Calibrated ~0.76 probability Democrats win ≥218 House seats in the Nov 3 2026 midterms. Markets range ~73–82%; we anchor to the model floor at ~0.76. Cook Political Report: 14 of 18 toss-ups are GOP-held; generic ballot Dem +5–6pts. Non-partisan, purely probabilistic.
Prediction market range: Polymarket ~82%; Kalshi / RaceToTheWH ~73%. Generic ballot: Dem +5–6pts (RCP aggregate, Jun 2026). Cook Political: 14 of 18 toss-up districts are currently GOP-held seats. We anchor to ~0.76 — the model floor, not the hottest market — for conservative calibration. Source: AP / NBC certified race calls.
We lock a calibrated binary claim: P(Democrats win ≥218 of 435 US House seats in the November 3 2026 midterm elections, certified by AP / NBC race calls) ≈ 0.76. This is a calibrated probability, not a partisan prediction. The complementary ~0.24 covers Republicans retaining House control.
Cook Political Report identifies 18 toss-up House districts; 14 are currently GOP-held — a structural map advantage for Democrats. The RealClearPolitics generic congressional ballot shows Democrats +5–6 points (Jun 2026). Prediction markets: Polymarket ~82% (higher), Kalshi / RaceToTheWH ~73% (lower). We calibrate at ~0.76 — anchored to the model floor (~73%), not the highest market reading — consistent with RAOSCAFF's doctrine of conservative calibration when market spread is wide. Non-partisan: we derive probability from public market synthesis, not political preference. Resolves against AP / NBC certified race calls determining House majority control.
RAOSCAFF locks P-44 on 2026-06-21. Non-partisan throughout: calibrated probability from market synthesis, not political preference. Resolves against AP / NBC certified race calls.
Non-partisan framing throughout — we derive probability from market synthesis, not political preference.