Anyone following the U.S. election horse-race will be familiar with sites like electoral-vote.com, real clear politics, the hedgehog report, and Professor Pollkatz.
Most of these sites (but not Pollkatz) simply average the state polls, and say no more. However, if numerous polls in a short space of time find a candidate has a lead within the margin of error, it's likely that lead is small but real.
The one site that does take these into consideration is Sam Wang's meta-analysis of state polls. I highly recommend checking it out.
Minnesota readers should note that while we're seeing boatloads of ads, and the local media seem to revel in the transition from most longstanding Democratic state to battleground, Wang's site puts the chances of Kerry winning here at near 100%. Indeed, Minnesota is a real example of the small margins replicated in many polls phenomenon.
If you want to use this information to make some money by betting, sorry "trading," on the outcome you're out of luck. The current asking price for a dollar share if Bush wins in Minnesota is already down to 34.5c (By comparison, if you wanted to take a dollar in the event that Bush wins Alabama the asking price is 98c. Bush will lose Alabama when he's caught shagging the Pope in prison, and maybe not even then ...)
Posted by robe0419 at October 11, 2004 4:24 PM