VotePredictor
← All pollsters

Lake Research Partners

Graded against the actual result across 37 races (from 41 polls, through 2022).

Races polled
37
Polls
41
Avg miss
7.33 pts
Most recent
2022

Head-to-head vs the VotePredictor model

The fair, apples-to-apples test: on the 32 races Lake Research Partners actually polled, how its final poll's margin compared to what the VotePredictor model predicted for those same races.

ModelAvg miss (pts)Called right
Lake Research Partners7.0566%
VotePredictor3.7684%

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 (36)

Each race Lake Research Partners 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
2022 SD SenateR+13.0R+43.530.5
2020 CA-1 HouseR+4.0R+14.010.0
2020 CA-4 HouseR+4.0R+11.97.9
2020 US PresidentD+3.0D+4.41.4
2018 AK-1 HouseR+4.0R+6.62.6
2018 AZ-8 HouseEVENR+4.74.7
2018 MI-2 HouseR+6.0R+12.36.3
2018 WA-3 HouseD+2.0R+5.37.3
2016 IA-1 HouseR+2.0R+7.75.7
2016 AK PresidentR+1.0R+14.713.7
2014 MI GovernorR+1.0R+4.13.1
2014 KY-6 HouseR+9.0R+20.011.0
2014 MI SenateD+9.0D+13.34.3
2013 SC-1 HouseD+3.0R+12.215.2
2012 WI GovernorEVENR+6.86.8
2012 CA-36 HouseD+6.0D+5.90.1
2012 NC-9 HouseR+1.0R+6.15.1
2012 NY-24 HouseD+2.0D+9.37.3
2012 NH PresidentD+5.0D+5.60.6
2012 OH PresidentD+2.0D+3.01.0
2010 AZ-1 HouseD+3.0R+6.09.0
2010 CA-11 HouseD+10.0D+1.18.9
2008 FL-18 HouseR+7.0R+15.78.7
2008 WA-8 HouseD+7.0R+5.612.6
2006 CA-11 HouseR+1.0D+6.57.5
2006 CA-50 HouseD+7.0R+4.611.6
2006 ID-1 HouseR+6.0R+5.20.8
2006 MI SenateD+18.0D+15.72.3
2004 NC PresidentEVENR+12.412.4
2004 NH PresidentEVEND+1.41.4
2004 OH PresidentEVENR+2.12.1
2004 WI PresidentD+4.0D+0.43.6
2002 CO-4 HouseR+5.0R+13.38.3
2002 NH-1 HouseD+2.0R+19.621.6
2002 AL SenateR+8.0R+18.710.7
2002 ME SenateR+9.0R+16.97.9

Accuracy by time to election

24686–9 wk3–6 wk1–3 wk≤1 wkavg miss (pts)
Lake Research PartnersAll 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 wk83.39-1.6863%
3–6 wk257.12+1.3872%
6–9 wk713.42+7.2771%

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)
2002412.1D+12.1
200444.9D+4.2
200655.0D+1.6
201283.3R+0.3
201447.5D+3.3
201855.6D+5.6
202036.4D+5.5

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.