Los Angeles Times
Graded against the actual result across 27 races (from 35 polls, through 2008).
Head-to-head vs the VotePredictor model
The fair, apples-to-apples test: on the 16 races Los Angeles Times 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 |
|---|---|---|
| Los Angeles Times | 4.02 | 81% |
| VotePredictor | 1.97 | 100% |
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 (27)
Each race Los Angeles Times 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 |
|---|---|---|---|---|
| 2008 FL President | D+9.0 | D+2.8 | 6.2 | ✓ |
| 2008 OH President | D+5.0 | D+4.6 | 0.4 | ✓ |
| 2008 US President | D+8.0 | D+7.4 | 0.6 | ✓ |
| 2006 CA Governor | R+17.0 | R+17.0 | 0.0 | ✓ |
| 2006 CA Senate | D+18.0 | D+24.4 | 6.4 | ✓ |
| 2006 MO Senate | R+3.0 | D+2.3 | 5.3 | ✗ |
| 2006 NJ Senate | D+4.0 | D+16.0 | 12.0 | ✓ |
| 2006 OH Senate | D+8.0 | D+12.3 | 4.3 | ✓ |
| 2006 TN Senate | R+5.0 | R+2.7 | 2.3 | ✓ |
| 2006 VA Senate | D+3.0 | D+0.4 | 2.6 | ✓ |
| 2004 CA President | D+17.0 | D+9.9 | 7.1 | ✓ |
| 2004 FL President | R+8.0 | R+5.0 | 3.0 | ✓ |
| 2004 OH President | D+6.0 | R+2.1 | 8.1 | ✗ |
| 2004 PA President | EVEN | D+2.5 | 2.5 | ✗ |
| 2004 US President | R+1.0 | R+2.4 | 1.4 | ✓ |
| 2004 CA Senate | D+22.0 | D+19.9 | 2.1 | ✓ |
| 2003 CA Governor | R+8.0 | R+17.1 | 9.1 | ✓ |
| 2002 CA Governor | D+9.0 | D+4.9 | 4.1 | ✓ |
| 2000 CA President | D+7.0 | D+11.8 | 4.8 | ✓ |
| 2000 FL President | R+4.0 | EVEN | 4.0 | ✓ |
| 2000 MI President | D+4.0 | D+5.1 | 1.1 | ✓ |
| 2000 PA President | R+2.0 | D+4.2 | 6.2 | ✗ |
| 2000 US President | R+6.0 | D+0.5 | 6.5 | ✗ |
| 2000 CA Senate | D+25.0 | D+19.3 | 5.7 | ✓ |
| 1998 CA Governor | D+11.0 | D+19.6 | 8.6 | ✓ |
| 1998 CA-46 House | D+14.0 | D+17.1 | 3.1 | ✓ |
| 1998 CA Senate | D+5.0 | D+10.0 | 5.0 | ✓ |
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 | 23 | 4.75 | -0.32 | 83% |
| 3–6 wk | 10 | 4.97 | -0.77 | 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) |
|---|---|---|---|
| 1998 | 4 | 8.0 | R+8.0 |
| 2000 | 6 | 4.7 | R+2.8 |
| 2003 | 3 | 20.4 | D+20.4 |
| 2004 | 9 | 3.5 | D+1.3 |
| 2006 | 7 | 4.7 | R+4.0 |
| 2008 | 4 | 2.1 | D+1.4 |
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.