or natural) frequency of events. We are not comparing a sample average to a population average. (In other words, to apply the CLT here, we'd make a statement comparing the average forecast of 100 people to the average fo

  • This can't work (Score:3, Interesting)

    by SparafucileMan ( 544171 ) on Thursday September 22, 2005 @09:28AM (#13621309)
    This only works when the distribution is NORMAL.

    Its "well documented" that financial markets are not normal. They are "fat-tailed"--extreme events happen with regularity. This attempt has been tried thousands of times by people with far more at stake than google. They also had more money, more PhDs, etc. And they failed. They always fail at some time or another. There is no algorithm for the financial markets.

    Those who doubt had best read the story of Long Term Capital Management. wikipedia link [wikipedia.org]

    What Google is doing is interesting, but it's no magic bullet. And I'm not even sure I'd use this system to place money on the markets. Fads come, fads go. Neural Networks. Genetic Algorithms. Fibonacci numbers. Eliiot Waves. what-have you. They're all nonsense. The big ol boys still make their money the old fashioned way: graft, theft, and deceipt.

    Anyone honestly think Google knows something that the trillion dollar financial industry doesn't? They have far more at stake that Google. They have their fair share of MIT PhDs, economic nobel prize winners, sparc computers, access to nearly every price in the world at any time. And they still make the majority of their cash through commission.

    Call me a cynic, but the math says this can't work. And I'll trust the math before I trust a company that doesn't even pay dividends any day.

    • Re:This can't work (Score:4, Insightful)

      by dajak ( 662256 ) on Thursday September 22, 2005 @10:10AM (#13621690)
      What Google is doing is interesting, but it's no magic bullet. And I'm not even sure I'd use this system to place money on the markets.

      I certainly will, if many people use the system. It is extremely useful to know what the majority believes. It is one thing to predict accurately what will happen, and another thing to gauge to what extent this knowledge is already factored into market prices. the point of reading the headlines of financial publications has always been to know where the herds are going, as far as I am concerned.
    • I have recently been working on a website http://www.consensusview.com/ [consensusview.com] with the aim of testing exactly these ideas. We were inspired by the book "The Wisdom of Crowds" by James Surowiecki, and thought it would be interesting to see how well the ideas worked in the financial markets. It's a simple "up or down" decision, made on a daily basis, but we're thinking about extending the time frame and providing a "rating" system as well. Initial results actually seem quite promising, with some hit-rates as high a

    • This is not about predicting the stock market. It is about using the idea of the stock market to make predictions.

      The idea is to create a market and use the prices to make predictions. This is not new. In fact, the US government got some PR trouble when someone proposed to use this idea to predict terrorist attacks. Look at this workshop [rutgers.edu] to get some pointers for further information.

    • This only works when the distribution is NORMAL.

      That's why you have special markets used to explicitly deal with non-normal distributions. For example, check out this prediction market [intrade.com] over at Intrade on what the federal funds rate will be by the end of 2005. Notice that they actually have several different tickets, for increases ranging from +1% to +5%.

      In the real stock market, it's also why you have things like options, which I believe can be used to get a handle on non-normal distributions.
  • I used to work for a company that did market forecasting using stuff like fractal theory, etc. I remember that one of our CERN physicists (we didn't have an economists on staff) did a correlation analysis of implied volatility versus actual volatility. Implied volatility is calculated from option prices and, according to the Black-Scholes model, indicates what the market as a whole thinks the future volatility of the underlying will be. He found there was absolutely no correlation between what the market ex

  • zLabs' project zocalo (Score:3, Informative)

    by Salamanders ( 323277 ) on Thursday September 22, 2005 @09:55AM (#13621587)
    http://zocalo.sourceforge.net/ [sourceforge.net]

    Great realtime prediction market, written in python, uses ajax for updates, very slick. (Disclaimer: I was an intern with zLabs over the summer and chatted with the developers often, very smart people)

  • Yahoo has one going you can actually play with (Score:5, Insightful)

    by X ( 1235 ) <x@xman.org> on Thursday September 22, 2005 @10:05AM (#13621663) Homepage Journal
    Yahoo has launched one of these [yahoo.com] to the public back at in March of 2005 at the O'Reilly Etech conference. They actually had a contest where the top performer got a Mac-mini.

  • This ought to raise the hairs on your neck (Score:5, Interesting)

    by TampaDeveloper ( 834876 ) on Thursday September 22, 2005 @10:11AM (#13621697)

    I see all these dismissals, as if this is a joke. But might I suggest that the researchers at Google aren't trying to get an acceptable sample-size, as many people here are naively suggesting. It seems much more likely that Google is researching the thought-process of prediction itself. It also seems likely that a number of fairly intelligent industry-insiders also believes they're close to figured out how to create a prediction engine that is accurate (ahref=http://slashdot.org/article.pl?sid=05/09/22 /1229238&tid=109&tid=120&tid=217rel=url2html-6813 [slashdot.org]h ttp://slashdot.org/article.pl?sid=05/09/22/1229238 &tid=109&tid=120&tid=217>)...

    Think about it; If they have enough information (IE the web), they can hone it by feeding it past events to "predict" other past events and then comparing their prediction to what actually occurred. If it's able to predict events that have occurred given what came before those events, then there's a good chance that it will work with current data to predict future events.

    Feel free to label me a wacko... Perhaps I've seen "Terminator" too many times. But I've also worked with a number of freakishly intelligent people; the kind that Google has been hiring right-and-left without any apparent reason. I've always said that technology could someday facilitate society's return to bondage... It may seem far fetched, until one considers the sheer breadth of people that hate "western" culture. Given the past as a predictor, I would be inclined to believe that few Middle Eastern leaders would hesitate to use technology to exterminate as much of western culture as they could... Folks, its not unrealistic to recognize the likelihood that technological-advancements will make sudden and drastic changes to our way of life... This ought to give us pause. What balances are essential to our culture? What imbalances will end it? Someday, somewhere, somebody is going to come up with a way to utilize technology to facilitate his/her agenda. Maybe we don't have to worry about Google, since the magnitude of the consequence is, at least in part, proportionate to the magnitude agenda of the individual(s). But things such as this should be cataloged in our brains as evidence that this mode of societal-failure is plausible.

    Ok, flame away. But I hope I never have to say I told you so.

    • If they have enough information (IE the web), they can hone it by feeding it past events to "predict" other past events and then comparing their prediction to what actually occurred. If it's able to predict events that have occurred given what came before those events, then there's a good chance that it will work with current data to predict future events.

      Do you think it could predict that someone would invent a machine that could make predictions, and then that people can use the predictions to completely

    • I would be inclined to believe that few Middle Eastern leaders would hesitate to use technology to exterminate as much of western culture as they could

      I doubt it, that would be stupid, think about it for a moment, who would they sell oil to? They would be destroying their primary source of income, and their economies on the whole are not diversified enough to withstand the loss of half of their oil revenue (assuming China remains, although China in turn is also greatly dependent on exports to 'the West' a

    • Prediction isn't a thought process. It's merely a statement about the future. The statement doesn't even have to come true. What Google has here is a rational way to evaluate the likelihood of predictions coming true. I personally am a pretty hardcore believer in the power of betting markets to make best possible estimates of predictions because they are the most effective means of aggregating information that we currently have. Google is just using that potential for valuable predictions about their corpor
  • Play the game that runs the code for the above sourceforge link.

    http://www.ideosphere.com/ [ideosphere.com]

    We have many players and you can have a lot of fun trying to beat the market. These markets don't always get things correct though... The market bet a 20% chance that Gas would hit $3 per gallon in the US. I always thought the chance of that to be far higher.

    Hence, these markets don't always predict things as well as people suspect. If the majority is wrong, the market will have a bias. Some have given guessing

  • I'm surprised no one has mentioned The Business Experiment. [thebusines...riment.com]

    Their new business will be selling "crowd wisdom!"

  • Slashdot Comment Moderation (Score:5, Insightful)

    by chuckw ( 15728 ) * on Thursday September 22, 2005 @01:44PM (#13623514) Homepage Journal
    CmdrTaco has been trying to do this for years with his comment moderation system. I think a key problem is that he dosen't use a big enough sample size. It only takes a few people to moderate a comment up to the highest level. The rules of crowd wisdom state that the larger your sample size, the more likely you are to arrive at the "correct" answer. Granted, with something like a story comment, there is no "correct" answer, only interesting and relevant responses. CmdrTaco's goal was not to tease out the interesting comments though, it was to filter out the irrelevant and wasteful spam.

    In essence, CmdrTaco had no choice. Spam was starting to choke slashdot comments and making them less than useful. The moderation system saved the comment system, but didn't, as many people assume that it should have done, make the comments more interesting.

    I believe that if the prevailing attitude among slashdot developers is to "weed out the spam", we'll see a slow decline of slashdot's popularity until it's made irrelevant by RSS feed aggregators.

    IMHO, the attitude *SHOULD* be to exploit slashdot's major differentiator over simple aggregators, which is the community it has created. In other words, they should invert the "weed out the spam" attitude into a "make the comments more interesting" attitude. It's a subtle difference and, on the face of it, it would appear that one begets the other. I contend that weeding out spam does not make comments more interesting and conversely, making comments more interesting won't weed out the spam. Thus we come to the root of the problem, two crosswise goals.

    CmdrTaco has to worry about the system from a performance standpoint. Weeding out the spam means less bandwidth and storage costs. That's immediate ROI, and a good thing on many levels. The community, however, needs more than 1,2,3,4 or 5 to determine what comments to read and which to ignore, to make them interesting. I can conjecture at a few ideas that would make it better, but I do not know the ultimate solution, and I doubt anyone else does either. I believe the problem requires more than just CmdrTaco playing whack-a-mole with ideas, meta-ideas and meta-meta-ideas etc. It requires serious PhD dissertation level study.
     

  • Re: (Score:2, Interesting)

    by account_deleted ( 4530225 )
    Comment removed based on user account deletion

  • It must be a scary experience to be in commercial competition with Google. Not only do they know literally everything about you, right down to where you grandmother buys porn, but they can predict you before you move. It must be rather like fighting the Kwisatz Haderach.

    Google is the singularity. You will want to be be assimilated.

Slashdot Top Deals

Some people manage by the book, even though they don't know who wrote the book or even what book.

Close