The Race for the Senate

[jpmassar]

Nate Silver's new Senate results confirm my 'good day for Democrats' post of yesterday with a slight rise in Democratic fortunes, from a projected loss of 7.5 seats last week to slightly less than 7 now. My, Intrade and Silver's estimates are converging as the race proceeds into its final weeks. Our biggest discrepancy is still Kentucky, where I am still holding out hope (and therefore probability) that Rand Paul will commit an unforgivable gaffe, while Silver is (I assume) pretty much looking straight at the polls and the lay of Kentucky's electorate.

Ignore the last row of the first table, and the last column of the second table. The first column of the first table represents expected number of Republican seat gains. The second table contains probabilities that each seat will switch parties.

Net R+R+D+

JP 5.95 6.90 0.95

InTrade 6.65 7.75 1.10

Silver 6.80 7.30 0.50

StochDem 4.50 5.70 1.20

StateNowJPInTradeSilverStochDem

AR D 100% 95% 100% 100%

CA D 15% 25% 15% 10%

CO D 65% 65% 70% 65%

CT D 5% 15% 0% 0%

DE D 0% 10% 0% 10%

IL D 45% 55% 45% 25%

IN D 100% 95% 100% 100%

ND D 100% 100% 100% 100%

NV D 40% 50% 55% 30%

NY D 0% 10% 0% 0%

PA D 75% 85% 90% 80%

WA D 25% 35% 20% 30%

WI D 80% 85% 95% 20%

WV D 40% 50% 40% 0%

----

AK R 15% 15% 10% 0%

AZ R 0% 5% 0% 0%

FL R 5% 15% 10% 40%

IA R 0% 20% 0% 0%

KY R 35% 20% 15% 20%

LA R 0% 5% 0% 0%

MO R 10% 10% 5% 20%

NH R 20% 10% 10% 0%

NC R 5% 5% 0% 15%

OH R 5% 5% 0% 25%