VotePredictor
← All pollsters

InsiderAdvantage

Graded against the actual result across 75 races (from 186 polls, through 2025).

Races polled
75
Polls
186
Avg miss
4.23 pts
Most recent
2025

Head-to-head vs the VotePredictor model

The fair, apples-to-apples test: on the 74 races InsiderAdvantage actually polled, how its final poll's margin compared to what the VotePredictor model predicted for those same races.

ModelAvg miss (pts)Called right
InsiderAdvantage4.6369%
VotePredictor2.8885%

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 (74)

Each race InsiderAdvantage 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
2025 VA GovernorD+10.2D+15.45.2
2024 AZ PresidentR+2.9R+5.52.6
2024 GA PresidentR+1.0R+2.21.2
2024 IA PresidentR+6.8R+13.36.5
2024 MI PresidentD+0.1R+1.41.5
2024 NC PresidentR+2.2R+3.21.0
2024 NV PresidentR+0.2R+3.12.9
2024 PA PresidentR+1.0R+1.70.7
2024 WI PresidentR+1.3R+0.80.5
2024 AZ SenateD+2.1D+2.40.3
2024 MI SenateD+0.6D+0.30.3
2024 NV SenateD+4.3D+1.72.6
2024 PA SenateD+1.1R+0.21.3
2024 WI SenateR+0.5D+0.91.4
2022 AZ GovernorR+3.0D+0.73.7
2022 FL GovernorR+10.0R+19.49.4
2022 GA GovernorR+5.0R+7.52.5
2022 MI GovernorR+0.2D+10.510.7
2022 NH GovernorR+15.3R+15.50.2
2022 NV GovernorR+5.2R+1.53.7
2022 PA GovernorD+7.7D+14.87.1
2022 GA-2 HouseD+3.3D+9.96.6
2022 AZ SenateEVEND+4.94.9
2022 FL SenateR+6.0R+16.410.4
2022 GA SenateR+2.0D+1.03.0
2022 NH SenateD+1.8D+9.17.3
2022 NV SenateR+6.0D+0.86.8
2022 PA SenateR+2.3D+4.97.2
2022 WA SenateD+1.6D+14.512.9
2021 VA GovernorR+2.0R+1.90.1
2020 FL PresidentR+1.2R+3.42.2
2020 GA PresidentR+2.0D+0.22.2
2020 IA PresidentR+1.7R+8.26.5
2020 MI PresidentD+2.1D+2.80.7
2020 NC PresidentR+4.4R+1.33.1
2020 PA PresidentR+1.3D+1.22.5
2020 GA SenateR+0.5D+2.12.6
2020 IA SenateR+5.6R+6.61.0
2018 GA GovernorD+0.9R+1.42.3
2017 GA-6 HouseD+0.3R+3.63.9
2017 AL SenateD+2.3D+1.60.7
2016 FL PresidentD+2.0R+1.23.2
2016 GA PresidentR+3.8R+5.11.3
2016 FL SenateR+3.6R+7.74.1
2016 GA SenateR+10.8R+13.83.0
2014 GA GovernorR+3.0R+7.94.9
2014 GA SenateR+3.0R+7.74.7
2012 FL PresidentR+5.0D+0.95.9
2012 GA PresidentR+21.0R+7.813.2
2010 GA GovernorR+5.8R+10.04.2
2010 SC GovernorR+14.0R+4.59.5
2010 GA SenateR+32.0R+19.312.7
2010 MA SenateR+9.1R+4.84.3
2009 VA GovernorR+3.7R+17.413.7
2008 CO PresidentD+8.0D+9.01.0
2008 FL PresidentD+1.0D+2.81.8
2008 GA PresidentR+1.0R+5.24.2
2008 MI PresidentR+1.0D+16.517.5
2008 MO PresidentR+3.0R+0.12.9
2008 NC PresidentEVEND+0.30.3
2008 NV PresidentEVEND+12.512.5
2008 OH PresidentD+10.0D+4.65.4
2008 PA PresidentD+9.0D+10.31.3
2008 US PresidentEVEND+7.47.4
2008 VA PresidentD+6.0D+6.30.3
2008 WV PresidentR+2.0R+13.111.1
2008 GA SenateR+5.0R+2.92.1
2007 KY GovernorD+10.0D+17.47.4
2006 FL GovernorR+19.0R+7.111.9
2006 GA GovernorR+14.0R+19.75.7
2006 FL-22 HouseR+1.0D+3.84.8
2006 GA-12 HouseD+3.0D+0.62.4
2004 FL PresidentEVENR+5.05.0
2004 FL SenateD+1.0R+1.12.1

Accuracy by time to election

24686–9 wk3–6 wk1–3 wk≤1 wkavg miss (pts)
InsiderAdvantageAll 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 wk493.18-1.1063%
1–3 wk674.72-0.3566%
3–6 wk473.99-1.7585%
6–9 wk235.50-0.6561%

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)
200493.9D+3.7
2006104.7D+1.3
2008445.1R+2.0
201095.7R+0.3
201237.1R+5.7
201485.6D+4.9
2016153.3D+3.3
2020203.0D+0.2
2022384.9R+3.4
2024221.5D+1.1

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.