Will Republicans hold 53+ Senate seats and ≤207 House seats after the 2026 midterms?

Note: only 5/8 ensemble models completed on this run.

Analysis

This is a conjunction market: R Senate ≥53 (zero net losses from exactly 53) AND R House ≤207 (losing 11+ seats). The House leg is likely (~55-65%) given D+6 generic ballot and near-universal midterm penalty, though mid-decade redistricting (TX +5R) compresses Dem gains. The Senate leg is unlikely (~15-25%): NC is the most probable Dem flip with Cooper at Lean D, Maine is genuinely at risk, and GOP offset opportunities (GA without Kemp, MI) have weakened. The two legs are negatively correlated — the environment that delivers House ≤207 is the same one that costs NC and Maine. Joint probability ~8-13%; I land at 10%, consistent with the trimmed ensemble of 9.7% (the low-end agents' race-specific evidence outweighs the high-end agents' structural prior, but the 2018 split-verdict precedent justifies staying above 4%). The 37.9% market price is demonstrably uninformative: the event's 12 mutually exclusive siblings sum to ~400% on trivial volume, so this is placeholder pricing, not smart money. Category is elections (our 71% hit-rate segment), evidence is specific and verifiable, the kill attempt failed, and the edge is ~28 points. TRADE_NO. Self-confidence 0.6 — I verified the market's structural incoherence decisively, but my probability simply confirms the ensemble rather than replacing it.

View on Polymarket