University of Massachusetts Lowell Center for Public Opinion
Graded against the actual result across 24 races (from 37 polls, through 2022).
Head-to-head vs VotePredictor Elections
The fair, apples-to-apples test: on the 24 races University of Massachusetts Lowell Center for Public Opinion actually polled, how its final poll's margin compared to what VotePredictor Elections predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| University of Massachusetts Lowell Center for Public Opinion | 3.57 | 83% |
| VotePredictor Elections | 3.41 | 88% |
VotePredictor Elections 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 (24)
Each race University of Massachusetts Lowell Center for Public Opinion 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.
| Race | Their call | Result | Miss | Call |
|---|---|---|---|---|
| 2022 MA Governor | D+27.0 | D+29.1 | 2.1 | ✓ |
| 2022 NH Governor | R+16.0 | R+15.5 | 0.5 | ✓ |
| 2022 GA Senate | D+5.0 | D+2.8 | 2.2 | ✓ |
| 2022 NH Senate | D+10.0 | D+9.1 | 0.9 | ✓ |
| 2020 NC Governor | D+12.0 | D+4.5 | 7.5 | ✓ |
| 2020 NH Governor | R+23.0 | R+31.8 | 8.8 | ✓ |
| 2020 NC President | EVEN | R+1.3 | 1.3 | ✗ |
| 2020 NH President | D+10.0 | D+7.4 | 2.6 | ✓ |
| 2020 TX President | R+1.0 | R+5.6 | 4.6 | ✓ |
| 2020 US President | D+10.0 | D+4.4 | 5.6 | ✓ |
| 2020 NC Senate | D+4.0 | R+1.7 | 5.7 | ✗ |
| 2020 NH Senate | D+19.0 | D+15.6 | 3.4 | ✓ |
| 2020 TX Senate | R+5.0 | R+9.6 | 4.6 | ✓ |
| 2018 MA Governor | R+39.0 | R+33.5 | 5.5 | ✓ |
| 2018 MA Senate | D+25.0 | D+24.2 | 0.8 | ✓ |
| 2016 NH Governor | R+4.0 | R+2.3 | 1.7 | ✓ |
| 2016 NH President | EVEN | D+0.4 | 0.4 | ✗ |
| 2016 NH Senate | D+1.0 | D+0.1 | 0.9 | ✓ |
| 2014 MA Governor | R+4.0 | R+1.9 | 2.1 | ✓ |
| 2014 NH Governor | D+4.0 | D+4.9 | 0.9 | ✓ |
| 2014 NH Senate | D+3.0 | D+3.2 | 0.2 | ✓ |
| 2013 MA Senate | D+20.0 | D+10.2 | 9.8 | ✓ |
| 2012 MA President | D+20.0 | D+23.0 | 3.0 | ✓ |
| 2012 MA Senate | R+1.0 | D+7.5 | 8.5 | ✗ |
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 |
|---|---|---|---|---|
| 1–3 wk | 19 | 3.18 | -1.89 | 84% |
| 3–6 wk | 14 | 3.52 | -2.22 | 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) |
|---|---|---|---|
| 2012 | 4 | 5.8 | R+5.8 |
| 2014 | 3 | 1.1 | R+1.1 |
| 2016 | 6 | 1.7 | D+0.6 |
| 2020 | 17 | 4.4 | D+4.3 |
| 2022 | 4 | 1.4 | D+0.1 |
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.