Thursday, January 3, 2008

Whats wrong with Intrade?

I'm watching the GOP Primaries futures on Intrade and scratching my head. Then I realize that the number of contracts and the volume are woefully shallow. Four years ago, I found Tradesports (the predecessor of Intrade) to be highly predictive in the last couple of days. But the volume and number of contracts were 10 times greater.

Prediction whether by polling or by profit/loss inducement require broad and statistically representative participation. The current future contracts represented on Intrade offer neither.

Also I postulate one other aspect is skewing predicative accuracy. On a parimutuel wager in Horse Racing, odds are determined by the wagers themselves. The more money laid down on a horse, the better odds are for that horse. The lesser money bet, the greater the odds against winning. Many wagers are pure gambles hoping to beat the odds. Hence after following a large number of races, the wagered odds don't stack up well to the actual results. This same gambling effect may be skewing the accuracy of the primary future contracts.

No comments: