Market Shares Corp.
Graded against the actual result across 33 races (from 52 polls, through 2012).
Head-to-head vs the VotePredictor model
The fair, apples-to-apples test: on the 15 races Market Shares Corp. 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 |
|---|---|---|
| Market Shares Corp. | 3.98 | 87% |
| VotePredictor | 2.28 | 80% |
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 (33)
Each race Market Shares Corp. 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 |
|---|---|---|---|---|
| 2012 IL-8 House | D+10.0 | D+9.5 | 0.5 | ✓ |
| 2012 IL President | D+19.0 | D+16.9 | 2.1 | ✓ |
| 2010 IL Governor | R+4.0 | D+0.9 | 4.9 | ✗ |
| 2010 IL Senate | R+3.0 | R+1.6 | 1.4 | ✓ |
| 2008 IL President | D+24.0 | D+25.1 | 1.1 | ✓ |
| 2006 IL Governor | D+15.0 | D+10.5 | 4.5 | ✓ |
| 2006 IL-6 House | R+4.0 | R+2.7 | 1.3 | ✓ |
| 2006 IL-8 House | D+19.0 | D+6.9 | 12.1 | ✓ |
| 2004 IA President | R+2.0 | R+0.7 | 1.3 | ✓ |
| 2004 IL President | D+8.0 | D+10.3 | 2.3 | ✓ |
| 2004 MN President | D+2.0 | D+3.5 | 1.5 | ✓ |
| 2004 OH President | D+4.0 | R+2.1 | 6.1 | ✗ |
| 2004 WI President | D+4.0 | D+0.4 | 3.6 | ✓ |
| 2004 IL Senate | D+47.0 | D+42.9 | 4.1 | ✓ |
| 2004 WI Senate | D+24.0 | D+11.2 | 12.8 | ✓ |
| 2002 IL Governor | D+10.0 | D+7.1 | 2.9 | ✓ |
| 2002 WI Governor | D+2.0 | D+3.7 | 1.7 | ✓ |
| 2002 IN-7 House | D+1.0 | D+9.0 | 8.0 | ✓ |
| 2002 IL Senate | D+20.0 | D+22.3 | 2.3 | ✓ |
| 2000 IN Governor | D+16.0 | D+14.9 | 1.1 | ✓ |
| 2000 IL President | D+10.0 | D+12.0 | 2.0 | ✓ |
| 2000 IN President | R+23.0 | R+15.6 | 7.4 | ✓ |
| 2000 MI President | D+2.0 | D+5.1 | 3.1 | ✓ |
| 2000 NJ President | D+18.0 | D+15.8 | 2.2 | ✓ |
| 2000 OH President | D+2.0 | R+3.5 | 5.5 | ✗ |
| 2000 PA President | D+5.0 | D+4.2 | 0.8 | ✓ |
| 2000 WI President | R+5.0 | D+0.2 | 5.2 | ✗ |
| 2000 IN Senate | R+59.0 | R+34.7 | 24.3 | ✓ |
| 2000 WI Senate | D+33.0 | D+24.5 | 8.5 | ✓ |
| 1998 IL Governor | R+15.0 | R+3.6 | 11.4 | ✓ |
| 1998 WI Governor | R+41.0 | R+21.0 | 20.0 | ✓ |
| 1998 IL Senate | R+10.0 | R+2.9 | 7.1 | ✓ |
| 1998 WI Senate | R+3.0 | D+2.1 | 5.1 | ✗ |
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 |
|---|---|---|---|---|
| 1–3 wk | 21 | 6.09 | +1.02 | 86% |
| 3–6 wk | 22 | 6.03 | +0.29 | 91% |
| 6–9 wk | 8 | 4.30 | -1.85 | 88% |
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 | 8 | 12.1 | R+12.1 |
| 2000 | 15 | 6.2 | R+0.9 |
| 2002 | 7 | 4.5 | R+0.4 |
| 2004 | 10 | 4.3 | D+2.7 |
| 2006 | 5 | 4.6 | D+4.0 |
| 2010 | 4 | 2.5 | R+0.6 |
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.