The Democrats' performance versus polls in 2010...
I'm going by just senate races decided by ten-points or less (I did include Kentucky, tho) solely because I think the bigger the margin, the chances of a larger disparity (being predicted to win by 25 and winning by 20 - a five-point swing - is insignificant in that scenario, however, being predicted to win by five and losing by .1% isn't).
So, did the Democrats out-perform polls in 2010?
Well...
California:
Final polls there put the race as a toss-up between Carly Fiorina and Barbara Boxer. The average poll sample had Boxer winning by 5. She eventually won the race over Fiorina by 9. 8 points - a difference of +4.8 points for Boxer.
Connecticut
The final polls there put the race as a lean Dem as Richard Blumenthal led Linda McMahon by 8.7 points. He won by 11.8 - a difference of +3.1 points for Blumenthal.
Kentucky
The final polls there put the race as leans GOP as Rand Paul held a lead of 11 points. He won by 11.6 - a difference of +.6 points for Paul.
Pennsylvania
The final polls there put the race as toss-up as Pat Toomey led Joe Sestak by a 4.5 margin. He won by +2 - a difference of +2.5 for Sestak (not good enough for the win, however).
Washington
The final polls there put the race as a toss-up as Patty Murray led in the polls by just .3%. She would win by +3.8 - a difference of +4.1 Murray.
Colorado
The final polls there had the race as a toss-up as Ken Buck led Michael Bennet by 3.0 points. Bennet would win by +.9 - a difference of +3.9 for Bennet.
West Virginia
The final polls there had the race as a toss-up as Joe Manchin led John Raese by a 4.5 margin. He would win by 10.1 points - a difference of +5.6 for Manchin.
Wisconsin
The final polls there had the race as a toss-up as Ron Johnson led Russ Feingold by a 7.7 margin. He would win by 4.9 points - a difference of +2.8 points for Feingold (not good enough for the win, however).
Nevada
The final polls there had the race as a toss-up as Sharron Angle led Harry Reid by 2.7 points. Reid would actually win by 5.6 points - a difference of +8.3 points for Reid.
Illinois
The final polls there had the race as a toss-up as Mark Kirk led Alexi Giannoulias by 3.3 points. He would win by 1.9 points - a difference of +1.4 for Giannoulias.
Those were the most contested senate races in 2010. The Democrats managed to win states where the polls had them losing in two races (Nevada & Colorado) - and there were zero states where the Dem candidate led and actually lost.
On the whole, the Democratic candidate, whether in a winning or losing effort, out-performed the polls by an average of +3.53.
Currently, these races are within that margin:
Iowa: R+2.1
North Carolina: D+1.6
Georgia: R+0.5
Alaska: R+2.2
New Hampshire: D+2.8
If the Democrats win all those seats, plus the 45 seats they will hold/expected to hold, that would give them a +1 margin (with Biden being the tie-breaker).
I can tolerate that - and that isn't counting other states like Colorado, Kentucky, Kansas, Louisiana and Arkansas - three of which are out of that margin (R+5.5 in KY & R+7 in AR & R+3.6 in CO - as well as Kansas, where the indy says he'll caucus with the majority party).
So, Democrats exceeded the final poll tallies in 2010. If they do so again at the same level, it's very likely they keep the senate.
Polls give you a good idea of where a race is at - it is not a prediction of the exact number, however. 2010 proved this.
We'll see if it holds up again in 2014.
*all polling data from RCP's 2010 average.
= new reply since forum marked as read
GOTV2014
