Marist College
Graded against the actual result across 147 races (from 257 polls, through 2024).
Head-to-head vs the VotePredictor model
The fair, apples-to-apples test: on the 140 races Marist College 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 |
|---|---|---|
| Marist College | 4.60 | 76% |
| VotePredictor | 3.15 | 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 (146)
Each race Marist College 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 | 40 | 4.47 | +0.19 | 75% |
| 1–3 wk | 92 | 4.92 | -0.15 | 77% |
| 3–6 wk | 81 | 5.59 | -0.15 | 82% |
| 6–9 wk | 44 | 4.33 | -1.82 | 89% |
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 | 4 | 11.8 | R+11.8 |
| 2000 | 12 | 6.4 | R+5.8 |
| 2002 | 3 | 5.7 | R+5.7 |
| 2004 | 9 | 5.6 | D+2.9 |
| 2005 | 4 | 4.7 | R+4.7 |
| 2006 | 6 | 10.2 | D+0.5 |
| 2008 | 23 | 3.4 | R+1.8 |
| 2010 | 24 | 7.6 | R+7.4 |
| 2012 | 48 | 3.0 | R+1.2 |
| 2014 | 22 | 6.6 | D+5.5 |
| 2016 | 28 | 4.4 | D+2.9 |
| 2018 | 24 | 4.7 | D+2.3 |
| 2020 | 14 | 6.0 | D+6.0 |
| 2022 | 22 | 3.8 | D+2.6 |
| 2024 | 13 | 3.5 | D+3.4 |
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.