Scrutinel— 2026 midterms

UpdatedOct 10, 2026, 11:27 AM EDT
To election day24 days
Simulations200,000

Texas 25th District

Currently held by a Republican · incumbent running · district redrawn for 2026

Democrat
Dione Sims
2%
Republican
Roger Williams
98%
ForecastSafe R
Projected margin
R+18.5
80% range
R+30.0 to R+7.0
Partisan lean
R+21.9
Uncertainty multiplier
1.80×

The 80% range comes from 200,000 simulations in which this race’s error is correlated with every other race through the national, state and demographic layers.

How the estimate is built
Fundamentals >99.9%
Polling average
no polls
Fundamentals
R+18.3
Blended
R+18.3
Similar-race shift
+1.4
Polling

No public polling of this matchup. The forecast rests on fundamentals plus the polling residuals of demographically and politically similar races — which is the normal case: most of the 435 House seats are never polled.

Fundamentals
Partisan lean
R+21.9
Elasticity
1.10
Incumbency
−1.9
Candidate experience
0.0
Fundraising
−0.8
National environment
D+7.1
Expected margin
R+19.7
How it voted for presidenttwo-party
2024 margin
R+24.3
2020 margin
R+16.1
Shift 2020 → 2024
R+8.3
Votes cast, 2024
309,215

Presidential results reaggregated onto the 2026 district lines. This is where the seat’s partisan lean comes from, and it carries most of the forecast in a race nobody has polled.

Who lives hereCensus ACS
Population
576,023
Median age
46.4
Median household income
$66,900
Bachelor’s degree or higher
26%
White
51%
Black
20%
Hispanic
24%
Asian
4%

Used to find demographically similar districts, so a race with no polling can borrow the polling error of races that look like it.

Campaign fundraisingFEC filings to Invalid Date
CandidatePartyRaisedCash on hand
Roger WilliamsREP$1.4M$1.1M
William James MarksDEM$79k$0

Receipts across the cycle, from FEC bulk filings. The model uses the ratio between the two sides rather than either total, and weights it lightly — money follows a competitive race at least as much as it causes one.

← Back to the full forecast