The (conventional) wisdom of crowds

The Harriet Miers nomination seems to have started a somewhat frantic feedback loop between bloggers and TradeSports.  Fueled by James Surowiecki's famous insight that groups composed of decentralized individuals can often be surprisingly good decision-makers, many bloggers seem to be attributing a sort of mystical predictive power to the peaks and valleys of Miers futures.   Anyone who hasn't previously followed TradeSports can't really be blamed for thinking this, since the accuracy of Tradesports is often taken for granted; witness NYT pet conservative John Tierney's recent repetition of the dubious claim that TradeSports (or its alter ego Intrade) had an inside line on the papal selection and the 2004 presidential results.*

As someone who's followed TradeSports since it pegged Howard Dean as a sure thing to win the presidential race, I can attest that it's just not right all that often.  That's not to say it's always wrong, but simply that it knows nothing more than the rest of us.  Its early predictions are quite inaccurate, but as the outcomes of uncertain events become more certain,  it becomes better at guessing them. By 10 pm on election night, TradeSports bettors had picked Bush as the winner of the 2004 election; but then that was also about the time the friends with whom I watched the results decided it was probably best to put that newly bought champagne away for a few more years.

What TradeSports is really good at producing is not wisdom, but conventional wisdom.  After all, part of the reason markets are supposed to be accurate is that people will act on private information, and individual expertise, more efficiently than that information can be filtered through the punditry.  But let's face it; anyone with insider knowledge, or even particularly good judgment, is unlikely to be mucking around on a low-stakes, Dublin-based gambling site conducting transactions of dubious legality.  What TradeSports users do seem to be is fanatical consumers of media and rapid responders to news events.   When a presidential candidate rises in the latest Gallup poll, TradeSports users will bid him up within hours after it is released; when Novak predicts that a Supreme Court justice's resignation is imminent, bettors will peg that justice as most likely to resign.  Thus, while TradeSports may not have any special ability to see into the future, it's extremely useful as a distillation of the state of conventional wisdom - at least on those days when reading the latest edition of the Note seems like just too much trouble.

With the Miers nomination, however, something different seems to be going on.  Miers contracts have been characterized by wild volatility.  Late last Friday night, they plunged from the 60s to the 20s (with a large increase in trading volume as well); by morning, they'd recovered somewhat, and now she's trading at around 35.  Nothing obvious happened last week to produce these results; there were wan, isolated Miers defenses amid a steady stream of Miers criticism, but what else is new?  What's interesting, however, is that a number of bloggers seem to have discovered the TradeSports Miers contracts around the end of last week.  Coupled with the fact that bloggers are far more hostile to the Miers nomination than the general public, it makes you wonder.  Are there people out there bidding down Miers in order to create the impression that her nomination is doomed?  Could the widespread exaggerated faith in TradeSports be manipulated in order to create a self-fulfilling prophecy?   It will be interesting to see how this plays out.

*Though initially decent predictors of the presidential results, TradeSports users got swept up uncritically in exit poll hype. They didn't have a clue about the Pope.

Comments

[Billy F]

"anyone with insider knowledge, or even particularly good judgment, is unlikely to be mucking around on a low-stakes [tradesports]2

I recall that they (tradesportS) had turnover over $10 million on just one of their futures. Perhaps this is small change to most people who might have inside info, but if you are teh guy on the exit poll or have other prece sensetive info I find it hard to believe that sums of this size = low-stakes for the majority.

Just my 2 cents however.

[Billy F]

"Could the widespread exaggerated faith in Tradesports be manipulated in order to create a self-fulfilling prophecy?"

This is VERY interesting.

If markets are low volumes, then a little $ invested could swing the price a lot, this in turn could influence some people.

I vaguealy recall this being said at the time of the 2004 election also. Soros selling GW future down or something like it.

[Eric]

The potential for manipulation is very real, especially given how thin the current-event markets on Tradesports tend to be. There are only seven asks (at 74-75) posted right now on the Libby conviction contract, so for $50 or so you could spike the ask to 100, if only temporarily.