VotePredictor
← All pollsters

ccAdvertising

Graded against the actual result across 82 races (from 83 polls, through 2014).

Races polled
82
Polls
83
Avg miss
7.21 pts
Most recent
2014

Head-to-head vs the VotePredictor model

The fair, apples-to-apples test: on the 82 races ccAdvertising actually polled, how its final poll's margin compared to what the VotePredictor model predicted for those same races.

ModelAvg miss (pts)Called right
ccAdvertising7.3182%
VotePredictor4.8489%

VotePredictor aggregates all the pollsters, so it's expected to beat any single one on margin — that's the value of averaging. The honest comparison among forecasters is on the combined board.

Every race (82)

Each race ccAdvertising polled, scored on its final poll — the call right before the vote — against the actual Dem−Rep result. Click a race for its full detail.

RaceTheir callResultMissCall
2014 CO SenateR+6.0R+1.94.1
2014 KY SenateR+9.0R+15.56.5
2014 MI SenateR+1.0D+13.314.3
2014 SD SenateR+22.0R+20.91.1
2013 SC-1 HouseD+0.5R+12.212.7
2013 MA SenateD+4.2D+10.26.0
2010 AR GovernorD+11.7D+30.819.1
2010 CA GovernorR+4.4D+12.917.3
2010 CO GovernorD+26.4D+39.913.5
2010 NH GovernorD+6.0D+7.61.6
2010 NV GovernorR+16.0R+11.74.3
2010 OH GovernorR+3.0R+2.01.0
2010 OR GovernorR+3.0D+1.54.5
2010 PA GovernorR+6.9R+9.02.1
2010 TX GovernorR+14.3R+12.71.6
2010 WI GovernorR+9.4R+5.83.6
2010 AR-1 HouseR+5.5R+8.32.8
2010 AR-2 HouseR+21.3R+19.61.7
2010 AZ-1 HouseR+19.9R+6.013.9
2010 AZ-5 HouseR+16.1R+8.87.3
2010 AZ-8 HouseR+0.2D+1.51.7
2010 CA-11 HouseR+8.2D+1.19.3
2010 CA-20 HouseR+23.3D+3.426.7
2010 CA-47 HouseR+16.4D+13.730.1
2010 CO-1 HouseD+7.4D+38.731.3
2010 CO-3 HouseR+23.4R+4.319.1
2010 CO-4 HouseR+12.4R+11.11.3
2010 CO-7 HouseD+7.3D+11.74.4
2010 DE-1 HouseD+12.0D+15.73.7
2010 FL-2 HouseR+27.1R+12.214.9
2010 FL-22 HouseR+6.4R+8.72.3
2010 FL-24 HouseR+17.2R+19.32.1
2010 FL-8 HouseR+21.3R+17.93.4
2010 HI-1 HouseR+18.1R+8.79.4
2010 IL-10 HouseR+9.4R+2.27.2
2010 IL-11 HouseR+12.3R+14.72.4
2010 IL-14 HouseR+8.0R+6.31.7
2010 IL-17 HouseR+18.5R+9.68.9
2010 NC-11 HouseD+0.5D+8.78.2
2010 NC-2 HouseR+18.0R+0.817.2
2010 NC-7 HouseD+2.8D+7.44.6
2010 NC-8 HouseD+2.1D+9.37.2
2010 NH-1 HouseR+6.0R+11.65.6
2010 NH-2 HouseR+10.0R+1.68.4
2010 NM-1 HouseR+7.5D+3.611.1
2010 NM-2 HouseR+20.1R+10.89.3
2010 NV-3 HouseR+9.0R+0.78.3
2010 NY-1 HouseD+0.2D+9.99.7
2010 NY-19 HouseR+0.6D+4.75.3
2010 NY-24 HouseR+2.8D+2.14.9
2010 OH-13 HouseD+4.9D+11.56.6
2010 OH-16 HouseR+14.4R+10.83.6
2010 OH-18 HouseR+13.1R+13.40.3
2010 OR-1 HouseD+6.4D+12.86.4
2010 OR-5 HouseD+3.0D+5.32.3
2010 PA-11 HouseR+11.8R+9.42.4
2010 PA-12 HouseD+0.1D+1.61.5
2010 PA-17 HouseD+3.9D+11.07.1
2010 PA-3 HouseR+13.8R+11.42.4
2010 PA-7 HouseR+4.9R+11.06.1
2010 TX-17 HouseR+11.0R+25.214.2
2010 TX-23 HouseR+11.0R+5.06.0
2010 TX-27 HouseR+10.0R+0.79.3
2010 WA-2 HouseR+3.2D+2.15.3
2010 WA-3 HouseR+7.6R+5.91.7
2010 WA-9 HouseR+0.8D+9.710.5
2010 WI-7 HouseR+15.9R+7.78.2
2010 WI-8 HouseR+10.0R+9.60.4
2010 WV-1 HouseR+6.4R+0.85.6
2010 WV-3 HouseD+3.6D+12.18.5
2010 AR SenateR+14.6R+21.06.4
2010 CA SenateR+4.1D+10.014.1
2010 CO SenateD+2.3D+1.60.7
2010 DE SenateD+15.3D+16.61.3
2010 NH SenateR+22.7R+23.20.5
2010 NV SenateR+5.9D+5.711.6
2010 OH SenateR+18.3R+17.40.9
2010 OR SenateD+13.7D+18.04.3
2010 PA SenateR+7.6R+2.05.6
2010 WA SenateD+0.2D+4.74.5
2010 WI SenateR+7.3R+4.82.5
2010 WV SenateD+3.2D+10.16.9

Accuracy by time to election

24686–9 wk3–6 wk1–3 wk≤1 wkavg miss (pts)
ccAdvertisingAll pollsters (field average)

Lower is better. Time to election runs right (election week) to left (~2 months out).

By the numbers

Time to electionPollsAvg missvs fieldCalled right
1–3 wk277.14+2.0785%
3–6 wk507.18+1.4478%
6–9 wk57.49+1.3480%

vs field is this pollster's average miss minus all pollsters' at the same lead time — green beats the field, redtrails it. Our historical polls reach ~2 months out; earlier polling isn't in the record.

Track record by cycle — getting better?

YearPollsAvg missLean (house effect)
2010777.2R+6.0
201446.5R+3.3

Do we credit a pollster for fixing its bias? Each cycle, the model re-estimates every pollster's lean from all its earlier polls (walk-forward) and subtracts it before using the poll. We tested weighting recent cycles more — it doesn't help: a pollster's lean in one cycle barely predicts the next (correlation 0.28), so the swings above are mostly noise, and averaging over more history beats chasing the latest cycle. The all-time estimate we use came out within ~0.5% of the best option.