Since final results have not, as of yet, been certified by the fifty Secretaries of State, it is still too early to compare my final poll averages of my Presidential and Senate forecasts to the final results in those states. The initial estimate seems to be that polling averages performed very well in states where there were a lot of polls, but did not do well in the less frequently polled states of Alaska (President, Senate and House showed massive, double-digit error), Iowa (6% error), Nevada (6-7% error) and North Dakota (6% error). Polling averages also seem to have been about 4-5% off in Arizona and New Mexico. Everywhere else, the averages seem to have nailed the final targets by 2.0% or less, even though inaccurate winners were projected in Indiana and North Carolina. Basically, it seems like the more polls in your averages, the more accurate the averages become. Makes sense.
While we wait for final results in the polling average states, I am happy to say that my House Forecast has, once again, done extremely well. If, as appears likely, MD-01 and VA-05 go to Democrats, and with CA-04, CA-44, LA-04, OH-15, and WA-08 still undecided, then Democrats will net 21-26 seats. My final projection was 21-27 seats, so I am feeling like I did pretty darn well. I also did well for each of the category projections (more in the extended entry):
Likely Democrat: I projected Dems to pickup four, and they pickup up four
Lean Democrat: I projected Dems to pickup 12-13, and with OH-15 undecided we picked up 12-13.
Toss-up: I projected Dems to pickup 6, but they have overperformed and are set to pickup 7-9 with CA-04 and WA-08 till undecided.
Lean Republican: I projected Dems to pick up 5-7, but they have underperformed and are set to pick up only 2-3 with LA-04 still undecided. This does, almost precisely, cancel out the over-performance in the toss-up category.
Likely Republican and Uncompetitive: I projected Democrats to pick up 0-1 seats, and they are set to pickup 0-1 seats with CA-44 outstanding. I admit, I didn't see CA-44 coming at all.
Democratic Held Seats
Likely Republican: I projected Republicans to pick up two, and they picked up two.
Lean Republican: I projected Republicans to pick up one, and they picked up zero. To put it a different way, I picked Kanjorski to lose, but he won.
Lean Democrat: I projected Republicans to pick up 1-2 seats, and they picked up one.
Likely Democrat and uncompetitive: I projected Republicans to pick up 0-1 seats, and they picked up one.
So, I screwed up Kanjorski, undershot the "toss-up" chances for Democratic pickups, and overshot the "Lean Republican" longshots for Democratic pickups. The overall number, however, was right on the money. If we win three of the five still undecided House races, my "best guess" of 24 will have been exactly right.
So, for the second cycle in a row, I feel as though I did very well in the House--as strong, or better, than any other forecaster around. Also, I think simple polling averages will have once again proven to be very effective, but I will wait for the final numbers before comparing the different methodologies. Mainly, I want to compare my results to Pollster.com and fivethirtyeight. If either of those sites outperformed my methodology in terms of average error, than perhaps election forecasting does require something more than my simplicity. Still, given the preliminary results, I doubt they will have outperformed simple polling averages by that much.
Finally, I'd like to figure out what the hell happened in Alaska--seriously, WTF? I'm even kind of surprised to write this, but what sort of voting machines do they use up there?
STATE BLOGS
