Harper Polling
Graded against the actual result across 16 races (from 22 polls, through 2020).
Head-to-head vs VotePredictor Elections
The fair, apples-to-apples test: on the 16 races Harper Polling actually polled, how its final poll's margin compared to what VotePredictor Elections predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| Harper Polling | 4.49 | 69% |
| VotePredictor Elections | 5.73 | 69% |
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 (16)
Each race Harper Polling 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 |
|---|---|---|---|---|
| 2020 NC Governor | D+10.0 | D+4.5 | 5.5 | ✓ |
| 2020 NC President | D+1.0 | R+1.3 | 2.3 | ✗ |
| 2020 NC Senate | D+3.0 | R+1.7 | 4.7 | ✗ |
| 2018 MI-11 House | D+1.0 | D+6.7 | 5.7 | ✓ |
| 2016 MI-1 House | R+5.0 | R+14.8 | 9.8 | ✓ |
| 2016 PA President | EVEN | R+0.7 | 0.7 | ✗ |
| 2016 PA Senate | EVEN | R+1.4 | 1.4 | ✗ |
| 2014 PA Governor | D+10.0 | D+9.9 | 0.1 | ✓ |
| 2014 NY-1 House | EVEN | R+8.7 | 8.7 | ✗ |
| 2014 NY-21 House | R+14.0 | R+20.5 | 6.5 | ✓ |
| 2014 WV-3 House | R+6.0 | R+10.7 | 4.7 | ✓ |
| 2014 NC Senate | R+2.0 | R+1.6 | 0.4 | ✓ |
| 2014 SD Senate | R+4.0 | R+20.9 | 16.9 | ✓ |
| 2013 VA Governor | D+5.0 | D+2.5 | 2.5 | ✓ |
| 2013 MA Senate | D+12.0 | D+10.2 | 1.8 | ✓ |
| 2013 NJ Senate | D+11.0 | D+10.9 | 0.1 | ✓ |
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 | 4 | 0.67 | -3.57 | 50% |
| 1–3 wk | 6 | 3.50 | -1.56 | 67% |
| 3–6 wk | 6 | 9.70 | +4.00 | 83% |
| 6–9 wk | 6 | 2.87 | -3.27 | 50% |
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) |
|---|---|---|---|
| 2013 | 3 | 1.5 | D+1.5 |
| 2014 | 7 | 7.1 | D+7.0 |
| 2016 | 5 | 3.2 | D+3.2 |
| 2020 | 6 | 3.9 | D+3.9 |
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.