Using betting odds as a guide to political prospects is certainly a worthwhile idea. I often monitor them here in the UK, where I live.
But it is worth mentioning a few things:
- Betting odds will be influenced by opinion polls to some extent, though not vice-versa to any appreciable extent.
- Betting activity also influences odds. If a lot of deluded people start pouring big amounts of money into bets for Kanye West, the odds on Kanye West will go down as part of a loss-minimisation strategy, even though the likelihood of Kanye West winning may be no greater.
- Odds tend to overstate what the bookmakers believe to be the actual chance of winning (in other words, if they are offering 10/1 on a certain candidate, they may believe that candidate *actually* has a 1/15 chance).
- This last one is just a guess, but I suspect very high odds on extremely unlikely choices are not calculated very rigorously, since there is no expectation of actually having to pay out on them.