Over at Greg Laden’s blog, he has a post up entitled On the Eve of GOP Debate, Only Two Candidates Matter.
He may be correct, the early polls may be predictive, the GOP may well have settled on two outsider candidates with zero experience in government. But I’m not betting on it, and neither are the people who are actually betting on the outcome of the GOP nomination. When you compare polling averages to the current prices at which prediction contracts are trading for the middle-tier candidates, you get a reasonably strong correlation between polls and prices:
This indicates that the people who are placing bets, on average, believe the polling is fairly indicative of probable outcome of those particular campaigns. When you add in the only two candidates who supposedly matter (because of their superlative poll numbers) the correlation completely falls apart:
For reasons which I will leave open to speculation, those who are betting money on the outcome of this particular nomination process do not (on average) trust the polls when it comes to Carson and Trump in particular. Perhaps bettors are factoring in data not captured in public polling, such as the size of soft- and hard-money political warchests, or the sophistication of ground organization in early voting states, or the probability of a career-ending gaffe.
Your thoughts?
