MassINC Polling Group
Graded against the actual result across 19 races (from 43 polls, through 2024).
Head-to-head vs VotePredictor Elections
The fair, apples-to-apples test: on the 19 races MassINC Polling Group actually polled, how its final poll's margin compared to what VotePredictor Elections predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| MassINC Polling Group | 3.23 | 90% |
| VotePredictor Elections | 3.66 | 90% |
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 (19)
Each race MassINC Polling Group 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 |
|---|---|---|---|---|
| 2024 MA President | D+30.0 | D+25.4 | 4.6 | ✓ |
| 2024 MA Senate | D+22.0 | D+19.8 | 2.2 | ✓ |
| 2022 MA Governor | D+30.0 | D+29.1 | 0.9 | ✓ |
| 2020 MA President | D+34.0 | D+33.5 | 0.5 | ✓ |
| 2020 MA Senate | D+31.0 | D+33.1 | 2.1 | ✓ |
| 2018 MA Governor | R+43.0 | R+33.5 | 9.5 | ✓ |
| 2018 MA Senate | D+22.0 | D+24.2 | 2.2 | ✓ |
| 2016 NH Governor | R+5.0 | R+2.3 | 2.7 | ✓ |
| 2016 MA President | D+26.0 | D+27.2 | 1.2 | ✓ |
| 2016 NH President | R+1.0 | D+0.4 | 1.4 | ✗ |
| 2016 NH Senate | R+6.0 | D+0.1 | 6.1 | ✗ |
| 2014 MA Governor | R+1.0 | R+1.9 | 0.9 | ✓ |
| 2014 MA-6 House | D+8.0 | D+13.8 | 5.8 | ✓ |
| 2014 MA Senate | D+25.0 | D+23.9 | 1.1 | ✓ |
| 2013 MA Senate | D+7.0 | D+10.2 | 3.2 | ✓ |
| 2012 MA-6 House | D+12.0 | D+1.1 | 10.9 | ✓ |
| 2012 MA President | D+20.0 | D+23.0 | 3.0 | ✓ |
| 2012 MA Senate | D+6.0 | D+7.5 | 1.5 | ✓ |
| 2010 MA-10 House | D+3.0 | D+4.5 | 1.5 | ✓ |
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 | 14 | 2.61 | -2.46 | 86% |
| 3–6 wk | 17 | 3.93 | -1.81 | 65% |
| 6–9 wk | 10 | 6.09 | -0.06 | 80% |
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 | 9 | 5.4 | R+0.8 |
| 2014 | 11 | 4.7 | D+3.6 |
| 2016 | 11 | 2.7 | D+0.3 |
| 2018 | 4 | 6.0 | R+5.1 |
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.