VotePredictor
← All pollsters

ABC News/The Washington Post

Graded against the actual result across 48 races (from 111 polls, through 2020).

Races polled
48
Polls
111
Avg miss
3.46 pts
Most recent
2020

Head-to-head vs the VotePredictor model

The fair, apples-to-apples test: on the 42 races ABC News/The Washington Post actually polled, how its final poll's margin compared to what the VotePredictor model predicted for those same races.

ModelAvg miss (pts)Called right
ABC News/The Washington Post3.9179%
VotePredictor2.5795%

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 (48)

Each race ABC News/The Washington Post 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 AZ PresidentR+1.0D+0.31.3
2020 FL PresidentR+2.0R+3.41.4
2020 MI PresidentD+7.0D+2.84.2
2020 MN PresidentD+16.0D+7.18.9
2020 NC PresidentD+1.0R+1.32.3
2020 PA PresidentD+7.0D+1.25.8
2020 US PresidentD+12.0D+4.47.6
2020 WI PresidentD+17.0D+0.616.4
2020 AZ SenateD+1.0D+2.31.3
2020 MI SenateD+6.0D+1.74.3
2020 NC SenateD+2.0R+1.73.7
2016 US PresidentD+4.0D+2.21.8
2013 VA GovernorD+12.0D+2.59.5
2012 FL PresidentD+4.0D+0.93.1
2012 MD PresidentD+24.0D+26.12.1
2012 OH PresidentD+8.0D+3.05.0
2012 US PresidentD+3.0D+3.90.9
2012 VA PresidentD+4.0D+3.90.1
2012 FL SenateD+14.0D+13.01.0
2012 MD SenateD+31.0D+29.61.4
2012 OH SenateD+12.0D+6.06.0
2012 VA SenateD+7.0D+5.91.1
2010 MD GovernorD+14.0D+14.50.5
2010 MD SenateD+37.0D+26.410.6
2009 VA GovernorR+11.0R+17.46.4
2008 OH PresidentD+6.0D+4.61.4
2008 US PresidentD+9.0D+7.41.6
2008 VA PresidentD+8.0D+6.31.7
2008 VA SenateD+31.0D+31.30.3
2006 MD GovernorD+10.0D+6.53.5
2006 MD SenateD+11.0D+10.01.0
2006 VA SenateR+2.0D+0.42.4
2005 VA GovernorD+3.0D+5.72.7
2004 FL PresidentEVENR+5.05.0
2004 OH PresidentD+3.0R+2.15.1
2004 PA PresidentR+3.0D+2.55.5
2004 US PresidentR+1.0R+2.41.4
2004 WI PresidentR+10.0D+0.410.4
2004 FL SenateEVENR+1.11.1
2004 OH SenateR+25.0R+27.72.7
2004 PA SenateR+17.0R+10.66.4
2004 WI SenateD+6.0D+11.25.2
2002 MD GovernorEVENR+3.93.9
2001 VA GovernorD+10.0D+5.14.9
2000 US PresidentR+3.0D+0.53.5
2000 VA PresidentR+4.0R+8.04.0
2000 VA SenateR+2.0R+4.62.6
1998 MD GovernorD+5.0D+10.35.3

Accuracy by time to election

24686–9 wk3–6 wk1–3 wk≤1 wkavg miss (pts)
ABC News/The Washington PostAll 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 wk72.32-1.9657%
1–3 wk473.20-1.8775%
3–6 wk313.21-2.5368%
6–9 wk264.53-1.6285%

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)
2000193.1R+1.6
2004213.3R+1.2
200642.3R+0.1
2008152.5R+0.9
200939.4D+9.4
201045.0D+3.0
2012182.5D+0.1
201661.6D+1.1
2020154.8D+4.4

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.