Michigan Senate Odds 2026: The El-Sayed Vs. Rogers AI Model Verdict Vs. The Kalshi Market

Michigan picks a Democratic Senate nominee on Tuesday, August 4, 2026, and then sends that winner against Republican Mike Rogers in a November 3 general election that Cook Political Report and Sabato's Crystal Ball both rate a toss-up. Kalshi has turned the whole thing into markets, and we asked our AI panel to score the general-election race without letting any model see a single price. What came back is one of the sharpest market-versus-model splits we have run: the market makes Democrat Abdul El-Sayed a clear favorite, and the panel flips the race, making Rogers a slight favorite instead.

New to Kalshi event markets? Start here → how Kalshi's yes/no contracts work.

The Quick Answer

Kalshi's general-election market makes Abdul El-Sayed (Democrat) the favorite near 64 cents, with Mike Rogers (Republican) at about 36 cents and the two other Democrats in low single digits. Our price-blind AI panel reaches the opposite verdict: it blends Rogers to about 57% and El-Sayed to about 34%, with Haley Stevens at about 8% and Mallory McMorrow near 1%. That is a roughly 30-point disagreement on the favorite, and it comes down to one fetched fact the models leaned on hard: El-Sayed trails Rogers by about 3.9 points on the non-partisan polling average, while Stevens runs about even with him.

• Market Favorite: El-Sayed (Democrat), ~64¢
• Model Favorite: Rogers (Republican), ~57% blend (about 21 points above his ~36¢ price)
• The Gap That Is The Story: the models put El-Sayed at ~34%, roughly 30 points below his market price
• The Tension In One Line: the market is pricing a Democratic-leaning midterm environment, and the panel is pricing the head-to-head polls, in which the likely Democratic nominee is the one running behind.
• The Honest Caveat Up Top: a provider outage knocked five of our eight standard seats offline for this run, so this board is scored by three Claude models only. Three correlated models agreeing is weaker evidence than eight independent ones, and we say so plainly below.

The full board, the Democratic primary that decides the nominee today, the head-to-head polling in detail, and a plain-English guide to reading these prices are all below.

Get the panel's verdict when this settles. We grade every model call in public once a market resolves. Michigan's primary settles tonight and the general in November — get the re-scored board and the settlement grade by grabbing our free updates at the end of this page. No pick, no code, just the receipts.

The Market At A Glance

Venue	Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026)
The General-Election Market	"Michigan Senate winner? (Person)" (Kalshi event KXMISENATE-26), one yes/no contract per person; the contract for whoever wins the November 3, 2026 election resolves YES, the rest No
The Primary Market	"Michigan Democratic Senate nominee?" (Kalshi event KXSENATEMID-26), one contract per candidate, settling on the August 4, 2026 Democratic primary
Liquidity (As Of August 4, 2026, 8:30 A.m. ET)	The two markets differ sharply. The primary market is deep and heavily traded on primary day, per Kalshi's live data: El-Sayed's nominee contract carries about 2.0 million contracts of open interest and Stevens's about 3.4 million, with hundreds of thousands changing hands in 24 hours. The general-election contracts are much lighter, roughly 180,000 and 135,000 contracts of open interest on the El-Sayed and Rogers lines with thin 24-hour volume, so their prices carry a wider spread and should be read as a lean, not a depth-tested consensus
Snapshot	Prices as of August 4, 2026, 8:30 a.m. ET (latest snapshot; prices on this page refresh during the news wave through the primary count tonight and on to November). Verdicts generated August 4, 2026

These are model estimates, not predictions of fact and not financial or voting advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate, and this article does not advocate for or against any candidate or party. Every number in this piece gets graded in public once the market settles, in the full graded Model Verdict scoreboard.

What Actually Decides This Race

Before any model number, six fetched facts define this board, and they truly pull against each other:

• It Is An Open Seat In A True Swing State. Democratic Sen. Gary Peters is retiring, so no incumbent runs. Michigan is about as evenly divided as states come: Trump carried it by roughly 1.5 points in 2024 while voters elected a Democratic senator on the same ballot.
• The Republican Already Came Within A Hair Of Winning This Exact Seat. Mike Rogers, a former U.S. representative and House Intelligence Committee chairman, is the GOP nominee, running unopposed. In 2024 he lost Michigan's other open Senate seat to Democrat Elissa Slotkin by 0.3 points (48.6% to 48.3%), the closest Senate race in the country that cycle. He is not a fringe candidate; he is a proven near-miss.
• The Democratic Nominee Is Decided Today, And The Two Finalists Are Not Interchangeable. Today's August 4 primary is effectively Dr. Abdul El-Sayed, a former Detroit and Wayne County health official and 2018 gubernatorial candidate running in the party's progressive lane, against Rep. Haley Stevens, from its moderate lane. State Sen. Mallory McMorrow suspended her campaign in early July but remains on the ballot.
• The Head-To-Head Polls Show The Two Democrats Performing Differently Against Rogers. In the live polling feed, the same pollsters that show Stevens tied or narrowly ahead of Rogers show El-Sayed trailing him. The late-July EPIC-MRA poll had Rogers 46, El-Sayed 43, and Stevens 44, Rogers 42. Across the non-partisan polls in the feed, El-Sayed averages about 3.9 points behind Rogers while Stevens runs essentially even, a real, measured gap between the two possible nominees.
• The Rater Consensus Is A Toss-Up. Both Cook Political Report and Sabato's Crystal Ball rate the open Michigan seat a toss-up, a live Republican flip chance in a Democratic-held seat.
• The National Environment Leans The Other Way. This is a midterm with a Republican in the White House, and historically the president's party is on defense in midterms; the live generic ballot in the fetched feed sits Democratic. That is the market's strongest argument for El-Sayed, and the panel weighed it.

That framework is the whole story of this board. The market prices the national environment and El-Sayed's primary strength into a clear Democratic favorite. The models, working only from those fetched facts, weighed the head-to-head polls and the nominee-strength gap more heavily and landed on the other side.

Here is the short version of the general-election board first, then every model's raw number.

The Board (General Election)

Outcome	Party & path	Market	Model blend
Mike Rogers	Republican; unopposed nominee, 2024 near-miss	~36¢	56.7%
Abdul El-Sayed	Democrat; primary favorite	~64¢	34.0%
Haley Stevens	Democrat; primary underdog, stronger vs. Rogers in polls	~2¢	8.3%
Mallory McMorrow	Democrat; suspended campaign, still on ballot	~1¢	1.0%

Every Seat's Number (General Election)

Outcome	Market	Fable	Opus	Sonnet	Blend	Gap (blend% − mkt¢)
Mike Rogers (Republican)	~36¢	54	58	58	56.7%	+21
Abdul El-Sayed (Democrat)	~64¢	37	33	32	34.0%	−30
Haley Stevens (Democrat)	~2¢	8	8	9	8.3%	+6
Mallory McMorrow (Democrat)	~1¢	1	1	1	1.0%	~0

These are model estimates, not predictions of fact and not financial or voting advice. The Blend column is the simple mean of the three model numbers shown in its row, after a revision round. The last column is a quick directional gap, the blend percentage minus the raw market price in cents. Because this is a closed field, the four prices already sum to about 100 cents (with a small spread), so the market barely needs de-vigging. The models scored who wins the November election, which is exactly how the contract settles.

A note on the panel, stated plainly. Our standard Model Verdict lineup runs eight seats. On this run five were unavailable: the ChatGPT and Gemini seats were not installed on the machine that produced these verdicts, and the provider that serves our GLM, Kimi, and DeepSeek seats returned a billing error on every call. This board therefore runs on three models, all from the Claude family: Claude Fable, Claude Opus, and Claude Sonnet. All three scored the race on August 4, 2026, without seeing any prices, working from a fetched, dated summary of the seat, the nominees, the race ratings, the structural voting history, and the complete live head-to-head polling. All three then read one another's anonymized reasoning and revised (two held, one nudged Rogers up and Stevens down). We flag the obvious limit honestly: three models from one family agreeing is weaker evidence than eight independent seats agreeing, and it raises the chance of a shared blind spot. Treat this board as a well-reasoned minority read against the market, not a confident call. Every model's number also lands in our back-tested Model Verdict ledger, where past price-blind calls are graded against settled results.

The board comes down to two forces the market and the models weigh differently: the national environment, and the head-to-head polls.

The one line that explains the whole split: the market is pricing a Democratic-leaning midterm, and the panel is pricing the polls, where the Democrat most likely to be nominated is the one running a few points behind a Republican who already came within 0.3 points of this seat. — the panel consensus

Why The Panel Makes Rogers The Favorite

All three seats made Rogers the favorite, and none put him below 54%. The case is not that Michigan leans Republican; it does not. The case is that the specific Democrat the primary looks set to nominate is the weaker of the two against Rogers, in a state that decides these races by a whisker.

Opus put the mechanism most directly: the "nominee-strength gap" between El-Sayed and Stevens is real and measured, and El-Sayed's roughly three-to-four-point head-to-head deficit against Rogers is the live state of the race unless you assume it closes. In a state where the last Senate race was decided by 0.3 points, a nominee polling behind starts as an underdog, not a 64-cent favorite. Fable and Sonnet reasoned along the same lines, with Fable noting that the polling average is dragged in both directions by a few partisan-sponsored surveys but still leaves El-Sayed on the wrong side of even.

The models did not ignore the market's best argument. Every seat explicitly weighed the Democratic-leaning midterm environment and the generic-ballot edge; that is exactly why none of them pushed Rogers past the high 50s. As Sonnet's revision put it, the environment is a real tailwind for the eventual Democrat, but a tailwind that has to overcome a measured polling deficit against a candidate who has run this state before is a reason the race is close, not a reason El-Sayed is a clear favorite.

Why The Market Makes El-Sayed The Favorite

The market is not irrational, and the models were careful not to treat it that way. The case for El-Sayed's ~64-cent price rests on three fetched facts the panel took seriously and simply weighed less:

• The Midterm Environment. The president's party has lost House seats in 18 of the last 20 midterms, and the live generic ballot sits Democratic. A Democratic-leaning national wind in a blue-leaning-in-midterms state is a real reason to expect the eventual Democrat to outperform the summer head-to-heads.
• Polls Tighten Toward The Nominee. Much of the head-to-head data was collected while the Democratic primary was still contested. Party voters often consolidate behind their nominee after the primary, which can lift a candidate several points; a market three months out may be pricing that expected consolidation rather than today's snapshot.
• Michigan's Recent Senate History Is Democratic. Democrats have held both Senate seats, and Slotkin won the 2024 open seat even as Trump carried the state. The market may simply trust that lean to reassert itself in a midterm.

The honest read is that this is a real disagreement about which signal dominates: the national environment (market) or the current head-to-head polls and the nominee-strength gap (panel). Michigan being a rated toss-up is the tell that neither side is being unreasonable.

Today's Catalyst: The Democratic Primary

The reason this page exists today is the August 4 primary, which settles the separate Kalshi nominee market and decides who actually faces Rogers. Late-July public polling had El-Sayed leading the primary by more than 15 points, and the primary market prices him as a heavy ~98-cent favorite.

These are model estimates, not predictions of fact and not financial or voting advice.

The Board (Democratic Primary)

Candidate	Lane	Market	Model blend
Abdul El-Sayed	Progressive	~98¢	87.6%
Haley Stevens	Moderate	~2¢	10.7%
Mallory McMorrow	Suspended campaign, still on ballot	~0.1¢	1.3%
Six Others (Anthony, Nessel, McDonald Rivet, Levin, Sahr, Tlaib)	Negligible support	~0.1¢ Each	<1% combined

The panel again lands a little softer than the market: it makes El-Sayed the heavy primary favorite at about 88%, but leaves Stevens a live ~11% chance rather than the market's ~2%. The seats' reasoning was consistent: a 15-point polling lead the day of a primary is a strong favorite, but primary polling is noisier than a 98-cent price implies, and a moderate candidate with real establishment support is not a 2-cent longshot until the votes are counted. This section will be refreshed once the primary is called tonight; if El-Sayed wins as expected, the general-election board above becomes an El-Sayed-versus-Rogers race, and the small share the models reserved for a Stevens nomination collapses into the general read.

Dark Horses The Panel Won't Dismiss

The general board looks like a two-name race, but the fetched record leaves a few scenarios the panel priced rather than ignored:

• Haley Stevens, The Panel's Dark Horse Who Polls Stronger Against Rogers. This is the panel's most interesting dark horse. In the same polls, Stevens runs even-to-ahead of Rogers while El-Sayed trails. If Stevens pulls the primary upset tonight, the general-election favorite likely flips back toward the Democrat, which is why the models give her ~8% in the general (four times her ~2-cent price) despite her long primary odds. Her value on this board is entirely gated by tonight's result.
• The Consolidation Swing. If post-primary consolidation and the midterm environment move El-Sayed several points, as the market is betting, his ~64-cent price is the correct read and the panel's fade is the mistake. The models' own spread is precisely the size of this disagreement, and every seat named it as the live risk to its own number.
• A Late Rogers Stumble In A State He Has Already Lost Once. Rogers came within 0.3 points in 2024 and still lost. A candidate who has narrowly lost a statewide race before is a proven competitor and a proven loser at the same time; the same fundamentals that make him the panel's favorite also make a second narrow defeat entirely ordinary here.

The Traps: Rows That Look Like Signal But Are Not

Transparency matters more than a clean story, so here is where this board could mislead you:

• A Three-Model Panel Is Not Eight. The single biggest caveat is structural, not about Michigan. With five seats offline, three closely related Claude models produced this verdict. Their agreement is real but correlated; do not read "the panel unanimously favors Rogers" as if eight independent methods converged. They did not this time.
• A 30-Point Gap Is A Disagreement, Not A Proven Edge. The panel sitting 30 points below El-Sayed's market price is a loud signal that two methods read the same race very differently. It is not proof the market is wrong. On a lightly traded general-election contract three months out, a wide gap can be the model missing something the crowd sees, especially the consolidation dynamic.
• Summer Head-To-Heads Carry A Shelf Life. Much of the polling that drives the panel's fade was taken during a live primary. If you believe those numbers move once the nominee is set, you should discount the panel's El-Sayed fade accordingly, which is exactly the market's bet.
• The General Contracts Are Thin. The primary market is deep; the general-election lines are lightly traded with a wider spread. Their prices are a lean, not a depth-tested consensus, so read the ~64 and ~36 as approximate.

How To Read This Board

The board sorts into three honest reads, and which one fits depends on whether you trust the national environment or the current polls more.

• You Trust The Head-To-Head Polls: The Rogers Case. El-Sayed trails Rogers by about four points on the non-partisan polling average, Michigan decides Senate races by a point or two, and the Republican already came within 0.3 points of this seat. That is the panel's read, and it makes Rogers a slight favorite.
• You Trust The Environment: The El-Sayed Case. A Democratic-leaning midterm, a Democratic generic-ballot edge, and expected post-primary consolidation are real forces that summer head-to-heads may understate. That is the market's read, and it makes El-Sayed a clear favorite.
• You Want The Honest Bottom Line: A Rated Toss-Up. Cook and Sabato both call this seat a toss-up for a reason. Whoever emerges from tonight's primary faces a real coin-flip environment, and the market-versus-model split on this page is the clearest sign that reasonable methods disagree about which way it tips.

How To Read These Prices

The models can tell you where they split from the crowd, but the number in front of you still has to be understood in context. Here is how to read a market like this one, plus the honest friction of an event-driven board.

Work an example. Say El-Sayed's contract shows a 64-cent YES price. Taken at face value, that implies about a 64% chance he wins the seat. To get a fair, spread-free read you normalize the whole field so every outcome's price sums to 100. On this board the four prices sum to roughly 103 cents, so there is a small spread to remove: 64 divided by 103 is about 62%, a touch below the raw price. The lesson holds anyway: always check what a fair version of a price looks like before comparing it to anything, because the favorite's number is usually a shade softer than it looks. That habit is the whole idea behind converting Kalshi's cents into real odds before comparing anything.

That same discipline of comparing a number to its fair value is what line shopping and finding the +EV side is about. OddsShopper's odds screen and free expert picks cover the sports slate rather than election futures, and the Michigan race itself lives on Kalshi, but the method is the same one our Texas Senate model verdict uses on a very different field, and it complements our broader 2026 Senate election odds breakdown of the whole chamber. Every number here also lands in our back-tested Model Verdict ledger.

A few honest points before reading too much into any number in the tables above:

• Eligibility. Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook. You must be 18 or older and located in an eligible state, and availability varies by state and can change.
• Liquidity, Spread, And Fees. The primary market is deeply traded; the general-election contracts are light, with a wider spread, so their prices are a lean rather than a depth-tested consensus. Your fill depends on the spread at the moment you act, and Kalshi charges trading fees on some markets that eat into any edge, so factor them in before acting.
• A Moving Story. This race has a primary settling tonight, close general-election polls, and an election three months out; new events before November 3 can move it, and these estimates reflect only what was known on the morning of August 4. Prices shown on this page refresh during the news wave.
• Not Advice. Everything here is analysis for information, not a recommendation to trade or to vote. These are model estimates, not predictions of fact and not financial or voting advice.

The Bottom Line

The market and the models are looking at the same Michigan Senate race and reaching opposite verdicts. Kalshi prices Democrat Abdul El-Sayed as a clear favorite near 64 cents; our price-blind panel makes Republican Mike Rogers a slight favorite at about 57%, with El-Sayed at 34% and Haley Stevens, the Democrat who polls stronger against Rogers, at about 8%. The disagreement is not noise: it is a real split between the market's read of a Democratic-leaning midterm and the panel's read of head-to-head polls that have the likely nominee running behind in a state that decides Senate races by fractions of a point. Take the honest caveats with it, three correlated models this run, thin general-election contracts, and summer polls that may tighten, and the fair conclusion is the one Cook and Sabato already reached: this is a toss-up, and tonight's primary sets the terms. If this method is your thing, our Texas Senate model verdict and our 2026 Senate election odds chamber breakdown are the natural next reads.

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.