Kalshi MLB Odds Today: The AI Panel Vs. The Market, August 12

Last updated: August 12, 2026, 1:10 PM ET. Edition 1. This page is refreshed daily at this same address.

Edition 1. Running record: 0-0. Every call below grades tomorrow against Kalshi settlement only: the market resolved yes or the market resolved no, nothing hand-graded, and the running record at the top of each edition carries the tally forward. Today it starts at zero because there is nothing to count yet.

The Quick Answer

Kalshi MLB odds price every game in cents, and a 54.5-cent contract is the market saying 54.5%. Each day we line that board up against the Stokastic Sims run projections and put a six-seat AI panel on the three biggest disagreements. Today the panel goes 5-1 with the model on the Orioles, deadlocks 3-3 in Pittsburgh-Miami, and votes 6-0 for the Mariners' side of a projected coin flip priced at 45.5 cents. The full 15-game board, the seat-by-seat reasoning, and the one price the whole panel refused to accept are below.

Wednesday's board is mostly an argument the market wins by forfeit. Across 15 games, the Sims' projected-runs edge and the market's favorite point the same direction on 13. That agreement is the normal state of a liquid market, and it is exactly why the seams matter: two games flip outright, and one price sits nine cents away from what the projections describe as a coin flip. The unanimous call comes last, and it is the one I keep coming back to.

A note on method before the board: the Sims produce projected team runs, not win percentages, so we never convert one into the other. A run edge is a lean, not a probability, and the panel's job is to decide whether that lean or the market's price deserves the benefit of the doubt. Prices below are the midpoint of the live Kalshi bid and ask, fetched as of 1:10 PM ET on August 12, before first pitch of the early game. If cents-to-moneyline translation is new to you, our Kalshi American odds guide covers it in full.

The Full Kalshi MLB Board, August 12

Game	First pitch	Kalshi price (mid)	Sims projected runs	Model and market agree?
Orioles At Twins	1:40 PM ET	MIN 50.5¢ / BAL 49.5¢	BAL 4.7 / MIN 4.5	No: flip
Phillies At Cardinals	2:15 PM ET	PHI 60.5¢ / STL 39.5¢	PHI 4.5 / STL 3.5	Yes
Rays At Athletics	3:05 PM ET	TB 67.5¢ / ATH 32.5¢	TB 5.8 / ATH 3.9	Yes
Rockies At Diamondbacks	3:40 PM ET	ARI 61.5¢ / COL 38.5¢	COL 4.4 / ARI 5.2	Yes
Astros At Giants	3:45 PM ET	HOU 57.5¢ / SF 42.5¢	HOU 4.7 / SF 4.0	Yes
Brewers At Padres	4:10 PM ET	MIL 51.5¢ / SD 48.5¢	MIL 4.3 / SD 3.9	Yes
Guardians At Tigers	6:40 PM ET	DET 54.5¢ / CLE 45.5¢	CLE 4.2 / DET 4.3	Barely
Pirates At Marlins	6:40 PM ET	MIA 51.5¢ / PIT 48.5¢	PIT 4.3 / MIA 4.2	No: flip
Cubs At Nationals	6:45 PM ET	CHC 60.5¢ / WSH 39.5¢	CHC 5.5 / WSH 4.3	Yes
Mariners At Yankees	7:05 PM ET	NYY 54.5¢ / SEA 45.5¢	SEA 4.5 / NYY 4.6	Barely
Red Sox At Blue Jays	7:07 PM ET	BOS 53.5¢ / TOR 46.5¢	BOS 3.9 / TOR 3.6	Yes
Mets At Braves	7:15 PM ET	ATL 62.5¢ / NYM 37.5¢	NYM 4.1 / ATL 4.9	Yes
Reds At White Sox	7:40 PM ET	CWS 57.5¢ / CIN 42.5¢	CIN 4.3 / CWS 4.7	Yes
Royals At Dodgers	10:10 PM ET	LAD 66.5¢ / KC 33.5¢	KC 4.1 / LAD 5.4	Yes
Rangers At Angels	10:10 PM ET	TEX 56.5¢ / LAA 43.5¢	TEX 4.9 / LAA 4.3	Yes

The row where model and market agree hardest is Tampa Bay. The Rays carry the heaviest price on the board at 67.5 cents, a 67.5% implied chance that outranks even the Dodgers at 66.5%, and the Sims back it with the day's widest run gap, 5.8 to 3.9 against the Athletics. When the projections and the price shout the same name, there is no edition of this page that will argue. The three games below are where they don't.

Disagreement 1: Orioles At Twins, The Board's First Flip

Market: Twins 50.5¢ (-102, a 50.5% implied chance) / Orioles 49.5¢ (+102, 49.5% implied). Sims: Orioles 4.7 projected runs, Twins 4.5.

Shane Baz against Zebby Matthews at Target Field, and the market calls it a pure coin flip with Minnesota a single cent above the line. Stokastic's Sims flip it: Baltimore projects for more runs, 4.7 to 4.5, which makes the Orioles the cheaper side of the model's favorite. Six seats read it, no seat names disclosed beyond the count:

• Seat 1 (Model): The projections' favorite is the plus-money side; that combination is the whole reason to look.
• Seat 2 (Model): Higher projected total at a coin-flip price is a small edge, but it is an edge.
• Seat 3 (Market): Two-tenths of a run is noise, and a one-cent spread already says coin flip.
• Seat 4 (Model): The market prices a pick'em while the model scores Baltimore ahead; take the cheaper lean.
• Seat 5 (Model): A 0.2-run edge against a near-pick'em price is thin value, but real.
• Seat 6 (Model): The projections put Baltimore ahead and the price has them as the dog; that mismatch is the read.

Panel call: Orioles, 5-1 with the model. This one grades tonight. First pitch was 1:40 PM ET, minutes after these prices were captured, so the game will be underway or finished by the time most readers see this. The call stands as written and grades on settlement like everything else here.

Disagreement 2: Pirates At Marlins, Where The Panel Split

Market: Marlins 51.5¢ (-106, 51.5% implied) / Pirates 48.5¢ (+106, 48.5% implied). Sims: Pirates 4.3 projected runs, Marlins 4.2.

The second flip is thinner. Carmen Mlodzinski against Janson Junk in Miami, the Marlins three cents above even, and the Sims leaning Pittsburgh by a tenth of a run. The panel could not resolve it:

• Seats 1, 2 And 3 (Market): All three landed on the same core objection: a 0.1-run gap sits inside model noise, and Miami's home park and better record are worth the three cents.
• Seats 4, 5 And 6 (Model): All three took the other horn: in a projected toss-up, the side priced under 50 cents is the value side by definition.

Panel call: none, 3-3 deadlock. A tie produces no gradeable call, and we would rather print a deadlock than manufacture a majority. That refusal is itself information: this is the kind of price where reasonable models simply differ, which usually means the market has it about right.

Disagreement 3: Mariners At 45.5 Cents, The Unanimous One

Market: Yankees 54.5¢ (-120, 54.5% implied) / Mariners 45.5¢ (+120, 45.5% implied). Sims: Yankees 4.6 projected runs, Mariners 4.5.

Here is the payoff of the promise up top. Bryce Miller against Will Warren at Yankee Stadium, and the market hands New York a nine-cent gap, the largest confidence spread of the three games in this edition. Stokastic's run projections see almost nothing between these teams: 4.6 runs to 4.5, and that razor-thin lean actually sits on New York's side, the same side the market likes. Nobody on the panel disputes that. What the seats refused to accept is the price: a nine-cent premium on a game the projections call even. Remember the Pittsburgh game, where a one-tenth run gap split the panel down the middle? The gap is the same size here, with one honest difference: in Miami the thin lean sat on the cheap side, while here it sits on the expensive one, so this case rests entirely on treating 4.6 to 4.5 as even. What moved every seat anyway is the price, three cents from even there, nine cents here:

• Seat 1 (Model): A projected coin flip priced nine cents apart is the board's clearest mispricing.
• Seat 2 (Model): The discount on Seattle overpays for the Yankee premium.
• Seat 3 (Model): The same seat that called the two run-gap flips noise takes the model here: this disagreement is about the price gap, not the run gap.
• Seat 4 (Model): A near coin flip with one side at 45.5 cents is the contrarian read that actually has data behind it.
• Seat 5 (Model): The market gap is the outlier, not the projection.
• Seat 6 (Model): Against a nine-cent spread, the priced-down side of a projected even game is the value.

Panel call: Mariners, 6-0. This one is a price call, not a projection call: the panel is not saying the model likes Seattle, it is saying 45.5 cents underprices an even game. Worth one footnote: the nearly identical Guardians-Tigers game (Detroit 54.5 cents, Sims 4.3 to 4.2) has the same shape, a nine-cent gap on a 0.1-run projection, and lost the slot only because the Bronx game is the bigger stage. For what it is worth, DraftKings pricing Foster Griffin under 5.5 strikeouts at -125 says the books expect a modest strikeout count from Cleveland's starter too.

Worked Example: What 45.5 Cents Actually Pays

Read the Mariners price the way the market does. A yes contract at 45.5 cents pays 100 cents if Seattle wins, so the market is treating the Mariners as a 45.5% proposition, roughly +120 in moneyline terms. A sportsbook quoting the same game will usually hang a number in that neighborhood with its own margin baked in; how the two venues charge you differs, and our Kalshi fees versus sportsbook vig breakdown walks the arithmetic. The practical move this page recommends every day is the boring one: compare the cents price against the moneylines on our live odds screen and shop the number across every major book before deciding where a position belongs. Whichever side of today's board you land on, the price you pay decides most of the outcome. That is the entire thesis of how to win on Kalshi, and it applies without modification here.

How Every Call Grades

The rules of this page, printed in every edition so nobody has to trust our memory:

• A call is a panel majority siding with the model or the market on a named team. Today produced two: Orioles (5-1, a projection call) and Mariners (6-0, a price call). The 3-3 deadlock produces none.
• Every call grades tomorrow against Kalshi settlement only: each contract settles once the final out is recorded, to 100 cents or zero, and today's two calls grade in the August 13 edition. The market resolved yes or it resolved no. No adjustments, no hand-grading, no "it would have hit if."
• The running record sits at the top of every edition. Edition 1 starts 0-0, plainly, because no call has settled yet.
• Seat counts are disclosed honestly every day. Six seats returned reads today; when a seat fails to answer, the count shrinks and we say so.

Nothing on this page is a recommendation to place a trade. It is a model, a market, and a scoreboard that will make one of them look smarter by tomorrow morning. If you want human picks with skin in the game alongside the machine reads, our free expert picks run every day, and the season-long version of this model-versus-market exercise lives on our Kalshi MLB playoffs odds board, with the MLB home run leader model verdict covering the season-long player side.

FAQ

What do Kalshi MLB odds in cents mean? Each contract settles at 100 cents if the named team wins and zero if it loses, so the price is the market's implied probability: 54.5 cents on the Yankees means the market gives New York roughly a 54.5% chance. The cents-to-American-odds translation is a straight formula, and how Kalshi makes money on those trades is its own answer.

Do the Stokastic Sims produce win probabilities? Not on this surface. What they produce is projected team runs for each side of every game. This page reports that projection as a lean, agree or disagree with the market, and never mints a win percentage the model did not produce.

When does the record update? Every edition grades the previous edition's calls against Kalshi settlement and carries the running tally at the top. Tomorrow's page will show 2-0, 1-1 or 0-2, and how the seats that made those calls voted.

The board resets tomorrow. The market will agree with the model on most of it, the way it did today, and this page will keep score on the gap between them: two flips, one deadlock, and a unanimous panel staring at a nine-cent spread the projections call even.