The Senate’s map changes every cycle. Comparing years means comparing different states, candidates, pollsters, and turnout conditions—not repeatedly measuring one fixed set of races.
Select a cycle to inspect its error and the number of eligible estimates. The changing sample is part of the story.
U.S. Senate
Original archive calculation: final 1–14 days by median field date; latest estimate per firm/race, same-date ties averaged. Includes retained general, special and runoff contests within source cycles; not AAPOR’s methodology. Outcomes and exclusions are documented below.
| Cycle | Mean absolute error | Mean signed error | Estimates | Contests |
|---|---|---|---|---|
| 1998 | 6.01 | −2.7 | 62 | 24 |
| 2000 | 5.12 | −2.8 | 101 | 28 |
| 2002 | 4.99 | +1.6 | 75 | 25 |
| 2004 | 4.80 | +0.3 | 83 | 25 |
| 2006 | 5.03 | −2.3 | 130 | 23 |
| 2008 | 4.60 | −0.6 | 122 | 33 |
| 2010 | 4.30 | −1.6 | 129 | 30 |
| 2012 | 4.70 | −3.5 | 139 | 28 |
| 2013 | Unavailable | — | 0 | 0 |
| 2014 | 4.89 | +3.3 | 98 | 22 |
| 2016 | 5.18 | +3.5 | 145 | 32 |
| 2017 | Unavailable | — | 0 | 0 |
| 2018 | 3.96 | +0.8 | 116 | 24 |
| 2020 | 5.59 | +4.9 | 138 | 26 |
| 2022 | 4.67 | −1.1 | 154 | 23 |
Mean absolute margin error was 6.01 points in 1998 and 4.67 in 2022, compared with 5.59 in 2020. Across the covered cycles, the smallest observed miss was 3.96 in 2018; the largest was 6.01 in 1998.
The latest cycle contains 154 selected firm/race estimates across 23 contests. Its signed error, −1.1 points, overstated Republican margins. Changing races and pollsters limit comparisons.
AAPOR’s weighted analysis · October 23–November 5 · raw D–R margin error
House district polls cover only 20 contests; this is not an estimate for all 435 districts. Presidential congressional-district polls concern electoral votes and are distinct from House races.
The observed error changes across cycles. The archive is not a balanced panel of the same races or firms, so its trend is descriptive. AAPOR’s separate evaluation reports Senate mean absolute margin error of 6.0 points in 2020, 4.8 in 2022, and 3.3 in 2024.
The historical archive groups some general, special, and runoff contests under source cycles. January 2021 Georgia runoffs belong to its 2020 cycle. The dataset exposes these race types and calendar years rather than silently relabeling every observation a November general election.
Different windows can contain different pollsters and races. These are descriptive comparisons, not a causal experiment.
The interactive timing chart loads with JavaScript. Every value is also in the downloadable dataset.
Latest per firm/race within each non-overlapping midpoint window.
