It's so easy to be swept away by the seductive embrace of Web 2.0 tools. The Browser got all excited this month reading about the launch of Betocracy, a Blogger-like tool that makes it relatively easy for any Web user to set up a prediction market, that is, a speculative market in which participants make bets about the outcome of events (in the case of Betocracy, not with real money). The potential applications for that are huge, and the collective data that could be harnessed might teach us a lot about market behavior in non-monetary situations.

Such markets are not new: the University of Iowa business school has been running a real-money political futures market since the 1988 presidential contest. The market's method is that a user purchases a futures contract that, say, a given Democrat or Republican will win a race, for whatever price the market will bear; it's redeemable for $1 if the correct choice is made. The price of the contract thus represents what the market as a whole believes the outcome will be; if the price is 50 cents, it's a dead heat, 0r so the crowd believes.

If that sounds a lot like gambling, well, it is. From a regulatory point of view, the Iowa market appears to be an academic exception to the general U.S. prohibition against betting on politics; in Britain and elsewhere, lots of people use the peer-to-peer gambling site Betfair to place bets on all sorts of things, including political contests, interest rate movements, or the closing price of the FTSE 100. (The legality of U.S. residents using such sites is very much up in the air.)

Aside from the tempting prospect of making money, these markets also theoretically aggregate information that is useful for making predictions. And that's why what IS new and potentially exciting about Betocracy is that it allows pretty much anyone to set up a market like this on any subject they want.

Here's the thing, though: the actual predictive power of these markets is very much in dispute. Anyone who followed the November elections closely on these markets will recall that even up until Election Day they showed a strong market prediction that the Republicans would retain the Senate. Now, the hardcore advocates will tell you that this does not represent a failure of the system, that it's a question of probabilities, that market participants had insufficient information, that the method still has greater predictive value than a poll, etc. Adjudicating that debate is well beyond the Browser's ken, but using a common-sense definition of "prediction," it certainly looked like the markets had a pretty lousy election (especially for the majority who lost money!).

But even the most passionate defenders of predictive markets argue that they will only work when there are a sufficient number of traders to make the market liquid. And that's where, we fear, Betocracy is severely limited. Granted, the site's been up for only a couple of weeks. But it should be obvious that people who set up a predictive market about who is going to come in second place in a charity 5k run are doing nothing more than amusing themselves. Even more serious-sounding markets, such as whether Barack Obama becomes the Democratic presidential nominee in 2008, are utterly meaningless if no or few people participate in them. The Browser can't help but feel that the wisdom-of-crowds meme is fueling a lot of experimentation with questionable payoff.