VotePredictor
← All pollsters

MRG Research

Graded against the actual result across 24 races (from 42 polls, through 2020).

Races polled
24
Polls
42
Avg miss
8.02 pts
Most recent
2020

Head-to-head vs the VotePredictor model

The fair, apples-to-apples test: on the 24 races MRG Research actually polled, how its final poll's margin compared to what the VotePredictor model predicted for those same races.

ModelAvg miss (pts)Called right
MRG Research8.3279%
VotePredictor5.8592%

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 (24)

Each race MRG Research 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
2020 HI PresidentD+33.0D+29.53.5
2018 HI GovernorD+21.0D+29.08.0
2014 HI GovernorD+6.0D+12.46.4
2014 HI-1 HouseEVEND+3.93.9
2014 HI-2 HouseD+50.0D+60.110.1
2014 HI SenateD+26.0D+42.116.1
2012 HI-1 HouseD+11.0D+9.21.8
2012 HI-2 HouseD+58.0D+61.13.1
2012 HI PresidentD+27.0D+42.715.7
2012 US PresidentD+3.0D+3.90.9
2012 HI SenateD+15.0D+25.210.2
2010 CT GovernorEVENR+1.71.7
2010 HI GovernorD+5.0D+17.112.1
2010 CT-1 HouseD+18.0D+20.62.6
2010 CT-2 HouseD+19.0D+18.30.7
2010 CT-3 HouseD+18.1D+27.49.3
2010 CT-4 HouseR+2.0D+4.06.0
2010 CT-5 HouseR+8.0D+7.115.1
2010 HI-1 HouseD+5.0D+6.51.5
2010 HI-2 HouseD+23.5D+46.923.4
2010 WA-2 HouseD+18.0D+2.115.9
2010 CT SenateD+8.0D+9.31.3
2010 HI SenateD+32.1D+53.221.1
2010 MA SenateR+9.6R+4.84.8

Accuracy by time to election

24686–9 wk3–6 wk1–3 wk≤1 wkavg miss (pts)
MRG ResearchAll 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 wk228.72+3.6582%
3–6 wk146.92+1.1886%
6–9 wk47.10+0.9575%

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)
2010238.2R+6.2
201297.9R+7.5
201488.1R+8.1

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.