Columbus Dispatch
Graded against the actual result across 17 races (from 29 polls, through 2016).
Head-to-head vs VotePredictor Elections
The fair, apples-to-apples test: on the 12 races Columbus Dispatch actually polled, how its final poll's margin compared to what VotePredictor Elections predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| Columbus Dispatch | 3.77 | 83% |
| VotePredictor Elections | 3.05 | 100% |
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 (17)
Each race Columbus Dispatch 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 |
|---|---|---|---|---|
| 2016 OH President | D+1.0 | R+8.1 | 9.1 | ✗ |
| 2016 OH Senate | R+21.0 | R+20.9 | 0.1 | ✓ |
| 2014 OH Governor | R+28.0 | R+30.6 | 2.6 | ✓ |
| 2012 OH President | D+2.0 | D+3.0 | 1.0 | ✓ |
| 2012 OH Senate | D+6.0 | D+6.0 | 0.0 | ✓ |
| 2010 OH Governor | R+2.0 | R+2.0 | 0.0 | ✓ |
| 2010 OH Senate | R+16.0 | R+17.4 | 1.4 | ✓ |
| 2008 OH President | D+6.0 | D+4.6 | 1.4 | ✓ |
| 2006 OH Governor | D+36.0 | D+23.9 | 12.1 | ✓ |
| 2006 OH Senate | D+24.0 | D+12.3 | 11.7 | ✓ |
| 2004 OH President | EVEN | R+2.1 | 2.1 | ✗ |
| 2004 OH Senate | R+24.0 | R+27.7 | 3.7 | ✓ |
| 2002 OH Governor | R+18.0 | R+19.5 | 1.5 | ✓ |
| 2000 OH President | R+10.0 | R+3.5 | 6.5 | ✓ |
| 2000 OH Senate | R+27.0 | R+24.1 | 2.9 | ✓ |
| 1998 OH Governor | R+1.0 | R+5.4 | 4.4 | ✓ |
| 1998 OH Senate | R+2.0 | R+12.9 | 10.9 | ✓ |
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 | 17 | 4.20 | -0.87 | 88% |
| 3–6 wk | 8 | 4.25 | -1.49 | 100% |
| 6–9 wk | 4 | 3.45 | -2.70 | 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) |
|---|---|---|---|
| 1998 | 3 | 5.5 | D+4.7 |
| 2000 | 4 | 3.0 | R+3.0 |
| 2004 | 4 | 4.1 | D+1.7 |
| 2006 | 4 | 9.0 | D+2.9 |
| 2012 | 4 | 2.8 | D+2.3 |
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.