SurveyUSA
Graded against the actual result across 532 races (from 1371 polls, through 2024).
Head-to-head vs the VotePredictor model
The fair, apples-to-apples test: on the 418 races SurveyUSA actually polled, how its final poll's margin compared to what the VotePredictor model predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| SurveyUSA | 4.24 | 87% |
| VotePredictor | 3.59 | 89% |
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 (523)
Each race SurveyUSA 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.
Accuracy by time to election
Lower is better. Time to election runs right (election week) to left (~2 months out).
By the numbers
| Time to election | Polls | Avg miss | vs field | Called right |
|---|---|---|---|---|
| ≤1 wk | 231 | 3.71 | -0.57 | 87% |
| 1–3 wk | 457 | 4.48 | -0.59 | 86% |
| 3–6 wk | 494 | 5.25 | -0.49 | 84% |
| 6–9 wk | 189 | 6.43 | +0.28 | 79% |
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?
| Year | Polls | Avg miss | Lean (house effect) |
|---|---|---|---|
| 1998 | 18 | 3.3 | R+1.3 |
| 2000 | 114 | 4.9 | R+2.4 |
| 2002 | 134 | 6.0 | D+4.2 |
| 2003 | 14 | 5.6 | D+2.7 |
| 2004 | 244 | 4.2 | D+0.2 |
| 2005 | 6 | 3.6 | R+3.6 |
| 2006 | 179 | 4.7 | R+1.5 |
| 2007 | 7 | 3.1 | D+2.6 |
| 2008 | 232 | 4.6 | R+1.8 |
| 2009 | 11 | 1.9 | D+1.5 |
| 2010 | 96 | 6.4 | R+2.8 |
| 2012 | 79 | 4.0 | R+2.8 |
| 2014 | 73 | 6.8 | D+5.1 |
| 2016 | 42 | 5.4 | D+2.3 |
| 2018 | 30 | 5.2 | R+1.9 |
| 2020 | 43 | 4.5 | D+3.0 |
| 2022 | 27 | 3.4 | D+1.3 |
| 2024 | 14 | 5.1 | D+4.8 |
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.