Quinnipiac University
Graded against the actual result across 152 races (from 413 polls, through 2025).
Head-to-head vs the VotePredictor model
The fair, apples-to-apples test: on the 136 races Quinnipiac University 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 |
|---|---|---|
| Quinnipiac University | 4.71 | 83% |
| VotePredictor | 3.31 | 87% |
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 (152)
Each race Quinnipiac University 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 | 53 | 4.15 | -0.13 | 76% |
| 1–3 wk | 127 | 4.62 | -0.45 | 80% |
| 3–6 wk | 133 | 5.28 | -0.46 | 84% |
| 6–9 wk | 100 | 6.76 | +0.61 | 85% |
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 | 12 | 11.7 | R+11.0 |
| 2000 | 20 | 5.8 | R+1.2 |
| 2001 | 4 | 2.8 | R+0.5 |
| 2002 | 12 | 5.0 | R+0.9 |
| 2004 | 31 | 4.7 | R+2.5 |
| 2005 | 4 | 3.2 | R+2.5 |
| 2006 | 27 | 6.3 | R+0.7 |
| 2008 | 41 | 4.2 | R+0.7 |
| 2009 | 4 | 3.3 | D+3.1 |
| 2010 | 43 | 5.8 | R+3.1 |
| 2012 | 36 | 3.6 | R+1.0 |
| 2013 | 13 | 4.9 | R+1.7 |
| 2014 | 35 | 5.1 | D+0.8 |
| 2016 | 50 | 4.6 | D+3.6 |
| 2017 | 7 | 4.7 | D+4.1 |
| 2018 | 21 | 4.7 | D+0.2 |
| 2020 | 35 | 8.3 | D+8.3 |
| 2022 | 11 | 4.3 | D+3.4 |
| 2024 | 6 | 2.8 | D+2.8 |
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.