So far, you've probably are already familiar with some of the more popular electoral vote predictors and election guesstimation sites. Two of the bigger ones are Election Projection (pro-Bush), and Electoral Vote (pro-Kerry). While both these sites seem to be making a solid effort to be, like, fair and stuff, they aren't bastions of statistical accuracy.
But there are some places to get much more accurate and thoroughly reasoned analysis of the elections with some links and current predictions, based on current polling data.
Prof. Andrea Moro, of the University of Minnesota gives Bush a 39.3% chance of winning the electoral vote, versus a 59.4% for Kerry.
Prof. Sam Wang of Princeton pegs the most likely outcome being Kerry with 257 EV, versus Bush with 281 EV.
The Federal Review, gives current chances of a Bush win at 81.7 % (expected to get 291 EV as of right now) versus Kerry at 17.2% (expected to get 247 EV as of right now).
Matthew Hubbard's binomial thingy at CSU Hayward gives us this breakdown: Kerry wins with 281 EV, versus Bush at 257 EV.
Then there's Larry Allen at Arrowhead Engineering, with this thing, chock 'full of graphs, which gives the odds of victory at Bush 77.7% (274 EV) to Kerry 22.3% (244 EV).
Finally, there's this bit of pro-level tea-leaf reading over at USS Clueless. Personally, I find this one to be pretty interesting, by virtue of the fact that the election is being analyzed as a system, rather than a pure stats exercise.
Long and short of it all, nobody's walking away with anything less than 240 electoral votes, and past that you can't predict the future. Except for one thing, given a lot of the crowing and expectations of victory I see on some websites, this is going to hit the press a lot like the Fall of Baghdad hit the Arab street - like a Mack Truck hitting a blind ferret.
(Cross-posted at the Jawa Report)
Launched by Bravo Romeo Delta at October 18, 2004 11:30 PM