Discover and track politics prediction markets.

Kalshi's CONTROLS and CONTROLH contracts auto-settle once the AP Decision Desk has called 51 Senate seats (or 218 House seats) for one party. The contract closes Feb 1 of the following year to allow runoffs and recounts to complete. If a runoff in Georgia or Louisiana tips control between Election Night and certification, the contract holds until AP confirms the final count. Settlement is final on the AP-confirmed party count.

Georgia and Louisiana use runoff rules that can push race calls weeks past Election Night. Kalshi control contracts close Feb 1 specifically to accommodate this. Individual race contracts may close earlier and resolve on Election Night calls; runoff contracts list separately. Read each contract's close date and resolution language. The cohort phrase: all of this is public and it is on you to find it.

Control contracts resolve on the aggregate party count (51 Senate seats, 218 House seats). They're a single binary on the cycle outcome. Individual-race contracts resolve on a single seat (Texas Senate, PA-7, etc.). Control contracts have deeper liquidity but slower resolution because the math depends on every race; individual races resolve faster but have thinner books. Most traders take control for size, individual races for edge.

Standard practice: control contracts hold open until certified counts are final. Individual-race contracts may resolve on Election Night AP call and not unwind even if a recount later flips the outcome, similar to the post-storm reanalysis rule on hurricane contracts. Read the resolution language. Polymarket race contracts typically hold until AP calls; Kalshi varies by series. The 2024 cycle had several individual-race contracts that resolved on the initial call and held through later recount-driven changes.

Because runoffs in Georgia and Louisiana, and recount windows in close states, can push a 51-seat call past Election Night. Kalshi's resolution language requires AP-confirmed party count, not network projections. The Feb 1 close date is the conservative window that accommodates every plausible certification timeline.

Cook Political Report's Partisan Voting Index measures how a district has voted relative to the country in the last two presidential cycles. R+5 means the district was 5 points more Republican than the national average. PVI is used to translate generic-ballot signals into district-level forecasts. A 3-point Democratic generic ballot lead translates to wins in R+0 and R+1 districts but losses in R+5 districts. PVI is the bridge between national polling and district-level contract price.

Special-election margins historically over-perform the next-cycle generic ballot by 3-5 points. Large out-party over-performance is read as a leading indicator of midterm outcomes. Special elections themselves don't directly resolve control contracts, but they move price the same way Cook re-ratings do. Watch for clusters of special-election results in May, June, and September.

Primary contracts resolve on the actual primary election (a specific date, e.g. Texas Republican Senate Primary on March 3). Nominee contracts resolve on whoever the party formally nominates at convention, which is usually the primary winner but not always. Primary contracts close on primary day; nominee contracts close at convention. The resolution language differs.

Most presidential contracts resolve on the popular-vote-and-electoral-college outcome that AP calls on Election Night. Faithless electors (electors who vote against their state's result) have not historically affected resolution because no faithless-elector cluster has swung an election. If it ever does, contracts will likely resolve on the ratified Congressional count in early January, not Election Night calls. Read the resolution language carefully on contracts that span Election Night to inauguration.

An open seat is a district where the incumbent is not running for re-election (retirement, running for higher office, primary loss). Open seats flip parties more often than incumbent-held seats because the incumbency advantage is gone. Race-raters frequently move open seats from Solid to Lean immediately on retirement announcement. Traders watch retirement news closely because the price impact is immediate.

Polymarket settlements run via UMA proposer. After a contract's deadline, a UMA tokenholder files a proposed outcome and the network confirms within 24 hours unless disputed. A dispute requires a $750 bond and triggers a tokenholder vote. Disputes on political contracts are more frequent than on weather contracts because resolution language is often ambiguous (what counts as "Trump acquires Greenland"). The March 2025 Ukraine mineral deal incident exposed governance risk in this layer; UMA later restricted proposer privileges to a whitelist (MOOV2). Read the proposer history before trading edge-case political contracts.

The Associated Press's election-call team. AP calls a race when, in its judgement, the trailing candidate cannot mathematically catch up given the remaining unreported vote. AP's call is faster than certified counts but slower than network calls. Most prediction markets use AP because it's institutionally credible, methodologically conservative, and a single source across every state. Kalshi names AP specifically; Polymarket usually defers to AP via UMA proposer.