3 ms·

Election Betting Odds

3 points by sndean 10y ago

ocdtrekkie 10y ago
The Presidential election odds seem about on point. Maybe giving more of a chance to others than warranted for the nominations. But the 74% chance for Hillary for the Presidency seems pretty spot on, where I'd guess it. The VICE President page is super interesting, because it's far less clear cut. There's a lot of options on the table, and nobody has a clear majority. Obviously Warren is a good choice for appeasing Sanders fans, Sanders himself has a tiny chance, unsurprisingly. Republican side I feel is super early to tell, Trump could go anywhere (it's Trump, I mean, come on). Gingrich is listed at the top on this page right now, which is an interesting choice. When I saw Gingrich interviewed, I thought he was a pretty clever guy, he could be a reasonable voice working in a Trump White House, until he gave up trying to bring order to chaos.
bostik 10y ago
This is a fabulous, if trivial, example of what proper prediction markets can do. We run a betting exchange, and one of our long term goals is to make universal prediction markets(+) a reality. These are pretty much the crystallisation of "Seldon Mathematics" - a way to accurately predict the probability of a sociologically meaningful event happening. Now, there are of course caveats. Prediction markets depend on having sufficiently large numbers of people making and taking bets on the event. Because in the world of betting the prediction markets are (at best) a niche of a niche, to become accurate these things require user base of tens to hundreds of millions of people. At that scale the tail end of user base - the people who do happen to care about their particular prediction markets - becomes large enough. Interestingly enough, prediction markets could also allow to experiment with certain theoretical market making models. The markets run for a very long time, and for that reason WILL be affected by all kinds of external factors. My personal favourite is something along the lines of liquidity-sensitive pricing [0]. EDIT: link layout fixed +: https://en.wikipedia.org/wiki/Prediction_market https://en.wikipedia.org/wiki/Prediction_market 0: https://www.semanticscholar.org/paper/A-Practical-Liquidity-Sensitive-Automated-Market-Othman-Pennock/72de958e664c0d3d053de62343a646b3da83f8ab https://www.semanticscholar.org/paper/A-Practical-Liquidity-...

3 points by sndean 10y ago

The Presidential election odds seem about on point. Maybe giving more of a chance to others than warranted for the nominations. But the 74% chance for Hillary for the Presidency seems pretty spot on, where I'd guess it. The VICE President page is super interesting, because it's far less clear cut. There's a lot of options on the table, and nobody has a clear majority. Obviously Warren is a good choice for appeasing Sanders fans, Sanders himself has a tiny chance, unsurprisingly. Republican side I feel is super early to tell, Trump could go anywhere (it's Trump, I mean, come on). Gingrich is listed at the top on this page right now, which is an interesting choice. When I saw Gingrich interviewed, I thought he was a pretty clever guy, he could be a reasonable voice working in a Trump White House, until he gave up trying to bring order to chaos.

bostik 10y ago

This is a fabulous, if trivial, example of what proper prediction markets can do. We run a betting exchange, and one of our long term goals is to make universal prediction markets(+) a reality. These are pretty much the crystallisation of "Seldon Mathematics" - a way to accurately predict the probability of a sociologically meaningful event happening. Now, there are of course caveats. Prediction markets depend on having sufficiently large numbers of people making and taking bets on the event. Because in the world of betting the prediction markets are (at best) a niche of a niche, to become accurate these things require user base of tens to hundreds of millions of people. At that scale the tail end of user base - the people who do happen to care about their particular prediction markets - becomes large enough. Interestingly enough, prediction markets could also allow to experiment with certain theoretical market making models. The markets run for a very long time, and for that reason WILL be affected by all kinds of external factors. My personal favourite is something along the lines of liquidity-sensitive pricing [0]. EDIT: link layout fixed +: https://en.wikipedia.org/wiki/Prediction_market https://en.wikipedia.org/wiki/Prediction_market 0: https://www.semanticscholar.org/paper/A-Practical-Liquidity-Sensitive-Automated-Market-Othman-Pennock/72de958e664c0d3d053de62343a646b3da83f8ab https://www.semanticscholar.org/paper/A-Practical-Liquidity-...