Mitch McConnell Resign Odds: The AI Model Verdict Vs. The Kalshi Market
Mitch McConnell is 84, has already announced he will not run again, and has spent much of the summer out of the Senate recovering from a fall. That combination is exactly what a prediction-market exchange turns into a contract, so Kalshi lists a yes/no market on whether he resigns his office early, before election day 2026. The crowd prices it around 12 cents. We asked three AI models to score it without letting any of them see a price, and they came back at roughly half the market, because the thing the contract actually pays on is narrower than the headline, and two of the most likely ways this ends are written out of it.
New to Kalshi event markets? Start here → how Kalshi's yes/no contracts work.
The Quick Answer
Kalshi's McConnell resignation contract trades around 12 cents, which reads as a roughly 12 percent implied chance that he resigns his Senate seat, or announces he will, before November 3, 2026. Our price-blind, three-model panel lands at about 6 percent, close to half of that. The gap is not a disagreement about whether McConnell's health is fragile. It is about what settles the contract. Kalshi's rules pay only on a voluntary early resignation or its announcement, and the secondary rule states plainly that leaving office due to death does not resolve the market to Yes. Finishing his term into the retirement he has already announced for January 2027 also pays nothing. So the models priced the one thing that does trigger it: McConnell himself choosing to step down early, from a man who keeps saying he intends to finish the job and whose own colleagues are publicly framing his absence as a delayed return.
- Market: ~12¢ (a deep, two-sided book)
- Model Blend: ~6% (about 5.7%), price-blind, after a revision round
- The Key Wrinkle Most Readers Miss: neither death nor simply serving out his term settles this market; only a voluntary early resignation does
- The Panel's Logic In One Line: most of an 84-year-old's health tail resolves as continued absence, incapacity, or death — none of which pay — so only a thin slice converts to an actual resignation.
The settlement mechanic that drives the whole thing, every seat's number, the paths that would actually settle it YES, and a plain-English guide to reading this price are all below.
Want the graded result? Every number in this piece gets scored in public once Kalshi settles this market. The settled record for this call and every other verdict lives on our Model Verdict scoreboard, and you can get the re-scored call the moment the market resolves through the email capture at the end of this article.
The Market At A Glance
| Venue | Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026) |
| The Event | "Will Mitch McConnell resign his office early?" (Kalshi market KXRETIREMM-26), a single yes/no contract, "Before election day 2026" |
| Settlement Rule, Verbatim | "If Mitch McConnell resigns their office, or announces they will resign their office, before Nov 3, 2026, then the market resolves to Yes." The secondary rule adds that "leaving office due to death will not resolve the market to Yes, and the market will remain open" |
| Settlement Sourcing | The binding language is this market's own rules above; the market can also close early if McConnell resigns |
| Deadline | November 3, 2026 (election day). The window from today is roughly three months |
| Liquidity | Deep and two-sided. About 1.68 million contracts traded over the market's life, roughly 763,000 in open interest, and about 121,000 traded in the prior 24 hours (Kalshi's public data, August 4, 2026) |
| Live YES Quote | About 12¢ last, on an 11¢ bid against a 12¢ ask, with NO at 88–89¢ (August 4, 2026). Unlike many of these contracts, this is a tight, firm market, not a soft mark |
These are model estimates, not predictions of fact and not financial advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate. Every number here gets graded in public once the market settles, in the full graded Model Verdict scoreboard.
What This Contract Actually Asks
The headline reads "will McConnell resign." The rules read more narrowly, and the difference is the entire trade. Kalshi pays YES only if McConnell "resigns their office, or announces they will resign their office, before Nov 3, 2026," and the secondary language carves out death explicitly. Three consequences follow, and none is obvious from the title.
- Serving Out His Term Resolves NO. On his 83rd birthday in February 2025, McConnell announced he would not seek reelection and would retire when his current term ends in January 2027, saying he had "some unfinished business to attend to." That already-planned, end-of-term retirement is not an early resignation, and it settles nothing. This is the same effective-now structure that governs Kalshi's athlete-retirement contracts, where a planned or future exit does not pay and only an immediate one does — our LeBron James retirement verdict and Novak Djokovic retirement verdict both turn on it.
- Death Resolves NO. The secondary rule is unusually blunt: leaving office due to death does not resolve YES, and the market stays open. For an 84-year-old with a serious 2026 health record, that clause quietly removes a large piece of what a casual reader might price toward YES.
- The Trigger Is A Voluntary Early Resignation. For the market to pay, McConnell has to decide, at some point before November 3, to step down early, or say publicly that he will. The realistic path is not a scheduled goodbye. It is his health forcing a decision to leave the seat rather than continue holding it while he recovers.
That is why the panel priced this as a thin voluntary-resignation tail rather than as a referendum on how sick McConnell is. The question is not "is his health failing." It is "does he choose to resign the seat early," and those are very different numbers.
The Board
One contract, one deadline. The market's YES price against the panel's price-blind blend:
| Outcome | Kalshi YES (last) | Implied probability | AI panel blend |
|---|---|---|---|
| McConnell Resigns His Office Early, Before Nov 3, 2026 | ~12¢ | ~12% | 5.7% |
These are model estimates, not predictions of fact and not financial advice. The blend is the simple mean of the three seats' round-two numbers; no model saw any market price. This is a liquid, tightly quoted book, so the 12¢ mark is a firm mid rather than a stale mark, which makes the model-versus-market gap a real disagreement rather than a spread artifact.
Every Seat's Number
Each model scored the market twice: once from the data card alone, then again after reading the other two anonymized rationales.
| Model | Round 1 | Round 2 |
|---|---|---|
| Claude Opus | 5% | 5% |
| Claude Sonnet | 9% | 7% |
| Claude Fable | 5% | 5% |
| Blend | 6.3% | 5.7% |
These are model estimates, not predictions of fact and not financial advice. The blend is the mean of the three round-two numbers, and the method is always disclosed. Three seats returned a verdict on this run. Three more, our GLM, Kimi and DeepSeek seats, could not run because their shared inference key returned a billing error, and the GPT, Gemini and Grok command-line seats are not installed on the machine that produced this run; we disclose that rather than paper over it. On a market where all three live seats converged this tightly, the missing seats would have had to break hard from the fetched picture to move the blend far.
Where The Panel Landed
The three seats agreed on the shape before they agreed on the number, and the shape is the settlement mechanic. Every rationale split McConnell's next three months into buckets, then noticed that the two heaviest buckets, serving out the term and death, both settle NO.
Claude Opus put it most directly:
"Repeated stated plan to finish his term to January 2027, no reporting of any early-resignation plan, and colleagues expecting his return." — Claude Opus
Claude Fable drew the line to the base rate:
"The exclusions do most of the work: the two most likely departure modes — serving out the term into his announced January 2027 retirement, and death in office — both settle NO, so I only price the narrow channel of a voluntary early resignation or its announcement within ~3 months. [...] Historical base rates agree — ailing senators overwhelmingly stay." — Claude Fable
That is why the number sits in the mid-single digits rather than near the market. Every seat assigned a real health tail: an 84-year-old with a post-polio mobility condition, recurring falls, a June hospitalization that began with a fall and brief unconsciousness and included a mild case of pneumonia, more than 40 days absent, and no medical clearance to return is truly at risk of an adverse event over three months. The seats put that severe-health tail somewhere between 20% and 35%. But most of it resolves as continued rehabilitation, ongoing incapacity while he still holds the seat, or the grim outcome the rules exclude, none of which pay YES. Only a thin slice, roughly a fifth of the tail on the panel's read, converts into McConnell voluntarily choosing to resign. Add his repeated intent to finish the term and colleagues who are actively framing this as a return, and the blend lands near 6 percent.
The Paths To YES The Panel Won't Dismiss
Six percent is not zero, and the seats were explicit about what lives inside it. These are the scenarios that would actually settle the market YES, ranked by how the panel weighted them.
- A Doctor-Driven Decision That He Cannot Continue. The single largest piece of the number. McConnell has been out more than 40 days, is not medically cleared to return, and told colleagues in early August he would not come back to vote "quite yet" on his doctors' advice. Claude Sonnet flagged exactly this path as its top mover: a rehab stall turning into a "cannot continue" moment, which is more acute here than the typical ailing-senator base rate.
- Quiet Succession Pressure That Surfaces. Every seat named the same falsifier: reporting that McConnell himself is weighing an early exit. Nothing like that is public as of this writing, and colleagues are signaling the opposite, but private family or medical pressure is the kind of thing none of the models can see, and it is the fastest way this repriced.
- A Structural Resignation To Allow A Concurrent Special Election. Under Kentucky's 2024 succession law, a Senate vacancy is filled by a special election rather than a governor's appointment, and to hold that election alongside the November 3 general the vacancy would have to occur early enough to clear a roughly 63-day proclamation window, around the start of September. That timing creates a narrow, mostly theoretical incentive to resign sooner rather than let the seat sit; the panel treated it as neutral-to-slightly-real, not a base case, because McConnell has shown no appetite to accommodate it and the term ends in January anyway.
None of these is a base case. Together they are why the panel prices this near 6 percent rather than near zero, and why it will not go higher: every one of them requires McConnell to choose to leave the seat, not merely to be unable to work it.
Why The Market May Sit Higher
We do not get to see the market's reasoning, only its price, so this is inference rather than fact. But the most likely explanation for a 12-cent contract is the exact trap the settlement rule sets. A trader who reads "will McConnell resign before the midterms" and thinks about the fall, the pneumonia, the 40-day absence, and the missed votes would price something visibly above zero, because it feels like a person whose time in office may be ending. That trader is answering "is McConnell's tenure in jeopardy," which is a fair bit higher than 6 percent. The contract asks a narrower question, "does he choose to resign early, without dying and without simply finishing his term," and the honest answer to that one is lower.
There is a real counterargument, and the panel held room for it: McConnell's current situation is more acute than a typical ailing-senator base rate. He is not merely old; he is out, uncleared, and declining to give a return date, and a prolonged rehab really could end in a decision to step down. That is priced in, and it is why the blend is 6 percent rather than 2. It is also not worth 12 cents, because the base rate of a sitting senator voluntarily resigning inside a specific three-month window, while his own party frames his absence as temporary, is low even for an 84-year-old.
How To Read This Market
- Eligibility. Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook. You must be 18 or older and in an eligible state, and availability varies by state and can change.
- A Firm Price, Not A Wide Spread. Unlike many of these contracts, this one is deeply traded, with an 11-cent bid against a 12-cent ask. The gap here is a real model-versus-market disagreement rather than an artifact of a thin book. Our guide to turning Kalshi's cents into real odds covers the arithmetic.
- Time And Capital. Absent a qualifying resignation, the contract resolves NO on November 3, 2026, so a position can sit for up to three months with no interim payout, and the market can also close early if McConnell resigns.
- What Would Move This. Every seat named the same falsifier: any reporting that McConnell himself is weighing an early resignation, or a doctor-driven statement that he cannot continue. Until then, there is no return date to read, which is a real gap in what anyone can know right now.
- Not Advice. Everything here is analysis for information, not a recommendation to trade, and nothing here is a comment on what McConnell should do. These are model estimates, not predictions of fact and not financial advice.
Every number in this piece gets graded in public once this market settles, on the full Model Verdict scoreboard, which is also where you can judge how this panel has done on every prior call. If you want the method applied to the broader 2026 map, our Senate balance-of-power verdict and House control verdict score the chambers the same way. Kalshi onboards United States traders only; if you came for the politics rather than the contract, our free expert picks run daily across the sports our own tools price.
Prices on this page refresh during any McConnell health or resignation-news wave, with the dated "as of" stamp in the market table above carrying the evidence. If McConnell makes news as the Senate returns from recess or beyond, check the timestamp before trusting a number.
FAQ
Is Mitch McConnell resigning? There is no announcement or reporting that he is. As of August 4, 2026, McConnell has been absent from the Senate for more than 40 days recovering from a June fall and is not yet medically cleared to return, but he has said he intends to finish his term, and Republican colleagues have framed his absence as a delayed return rather than an exit. Kalshi's contract on an early resignation before election day trades around 12 cents, and our price-blind model panel puts it near 6 percent.
When does Mitch McConnell's term end? McConnell's current Senate term ends in January 2027. On his 83rd birthday in February 2025 he announced he would not seek reelection in 2026 and would retire when the term ends. That planned, end-of-term retirement is different from an early resignation, and it does not settle Kalshi's contract YES.
What happens to McConnell's Senate seat if he resigns early? Under a 2024 Kentucky law, a U.S. Senate vacancy is filled by a special election, not by a governor's appointment. To hold that special election alongside the November 3 general, the vacancy would generally have to occur early enough to clear a roughly 63-day proclamation window, around the start of September 2026; after that, a concurrent election is no longer possible.
Why is the model estimate lower than the Kalshi price? Because the contract only pays on a voluntary early resignation, and it explicitly does not pay if McConnell leaves office due to death or simply finishes his term. A casual read of the price looks like it is answering "is McConnell's time in office in jeopardy," which feels higher given his health. The models answered the narrower question the rules actually ask, "does he choose to resign early," which is a lower, mid-single-digit probability.
The Bottom Line
This is not a close market, and the panel does not pretend it is. The 12-cent price is best read as a misframing: it answers how fragile McConnell's hold on the seat looks, and the answer to that is deeply uncertain. The contract asks something narrower, whether he chooses to resign early before November 3, and the fetched record, an already-announced plan to finish his term, colleagues framing his absence as a return, and a settlement rule that pays on neither death nor a term-end retirement, makes that a mid-single-digit probability. The content worth your time is not the 6. It is the settlement rule that quietly excludes both the ending everyone fears and the retirement he has already scheduled. Read the same rule the other way, and it explains our LeBron James and Novak Djokovic retirement numbers too.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $166M traded
- Rocket Classic Winner · $124M traded
- 2027 Pro Football Champion · $51M traded
- Pro Baseball Champion · $40M traded
- Bitcoin price at the end of 2026 · $30M traded
- 2027 Pro Basketball Champion · $19M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.
Verdicts generated August 4, 2026, roughly 9:00 p.m. UTC. Prices as of August 4, 2026, roughly 9:00 p.m. UTC. Panel: Claude Fable, Claude Opus and Claude Sonnet, all price-blind, followed by an anonymized revision round. Three additional seats (GLM, Kimi, DeepSeek) were unavailable on a billing error, and the GPT, Gemini and Grok command-line seats are not installed on this machine. These are model estimates, not predictions of fact and not financial advice.
