Texas Polls Tracker

Morning digest

Past digests

Texas courts and State Board of Education

Statewide judicial races have no public polling and usually land within a point or two of the generic down-ballot vote, so each Supreme Court and Court of Criminal Appeals seat is shown with the statewide-environment baseline from the Legislature model. SBOE districts list nominees only.

State Board of Education

Early vote

Texas early voting runs October 19 through October 30, 2026. The Secretary of State publishes daily turnout, but its site sits behind a bot wall that this tracker will not bypass, so the daily series is entered by hand into public/data/early-vote.json (a paste from the SOS daily report). Benchmarks for 2022 and 2024 use the same file. Charts appear once the first day of data is in.

Governors

The 36 governorships on the ballot. Ratings from Cook, Inside Elections, Sabato, RealClearPolitics, Fox, Silver Bulletin, and Decision Desk HQ via Wikipedia; polls from VoteHub. The Texas race uses the fuller poll set from the Texas statewide tab.

Battlegrounds

Scenarios

Move the assumptions and the models re-run in your browser with today's data. Nothing here is saved or published; reload to reset.

Texas Legislature

U.S. Senate

U.S. House

Trends

One point per daily snapshot. Lines fill in as the archive grows; the first snapshot was October 4, 2026.

Rating and model changes, by day

Pollsters

Every pollster seen in the tracked races. The 538 grade drives the weight each poll gets in the averages. "Lean" is the pollster's average deviation from the quality-weighted average of the race it polled, so a D+2 lean means its margins run two points more Democratic than the consensus. Partisan sponsors are noted; those polls are half-weighted.

U.S. Senate control

U.S. Senate map

Hover a state for details. Seats not up in 2026 are left blank. Ratings are the average of ten public forecasters (Cook, Decision Desk HQ, The Economist, FiftyPlusOne, Fox, Inside Elections, RealClearPolitics, Sabato, Silver Bulletin, Split Ticket).

Battlegrounds

Texas statewide

Polling average: last 30 days (minimum three polls), weighted by recency (14-day half-life), sample size, and pollster quality (FiveThirtyEight numeric grade: A+ = 1.0× down to 0.35×, unrated 0.5×), partisan/internal polls halved.

Texas House

Scroll to zoom, drag to pan, double-click to zoom in. Hover or tap a district for its numbers.

Legislative generic ballot

Closest districts

Texas Senate

Scroll to zoom, drag to pan, double-click to zoom in. Hover or tap a district for its numbers.

Seats up in 2026

U.S. House

Texas congressional districts

The 2025 map (Plan C2333) is in force for 2026. Colors show our model's Democratic win probability; ratings for all 38 districts come from the Texas race pages, and district polls are folded in where they exist.

Scroll to zoom, drag to pan, double-click to zoom in. Hover or tap a district for its numbers.

Generic congressional ballot

National battlegrounds

The forecasters' competitive list, sorted by our model. Texas districts are marked.

Sources and method

Polls model. Each race's quality-weighted polling average (D − R margin) is turned into a win probability with a Student-t error model (5 degrees of freedom, so misses of 2–3σ are not treated as freakish). σ = √(5.5² + (0.05 × days to election)² + 3²/effective polls): 5.5 points is roughly the historical root-mean-square error of final statewide polling averages, the time term adds ~1.5 points a month out, and the sparsity term widens thinly polled races. Pollster quality comes from FiveThirtyEight's pollster ratings (vendored copy, 2024 methodology); unrated pollsters get half weight. Races with no polls fall back to a forecaster-rating prior and say so. Prediction-market odds are not used anywhere on this site.

Texas Legislature model. Every seat starts from its district's 2024 presidential margin, shifted by three-quarters of the statewide swing implied by today's environment (a 60/40 blend of the legislative generic-ballot polling average and the mean of the down-ballot statewide race averages, measured against Trump's 13.7-point 2024 margin), plus 3 points for an incumbent who is running again. Uncertainty has two parts: a shared statewide error (Student-t, scale 4.5) and an independent district error (scale 6). Where State Navigate rates a district, its implied probability is averaged with the fundamentals in log-odds. Chamber control is the probability that Democratic seats reach the majority line, integrating over the statewide error. No district-level public polls exist yet; if any appear on the race pages they will be folded in.

U.S. House model. Forecasters publish ratings for about 144 competitive districts (Cook, Inside Elections, Sabato, The Economist, Split Ticket, Decision Desk HQ, FiftyPlusOne, Fox, Silver Bulletin, RealClearPolitics). Each rated seat's consensus rating becomes an implied margin (Tossup 0, Tilt 2.5, Lean 5.5, Likely 10, Safe 17), averaged half-and-half with the district's polling average where at least two district polls exist (so far only in Texas). Seats outside the list stay with the party that holds them; the three vacancies are left out. Uncertainty is a shared national error (Student-t, scale 3.5) plus independent district error (scale 5.5); control is the probability Democrats reach 218.

Seat distributions are Poisson-binomial over the races in a chamber and assume independence beyond the shared statewide or national error term, so they understate the chance of a uniform swing. Ratings tables, poll tables, and legislative district tables are read from Wikipedia (CC BY-SA 4.0), which cites each forecaster and pollster; polls also come from VoteHub's public feed. District boundaries: Texas Legislative Council plans H2316, S2168, and C2333. Not affiliated with any campaign or forecaster.