• Markets
• Economy
• Federal Reserve
• Housing
• Energy
• Digital Media

Me Media: Prediction Markets in the NYT

(With apologies to Paul for stealing the headline) Last month, we took a look at Why Prediction Markets Fail, reviewing some of the more blatant misses. That theme was picked up in a longer NY Times piece today, titled, Looking for Sure Political Bets at Online Prediction Market.

For those of you who are as intrigued about these markets as I am -- warts and all -- its a must read.

Here's the NYT Ubiq-cerpt:™ :

It’s no wonder, then, that the site has become a phenomenon. In recent months, when I have asked former advisers to Mr. Bush or Bill Clinton what they think will happen in 2008, they’ve often talked about Intrade. The Wall Street Journal and The Financial Times run regular online features based on the odds. Journalists — me included — have praised Intrade as a miniature version of the stock market, where the collective wisdom of the masses reveals a larger truth.

But now a little backlash has begun. Barry Ritholtz, author of the popular Big Picture economics blog, has put together a list of Intrade’s misses. Last Tuesday, meanwhile, I e-mailed a high-profile Democratic economist and asked what he made of the dueling numbers coming from the California polls and from Intrade. He replied with a salty

two-word message, the second word of which was “Intrade.”

Fun stuff . . .

>

Sources:
Looking for Sure Political Bets at Online Prediction Market
DAVID LEONHARDT
NYT, February 13, 2008
http://www.nytimes.com/2008/02/13/business/13leonhardt.html

Previously:

Why Prediction Markets Fail   
Friday, January 11, 2008 | 07:50 AM   http://bigpicture.typepad.com/comments/2008/01/prediction-mark.html

See Also:

Prediction Traders Put Their Money on Obama
JUNE KRONHOLZ
WSJ, February 13, 2008; Page A8
http://online.wsj.com/article/SB120286576982063999.html

Wednesday, February 13, 2008 | 03:30 PM | Permalink | Comments (8) | TrackBack (0)
add to de.li.cious | digg this! | add to technorati | email this post

Comments

Prediction markets fall flat if a majority of the people on them are getting their information via Cable TV News, and that information is false.

But it's a good source of income for those who know what's really going on.

Posted by: Drewbert | Feb 13, 2008 3:43:57 PM

For all the media whoring Intrade does, the fact remains that even their bread and butter, sports markets, are incredibly illiquid, and the politics trading is thinner still.

Posted by: Ramstone | Feb 13, 2008 3:46:55 PM

I feel happy today cuz the market goes up. People may say where's the volume? Forget the volume, the result speaks.

Cheers

Posted by: Ben | Feb 13, 2008 3:47:24 PM

The failure of prediction markets should be fairly simple to understand. A statistically significant sample of a population generates signal for that population only. Extending that signal to other populations is just plain daffy.

Posted by: Estragon | Feb 13, 2008 3:49:36 PM

The correct question is not: are these markets always right? The question is: when these markets predict that X will happen with Y percent probability, and you look at many different predictions with the same Y percent probability, are Y percent right and 100-Y percent wrong? The second part is important. Markets trading at 75 should be wrong 25% of the time; if they are not, there's an arbitrage opportunity.

For instance, the "GOP Senate 2006" was trading about 75. Would you start crying about a coin being unfairly loaded if you tossed tails twice in a row? That is the same probability as a 75% prediction being wrong.

The Dean prediction was about 50%. You're telling me I'm supposed to draw a conclusion on the inaccuracy of a predictor, when that predictor predicted a 50% chance of it happening, and then it didn't?

Not even the Obama case would reach a significance level of P=0.05.

I've never seen a study that systematically observes whether, when the Intrade market predicts something will happen with Y probability, whether it happens Y% of the time. Maybe someone here can point me to such a study.

That's not to say I think they are always perfect and efficient; just that these anecdotes don't do anything to address the question of whether they are. (Yes, just before NH, I also made a boatload of money off of Ron Paul, having kicked myself in Nov. 06 for not selling 10 cent contracts for GOP retaining the House).

Posted by: Andrew Foland | Feb 13, 2008 5:40:42 PM

I predict that those who predict will be unable to predict.

Posted by: getting older | Feb 13, 2008 9:21:51 PM

BR: FYI - Me Media has now gone global - Your comments were picked up and featured on the Business News Network (BNN) in Canada today

Posted by: Shadow | Feb 13, 2008 10:15:54 PM

Andrew Foland is correct. The folks at Google have made a definitive study of prediction markets by applying them to in house events. See http://googleblog.blogspot.com/2005/09/putting-crowd-wisdom-to-work.html. Basically, they collected and analyzed the results of the prediction process, and determined there was a measurable decrease in the randomness of the predicted events. Most of the other comments here completely miss the point.

Posted by: big99 | Feb 13, 2008 11:24:46 PM