Where they agree
13 / 14
Races within 2.5 pts. Mean divergence: 1.9 pts. Almost the same map.
Season 1 · 2026 Midterms
Awaiting resultsPolymarket vs Kalshi — Prediction Showdown
The 2026 US Senate, through two prediction markets. The prices are frozen. The verdict comes from the results.
Venue 01
Polymarket
- Brier / log loss
- Awaiting results
- Snapshot · Nov 3
- 18:53 ET
Venue 02
Kalshi
- Brier / log loss
- Awaiting results
- Snapshot · Nov 3
- 18:54–19:04 ET
01 / At a glance
The one big split
9.6 pts
Florida: Polymarket 15.5% vs Kalshi 5.9% for the Democrat. Tonight’s sharpness test.
The trust test
86–87%
Both priced Democratic Senate control here. PM 86.5% · K 87.0%. A shared conviction to grade.
02 / The frozen board
14 Senate races · 2 control markets
Democrat win probability (%), except Nebraska*. Gap = absolute difference in percentage points. Scores will be shown per venue after verified results.
Scroll the table horizontally to see all columns →
| Race / priced subject | Polymarket | Kalshi | Divergence | Close ET | Result | Brier · PM / K | Log loss · PM / K |
|---|---|---|---|---|---|---|---|
| Senate (D)Democratic Senate control | 86.5% | 87.0% | 0.5 pts | — | Awaiting results | Awaiting results | Awaiting results |
| House (D)Democratic House control | 98.1% | n/a | n/a | — | Awaiting results | Awaiting results | Awaiting results |
| SENATE / STATE RACES | |||||||
| TexasTalarico (D) vs Paxton (R) | 83.5% | 81.0% | 2.5 pts | 20:00 | Awaiting results | Awaiting results | Awaiting results |
| MichiganEl-Sayed (D) vs Rogers (R) | 85.5% | 84.0% | 1.5 pts | 20:00 | Awaiting results | Awaiting results | Awaiting results |
| Ohio specialBrown (D) vs Husted (R) | 82.5% | 81.0% | 1.5 pts | 19:30 | Awaiting results | Awaiting results | Awaiting results |
| MaineJackson (D) vs Collins (R) | 74.5% | 73.0% | 1.5 pts | 20:00 | Awaiting results | Awaiting results | Awaiting results |
| AlaskaPeltola (D) vs Sullivan (R) | 73.5% | 74.0% | 0.5 pts | 24:00 | Awaiting results | Awaiting results | Awaiting results |
| KansasHamilton (D) vs Marshall (R) | 72.5% | 74.0% | 1.5 pts | 20:00 | Awaiting results | Awaiting results | Awaiting results |
| IowaTurek (D) vs Hinson (R) | 64.0% | 62.0% | 2.0 pts | 21:00 | Awaiting results | Awaiting results | Awaiting results |
| North CarolinaCooper (D) vs Whatley (R) | 98.4% | 98.8% | 0.4 pts | 19:30 | Awaiting results | Awaiting results | Awaiting results |
| GeorgiaOssoff (D) vs Collins (R) | 98.8% | 98.4% | 0.4 pts | 19:00 | Awaiting results | Awaiting results | Awaiting results |
| South CarolinaAndrews (D) vs Graham (R) | 36.5% | 39.0% | 2.5 pts | 19:00 | Awaiting results | Awaiting results | Awaiting results |
| New HampshirePappas (D) vs Sununu (R) | 96.0% | 94.6% | 1.4 pts | 19:00 | Awaiting results | Awaiting results | Awaiting results |
| MinnesotaFlanagan (D) vs Tafoya (R) | 94.5% | 93.9% | 0.6 pts | 21:00 | Awaiting results | Awaiting results | Awaiting results |
| Florida specialNixon (D) vs Moody (R) | 15.5% | 5.9% | 9.6 pts | 20:00 | Awaiting results | Awaiting results | Awaiting results |
| Nebraska *Ricketts (R incumbent) vs independent challenger | 63.0% | 62.0% | 1.0 pts | 21:00 | Awaiting results | Awaiting results | Awaiting results |
* Nebraska prices Ricketts, the Republican incumbent, to win against an independent challenger. It is not a Democratic win probability. Alaska’s 24:00 ET close means midnight on November 4. House Kalshi data is n/a.
03 / Methodology
Prices are implied probabilities (%) from both venues’ public APIs, snapshotted on Election Day, November 3, 2026: Polymarket at 18:53 ET, Kalshi at 18:54–19:04 ET, ahead of later battleground poll closes.
The Kalshi window overlaps the listed 19:00 ET closes. These are fixed Election Day snapshots, not verified final quotes at every individual poll close. The supplied dataset does not include per-market Kalshi timestamps or the exact bid/ask/last-price basis.
Brier score = (p − y)². Log loss = −ln(probability assigned to the actual outcome), with p on a 0–1 scale. Lower is better. Venue averages will use the same resolved, paired state races; control markets stay separate. No outcomes or scores are inferred from prices.
After results, calibration bins will compare stated probabilities with observed hit rates, alongside sample counts. Until then: Awaiting results.
Read the scoring guide →