Becker Institute
Historical accuracy is measured against certified results. Current-cycle records come from the canonical, deduplicated archive and do not by themselves change the sealed forecast.
- Historical polls
- 5
- Mean absolute error
- 7.66percentage points
- Historical bias
- D+1.54
- Represented cycles
- 21998–2002
- Current archive
- 0canonical records
- History span
- Oct 17, 1998through Oct 23, 2002
Known names and relationships
- Known aliases
- No additional reviewed aliases
- Organization type
- Single canonical organization
- Cycles represented
- 1998 · 2002
5 eligible historical polls is a limited denominator; both MAE and bias may move materially as another cycle is added.
Current-cycle records
Archive source date 2026-09-09 · bundled copy. Archive records may be newer than the sealed forecast; inclusion here does not mean the forecast uses them.
0 shown · 0 total
No record in the currently authorized canonical archive is attached to this identity.
This is not evidence that the pollster released no 2026 polling; the source-audit gaps remain documented.
Eligible historical observations
5 shown · 5 total
| Cycle | Race | Type | Field end | Sample | Poll | Result | Error |
|---|---|---|---|---|---|---|---|
| 2002 | 2002_Sen-G_NH | Sen-G | 2002-10-23 | 401 | D+6.0 | R+4.4 | D+10.4 |
| 2002 | 2002_House-G_NH-2 | House-G | 2002-10-23 | 200 | R+22.0 | R+16.0 | R+6.0 |
| 2002 | 2002_House-G_NH-1 | House-G | 2002-10-23 | 200 | R+8.0 | R+19.6 | D+11.6 |
| 2002 | 2002_Gov-G_NH | Gov-G | 2002-10-23 | 401 | R+20.0 | R+20.4 | D+0.5 |
| 1998 | 1998_Gov-G_NH | Gov-G | 1998-10-17 | 355 | D+45.0 | D+35.3 | D+9.7 |
Method · historical final-three-week D/R general-election rows plus verified historical supplement; min n=5
Definition · bias_D is poll D-R margin minus actual D-R margin; positive overstates Democrats.
A missing sample, sponsor or source is labeled unavailable; it is never reconstructed.
Poll data via VoteHub, used under CC BY 4.0. Poll data from The New York Times, used under CC BY 4.0.