The Volokh Conspiracy - More on the Accuracy of Electronic Markets:

Jim writes, in the post below: "The question is whether experts can usually do a better job than markets. It would seem that the answer is generally No." This is far, far from my expertise, but I'm not sure that is the question. I think the question is whether markets do a good job predicting the discretionary decision of one person, as compared to predicting the collective outcome of the individual decisions of many. I can see markets doing a good job predicting collective decisionmaking, but I don't see the advantage they have in predicting what one person is thinking. Thus, it seems to me that Tradesport users incorrectly predicted Rehnquist would decide to retire because that's what newspapers were incorrectly predicting at the time; ditto for the idea that Clement would be nominated to replace Rehnquist.

Related Posts (on one page):

(link)

Proud Generation Y Slacker:

I hope Tradesports is wrong, because I'll win $950 if Estrada gets picked.

7.19.2005 6:51pm

JF (mail):

There are two sets of people: those who know nothing and those who do. Note that a subset of the former is those who think they know something but don't. Furthermore, you don;t have to know the answer to be at least partly knowledgable. Suppose, for example, you're a buddy of Mike McConnell and he tells you he's heard nothing from W in a week -- well, you can probably cross him off and you're already more knowledgable than the completely uninformed. Each person who possesses some disaggregated piece of information which doesn't give the whole picture bets that way, and the aggregation contains everyone's information. The problem, of course, is that the noise traders (those who know nothing) if they're a big enough piece of the market, can make this go just about anywhere.

A final question: there's no way to judge this market by the result. Suppose ol' Mike, currently about 0.5 in fact gets it. How do you know that that wasn't his true ex ante probability of winning? This gets back to your original point Orin... you don't think there's a probability at all. There's a 1 by somebody's name and a zerop by everybody else's. Unless the result (before the announcement) is 100 for somebody, isn't that evidence that the market is "wrong?"

7.19.2005 6:57pm

Isn't the probability that something happened always 1 after it happened?

7.19.2005 7:02pm

M (mail):

Steven Bainbridge had some useful thoughts on the limits of this sort of betting market on his blog a few weeks (or something like that) ago. Daniel Davis also had some similar skeptical thoughts on Crooked Timber when the idea of a "terrorism futures" market came up. Davis is here:
hayekian makets

Bainbridge here:
insider info
(forgive me if the links don't go in properly- I don't know how to do that well.)

7.19.2005 7:06pm

anonymous coward:

Check out today's graph for Clement (click on her contract, then "Today"). It's clear (also see the reaction to the exit polls during the Presidential election) that the Tradesports political markets are excessively influenced by vague, unreliable data. The tradesports traders seem to be media consumers just like us, not rocket scientists poring over balance sheets and covariance matrices 12 hours a day.

It's worth noting that new markets are often inefficient for years before people figure out how to price them. Learning how to weigh rumors from Drudge or Wonkette is undoubtedly an art, and one we don't have much chance to perfect: in "real" markets, traders have their predictions proved or disproved every day.

7.19.2005 7:42pm

Moral Hazard (mail):

Tradesports political markets are excessively influenced by vague, unreliable data.

Unless of course the rumors they respond to turn out to be true.

7.19.2005 8:48pm

	[Orin Kerr, July 19, 2005 at 5:27pm] Trackbacks More on the Accuracy of Electronic Markets: Jim writes, in the post below: "The question is whether experts can usually do a better job than markets. It would seem that the answer is generally No." This is far, far from my expertise, but I'm not sure that is the question. I think the question is whether markets do a good job predicting the discretionary decision of one person, as compared to predicting the collective outcome of the individual decisions of many. I can see markets doing a good job predicting collective decisionmaking, but I don't see the advantage they have in predicting what one person is thinking. Thus, it seems to me that Tradesport users incorrectly predicted Rehnquist would decide to retire because that's what newspapers were incorrectly predicting at the time; ditto for the idea that Clement would be nominated to replace Rehnquist. Related Posts (on one page): Trading in Roberts Futures at Tradesports.-- According to Tradesports Betting, President Bush Will Pick John Roberts: More on the Accuracy of Electronic Markets: The Accuracy of Small-Volume Electronic Markets.-- Tradesports and Supreme Court Nominations: According to Tradesports betting, Bush will choose Edith Clement [or Edith Jones].-- (link) Proud Generation Y Slacker: I hope Tradesports is wrong, because I'll win $950 if Estrada gets picked. 7.19.2005 6:51pm (link) JF (mail): There are two sets of people: those who know nothing and those who do. Note that a subset of the former is those who think they know something but don't. Furthermore, you don;t have to know the answer to be at least partly knowledgable. Suppose, for example, you're a buddy of Mike McConnell and he tells you he's heard nothing from W in a week -- well, you can probably cross him off and you're already more knowledgable than the completely uninformed. Each person who possesses some disaggregated piece of information which doesn't give the whole picture bets that way, and the aggregation contains everyone's information. The problem, of course, is that the noise traders (those who know nothing) if they're a big enough piece of the market, can make this go just about anywhere. A final question: there's no way to judge this market by the result. Suppose ol' Mike, currently about 0.5 in fact gets it. How do you know that that wasn't his true ex ante probability of winning? This gets back to your original point Orin... you don't think there's a probability at all. There's a 1 by somebody's name and a zerop by everybody else's. Unless the result (before the announcement) is 100 for somebody, isn't that evidence that the market is "wrong?" 7.19.2005 6:57pm (link) Proud Generation Y Slacker: Isn't the probability that something happened always 1 after it happened? 7.19.2005 7:02pm (link) M (mail): Steven Bainbridge had some useful thoughts on the limits of this sort of betting market on his blog a few weeks (or something like that) ago. Daniel Davis also had some similar skeptical thoughts on Crooked Timber when the idea of a "terrorism futures" market came up. Davis is here: hayekian makets Bainbridge here: insider info (forgive me if the links don't go in properly- I don't know how to do that well.) 7.19.2005 7:06pm (link) anonymous coward: Check out today's graph for Clement (click on her contract, then "Today"). It's clear (also see the reaction to the exit polls during the Presidential election) that the Tradesports political markets are excessively influenced by vague, unreliable data. The tradesports traders seem to be media consumers just like us, not rocket scientists poring over balance sheets and covariance matrices 12 hours a day. It's worth noting that new markets are often inefficient for years before people figure out how to price them. Learning how to weigh rumors from Drudge or Wonkette is undoubtedly an art, and one we don't have much chance to perfect: in "real" markets, traders have their predictions proved or disproved every day. 7.19.2005 7:42pm (link) Moral Hazard (mail): Tradesports political markets are excessively influenced by vague, unreliable data. Unless of course the rumors they respond to turn out to be true. 7.19.2005 8:48pm