Texas Senate Odds: Talarico Vs. Paxton At A Dead-Even 50¢

The Quick Answer

The Texas Senate odds on Kalshi are the most balanced high-volume race on the exchange: the Democratic contract (James Talarico) and the Republican contract (Ken Paxton) both trade at 50.5¢, on 7.6 million contracts. Our seven-model AI panel read the complete post-nomination polling file and the structural history price-blind and came out 54.8% Republican, 43.7% Democratic, a modest but unanimous lean the market is not pricing. The polls, the streak, the money, and the exact tension between them are all below.

Free: The Weekly PM Market Brief — the panel's graded record, the week's biggest market-vs-model gaps, and what is spiking next. One email, Sundays, from the signup box on this page.

The Market

This market settles on party, not candidate. The rule, verbatim: "If a representative of the Republican party is sworn in as a Senator of Texas for the term beginning in 2027, then the market resolves to Yes." The Democratic contract mirrors it. Note the wrinkle that creates: a third-party winner, or a party switch between the election and swearing-in, would resolve both contracts No. The panel prices that combined tail at 1.5%.

The path here matters for reading the race. Ken Paxton reached the general by defeating four-term incumbent John Cornyn in the Republican runoff, 63.8 to 36.2, after a 42.0-40.5 first round. James Talarico won his primary outright at 52.4%. Prices are bid-ask midpoints as of August 9, 2026, and both contracts held a one-cent spread; panel verdicts were generated the same evening, price-blind.

The Board

More live boards from the same panel: 2028 Republican nominee odds (J.D. Vance 41.5¢) · 2028 Democratic nominee odds (Gavin Newsom 16.5¢) · The full markets hub

What The Polls Say: Talarico By A Nose

The card carried every head-to-head poll in the feed since the nominees were set on May 26, thirteen in all, not a curated subset. The average: Talarico +1.8, and +1.5 after excluding partisan or aligned-sponsor surveys, which means the lead is not a house-effect artifact. The three most recent, all fielded in the last two weeks of July: Wedgewood 48-46 Talarico, Texas Southern University's Barbara Jordan Center 47-45, and a Fox News-sponsored Beacon/Shaw survey at 51-48. The New York Times/Siena poll in late June had it dead even at 47-47.

The national backdrop leans the same direction: Democrats lead the generic ballot by 6.7 points across six public aggregators, and presidential net approval sits at −21.5, a historically hostile environment for the president's party. And the panel repeatedly cited one more fact from the card's archive: Trump-era midterm polling was nearly unbiased (2018 missed by just 0.3 points, 2022 by 0.9), which forecloses the reflexive "shave the Democrat's polls" adjustment.

What The Structure Says: 1994

The other column of the card is just as one-sided. No Democrat has won any statewide office in Texas since 1994, a 32-year streak. Trump carried the state by 13.68 points in 2024, and Colin Allred lost the Senate race that same day by 8.49. The modern high-water mark for Texas Democrats is Beto O'Rourke's 2.56-point loss to Ted Cruz in 2018, in a national blue wave. The professional raters split on exactly this tension: Cook, Sabato, and Inside Elections all keep the race on the Republican side of the line (Lean or Likely R), while The Economist, RealClearPolitics, and Race to the WH call it a tossup.

"Texas's fundamentals are stark: no Democrat has won statewide since 1994, Cruz beat O'Rourke by 2.6 even in 2018's wave, and Allred lost by 8.5 just two years ago on a similar map. That floor has survived worse." (Claude Sonnet)

The Money

One asymmetry the polls do not capture: Talarico announced $30 million raised in the second quarter, described as a record for a Senate candidate in the second quarter of an election year, and over $70 million for the cycle from more than 780,000 donors, 97% of contributions at $100 or less. Paxton announced $9 million for the same quarter. Neither campaign's cash-on-hand was disclosed in the filings the card cites, so treat this as a fundraising-pace fact, not a bank-balance one.

The Panel's Verdict: A Lean The Market Won't Take

Every seat framed the race the same way: the polls say one thing, the structure says another, and the entire forecast is how you weight them. What is striking is that all seven seats, after deliberation, landed within two points of each other, and all seven on the Republican side of the market's coin flip.

"The post-nomination polling average is meaningful: Talarico leads by 1.8 points overall and 1.5 excluding partisan or aligned polls, so the Democratic edge is not just a sponsor artifact." (ChatGPT, which still finished at 56% Republican)

The panel's summary logic: a two-point polling lead in August is smaller than the structural cushion Texas Republicans have repeatedly shown in November, but a 32-year streak is worth much less than usual in a cycle with a D+6.7 generic ballot and an incumbent-party approval deficit this deep. Put together, that is a modest Republican edge, not the market's pure toss-up and not a safe hold either.

Where The Panel Changed Its Mind

Round one spread from 36% to 52% on the Democratic side; the revision round collapsed it to 42-45. The two largest moves came from opposite directions, which is exactly what the round is for.

"The polls deserve serious credit: 13 post-nomination surveys, the nonpartisan cut (+1.5) essentially matching the full average (+1.8), so the lead isn't a partisan artifact, and Trump-era midterm poll bias was near zero (2018 −0.3, 2022 +0.9), so no reflexive shave applies." (Claude Opus, which moved from 36% to 44% Democratic after starting as the panel's most structure-weighted seat)

"Texas has not elected a Democrat statewide since 1994, and recent Senate races saw polls underestimate Republican margins." (DeepSeek V4, which moved the other way, from 52% Democratic to 45%, after starting as the panel's most polls-weighted seat)

What Would Change The Panel's Mind

• A nonpartisan poll putting Talarico up 4 or more. The panel discounted a two-point lead; a lead outside the structural cushion is a different race.
• Movement in the national environment. The generic ballot and approval numbers carried real weight in every rationale; a meaningful recovery compresses the Democratic case.
• September ad translation. The fundraising gap only matters if it shows up as a sustained paid-media edge; several seats flagged this as the mechanism to watch.
• A late Republican consolidation. The 2018 pattern, a close summer race widening by November, is the structural argument's favorite precedent; early evidence of it would push the blend past 60.

Every Seat's Round-1 Number

The per-model split before the revision round. The board above carries the headline numbers, which average all seven seats' revised round-two estimates. Seven of the panel's eight seats ran this board.

Every seat on this panel is graded against real market settlements — records to date: GPT 83% on 3,225 graded calls · Claude Fable 82% on 450 graded calls · Claude Opus 84% on 514 graded calls · Claude Sonnet 82% on 498 graded calls · GLM 81% on 2,307 graded calls · Kimi 83% on 2,285 graded calls · DeepSeek 80% on 2,332 graded calls. Recomputed daily; the full scoreboard is public.

These are model estimates, not predictions of fact and not financial or trading advice. Models are frequently wrong; the market price reflects real traders' money. Kalshi is a CFTC-regulated exchange; 18+, availability varies by state.

Every number in this piece gets graded in public when the race settles, on the same scoreboard as every other call the panel has made.

Tracked Like A Projection

This page treats the race the way a projections product treats a player: the polling average is the observed line, the structural baseline is the prior, and the forecast is the disciplined blend, error bars and all. The market's 50.5/50.5 says traders weight the polls at full face value; the panel's 55/44 says the prior still deserves a thumb on the scale.

A Worked Example: Reading Two Prices Against One Forecast

The arithmetic is worth doing once, slowly. At fetch time both contracts showed a 50¢ bid against a 51¢ ask, so the market's implied probability for each side is 50.5%. The panel's blend converts to a fair value near 55¢ on the Republican contract and about 44¢ on the Democratic one. That is a 4-point disagreement on each side of the same race, before fees, which is meaningful on a two-outcome market this liquid but far from a table-pounding gap. The honest read: the panel sees the Republican side as slightly cheap at 50.5¢, and the size of "slightly" is the entire forecast. Both positions will be graded by the same November result, which is what separates this from punditry. It is the same discipline the Stokastic Sims apply to a DFS slate, pointed at the biggest unclaimed race on the 2026 map, and this page re-scores into every polling wave between now and November.

Settlement Timeline

• November 3, 2026: election day in Texas.
• January 2027: the winner is sworn in; the party contracts settle on who takes the seat.
• January 4, 2027: listed market expiration on Kalshi.
• This page re-scores when the story moves; prices above are as of August 9, 2026.

The Bottom Line

Two prices at 50.5¢ say the market sees a pure coin flip. Seven graded models, reading the identical file, all landed on the same side of that flip and none further than 64% from the middle, which is its own message: this is a real race, the first one Texas has hosted in a generation, and the disagreement between a two-point polling lead and a 32-year streak is the entire question. The panel wrote 55-44 next to its name. November grades it, and the full markets hub carries the live price until then.