Stockton University William J. Hughes Center for Public Policy
Graded against the actual result across 19 races (from 30 polls, through 2021).
Head-to-head vs VotePredictor Elections
The fair, apples-to-apples test: on the 19 races Stockton University William J. Hughes Center for Public Policy actually polled, how its final poll's margin compared to what VotePredictor Elections predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| Stockton University William J. Hughes Center for Public Policy | 4.09 | 90% |
| VotePredictor Elections | 4.29 | 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 Stockton University William J. Hughes Center for Public Policy 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 |
|---|---|---|---|---|
| 2021 NJ Governor | D+9.0 | D+3.2 | 5.8 | ✓ |
| 2020 NJ-2 House | D+1.0 | R+5.8 | 6.8 | ✗ |
| 2020 NJ President | D+20.5 | D+15.9 | 4.6 | ✓ |
| 2020 NJ Senate | D+25.0 | D+16.3 | 8.7 | ✓ |
| 2018 NJ-2 House | D+17.0 | D+7.7 | 9.3 | ✓ |
| 2018 NJ-3 House | R+1.4 | D+1.3 | 2.7 | ✗ |
| 2018 NJ Senate | D+12.0 | D+11.2 | 0.8 | ✓ |
| 2017 NJ Governor | D+14.0 | D+14.1 | 0.1 | ✓ |
| 2016 NJ-2 House | R+25.5 | R+22.0 | 3.5 | ✓ |
| 2016 NJ President | D+11.0 | D+14.1 | 3.1 | ✓ |
| 2014 NJ-2 House | R+18.0 | R+24.2 | 6.2 | ✓ |
| 2014 NJ-3 House | R+5.0 | R+9.6 | 4.6 | ✓ |
| 2014 NJ Senate | D+24.0 | D+13.5 | 10.5 | ✓ |
| 2013 NJ Governor | R+24.0 | R+22.1 | 1.9 | ✓ |
| 2013 NJ Senate | D+11.0 | D+10.9 | 0.1 | ✓ |
| 2012 NJ-2 House | R+20.0 | R+17.4 | 2.6 | ✓ |
| 2012 NJ-3 House | R+10.0 | R+8.9 | 1.1 | ✓ |
| 2012 NJ President | D+15.0 | D+17.7 | 2.7 | ✓ |
| 2012 NJ Senate | D+22.0 | D+19.5 | 2.5 | ✓ |
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 | 12 | 4.39 | -0.68 | 92% |
| 3–6 wk | 13 | 6.76 | +1.02 | 92% |
| 6–9 wk | 5 | 6.80 | +0.65 | 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 | 4 | 2.3 | R+1.0 |
| 2013 | 4 | 5.8 | D+1.9 |
| 2014 | 7 | 7.9 | D+6.5 |
| 2016 | 3 | 4.8 | R+4.8 |
| 2018 | 5 | 7.5 | D+2.7 |
| 2020 | 3 | 6.7 | D+6.7 |
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.