VotePredictor
← All pollsters

EPIC-MRA

Graded against the actual result across 38 races (from 98 polls, through 2024).

Races polled
38
Polls
98
Avg miss
5.24 pts
Most recent
2024

Head-to-head vs the VotePredictor model

The fair, apples-to-apples test: on the 26 races EPIC-MRA actually polled, how its final poll's margin compared to what the VotePredictor model predicted for those same races.

ModelAvg miss (pts)Called right
EPIC-MRA4.2589%
VotePredictor4.1292%

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 (38)

Each race EPIC-MRA 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
2024 MI PresidentD+3.0R+1.44.4
2024 MI SenateD+5.0D+0.34.7
2022 MI GovernorD+11.0D+10.50.5
2020 MI PresidentD+7.0D+2.84.2
2020 MI SenateD+5.0D+1.73.3
2018 MI GovernorD+5.0D+9.64.6
2018 MI SenateD+7.0D+6.50.5
2016 MI PresidentD+4.0R+0.24.2
2014 MI GovernorR+2.0R+4.12.1
2014 MI SenateD+15.0D+13.31.7
2012 MI PresidentD+6.0D+9.53.5
2012 MI SenateD+21.0D+20.80.2
2010 MI GovernorR+18.0R+18.20.2
2010 IN-2 HouseD+5.0D+1.33.7
2010 IN-7 HouseD+13.0D+21.18.1
2010 MI-1 HouseR+2.0R+11.19.1
2010 MI-15 HouseD+17.0D+16.70.3
2010 MI-3 HouseR+9.0R+22.213.2
2010 MI-7 HouseD+6.0R+4.810.8
2010 MI-9 HouseD+5.0D+2.52.5
2010 IN SenateR+18.0R+14.63.4
2008 MI PresidentD+12.0D+16.54.5
2008 MI SenateD+18.0D+28.810.8
2006 MI GovernorD+9.0D+14.15.1
2006 MI SenateD+13.0D+15.72.7
2004 MI PresidentD+6.0D+3.42.6
2002 MI GovernorD+9.0D+4.05.0
2002 MI-11 HouseR+15.0R+17.52.5
2002 MI SenateD+30.0D+22.77.3
2000 MI-8 HouseD+1.0R+0.11.1
2000 IL PresidentD+15.0D+12.03.0
2000 MI PresidentD+1.0D+5.14.1
2000 OH PresidentR+2.0R+3.51.5
2000 PA PresidentD+18.0D+4.213.8
2000 MI SenateR+5.0D+1.66.6
2000 OH SenateR+21.0R+24.13.1
2000 PA SenateR+4.0R+6.92.9
1998 MI GovernorR+32.0R+24.47.6

Accuracy by time to election

24686–9 wk3–6 wk1–3 wk≤1 wkavg miss (pts)
EPIC-MRAAll 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 wk55.45+1.1760%
1–3 wk454.56-0.5182%
3–6 wk266.17+0.4381%
6–9 wk225.49-0.6686%

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)
2000187.0R+4.2
2002106.7D+6.7
2006114.7R+4.1
200875.9R+5.9
2010145.3D+1.5
201273.5R+3.4
201462.7R+1.4
201646.5D+6.5
201845.8D+2.7
202084.5D+4.5
202241.7D+1.7

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.