Hendrix College
Graded against the actual result across 26 races (from 32 polls, through 2022).
Head-to-head vs the VotePredictor model
The fair, apples-to-apples test: on the 23 races Hendrix 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 |
|---|---|---|
| Hendrix College | 7.28 | 96% |
| VotePredictor | 6.61 | 96% |
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 (23)
Each race Hendrix 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.
| Race | Their call | Result | Miss | Call |
|---|---|---|---|---|
| 2022 AR Governor | R+10.0 | R+27.8 | 17.8 | ✓ |
| 2022 AR Senate | R+20.0 | R+34.6 | 14.6 | ✓ |
| 2020 AR-2 House | R+0.5 | R+10.7 | 10.2 | ✓ |
| 2020 AR President | R+24.5 | R+27.6 | 3.1 | ✓ |
| 2018 AR Governor | R+36.0 | R+33.6 | 2.4 | ✓ |
| 2018 AR-2 House | R+12.0 | R+6.3 | 5.7 | ✓ |
| 2016 AR President | R+23.5 | R+26.9 | 3.4 | ✓ |
| 2016 AR Senate | R+18.0 | R+23.6 | 5.6 | ✓ |
| 2014 AR Governor | R+8.0 | R+13.9 | 5.9 | ✓ |
| 2014 AR-1 House | R+22.0 | R+30.9 | 8.9 | ✓ |
| 2014 AR-2 House | D+4.5 | R+8.3 | 12.8 | ✗ |
| 2014 AR-4 House | R+2.0 | R+11.2 | 9.2 | ✓ |
| 2014 AR Senate | R+8.5 | R+17.1 | 8.6 | ✓ |
| 2012 AR-1 House | R+25.0 | R+17.1 | 7.9 | ✓ |
| 2012 AR-2 House | R+20.0 | R+15.7 | 4.3 | ✓ |
| 2012 AR-4 House | R+29.0 | R+22.8 | 6.2 | ✓ |
| 2012 AR President | R+21.0 | R+23.7 | 2.7 | ✓ |
| 2010 AR Governor | D+16.0 | D+30.8 | 14.8 | ✓ |
| 2010 AR-1 House | R+8.0 | R+8.3 | 0.3 | ✓ |
| 2010 AR-2 House | R+12.0 | R+19.6 | 7.6 | ✓ |
| 2010 AR-3 House | R+38.0 | R+44.9 | 6.9 | ✓ |
| 2010 AR-4 House | D+18.0 | D+17.4 | 0.6 | ✓ |
| 2010 AR Senate | R+13.0 | R+21.0 | 8.0 | ✓ |
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 | 16 | 7.16 | +2.09 | 94% |
| 3–6 wk | 5 | 9.54 | +3.80 | 100% |
| 6–9 wk | 11 | 9.18 | +3.03 | 100% |
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) |
|---|---|---|---|
| 2010 | 7 | 6.3 | D+0.4 |
| 2012 | 5 | 7.8 | R+3.9 |
| 2014 | 6 | 9.9 | D+9.1 |
| 2016 | 4 | 4.3 | D+3.1 |
| 2020 | 4 | 6.1 | D+7.5 |
| 2022 | 4 | 17.7 | D+17.7 |
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.