JMC Analytics
Graded against the actual result across 13 races (from 21 polls, through 2019).
Head-to-head vs VotePredictor Elections
The fair, apples-to-apples test: on the 13 races JMC Analytics actually polled, how its final poll's margin compared to what VotePredictor Elections predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| JMC Analytics | 7.88 | 85% |
| VotePredictor Elections | 7.47 | 92% |
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 (13)
Each race JMC Analytics 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 |
|---|---|---|---|---|
| 2019 LA Governor | D+0.5 | D+2.7 | 2.2 | ✓ |
| 2017 AL Senate | R+5.0 | D+1.6 | 6.6 | ✗ |
| 2016 GA President | R+6.0 | R+5.1 | 0.9 | ✓ |
| 2016 LA President | R+7.0 | R+19.6 | 12.6 | ✓ |
| 2016 NV President | EVEN | D+2.4 | 2.4 | ✗ |
| 2016 GA Senate | R+13.0 | R+13.8 | 0.8 | ✓ |
| 2016 NV Senate | D+2.0 | D+2.4 | 0.4 | ✓ |
| 2015 LA Governor | D+4.0 | D+12.2 | 8.2 | ✓ |
| 2014 LA-6 House | R+26.0 | R+24.9 | 1.1 | ✓ |
| 2014 PA-10 House | R+12.0 | R+37.8 | 25.8 | ✓ |
| 2014 LA Senate | R+15.0 | R+11.9 | 3.1 | ✓ |
| 2012 NY-20 House | D+19.0 | D+34.5 | 15.5 | ✓ |
| 2010 MS-2 House | D+1.0 | D+23.8 | 22.8 | ✓ |
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 | 6 | 6.10 | +1.04 | 67% |
| 3–6 wk | 6 | 4.35 | -1.35 | 100% |
| 6–9 wk | 6 | 12.56 | +6.42 | 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) |
|---|---|---|---|
| 2014 | 3 | 10.0 | D+7.2 |
| 2015 | 3 | 6.6 | D+1.1 |
| 2016 | 8 | 3.7 | D+2.7 |
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.