NHL Hart Trophy Odds 2026-27: McDavid, MacKinnon And A Board Full Of Traps
The 2026-27 Hart Trophy race opens with a fact that sounds backwards: Connor McDavid put up 138 points last season, led the league by eight, and did not win MVP. Nikita Kucherov did, with 130 points on a Tampa Bay team that finished 13 standings points clear of Edmonton. That vote is the single most important input on Kalshi's new 28-player Hart board, because it told everyone exactly how the voters break a tie: production first, team quality as the tiebreak. We put the whole board in front of an eight-model AI panel, price-blind, and the panel's answer is a dead heat at the top, a reigning winner the market has not even traded yet, and four thin offseason prices it thinks are badly wrong.
The Quick Answer
Eight AI models, given the full fetched data and no prices, make the 2026-27 Hart a two-man race: Connor McDavid at 18.1% and Nathan MacKinnon at 18.0%, with defending winner Nikita Kucherov third at 11.3%. The panel's sharpest disagreements with the early market are fades: it prices Connor Bedard near 1% against a 41-cent last trade, and Kirill Kaprizov near 4% against 30 cents. The full model-by-model board, the traps in the thin prices, and the dark horses at least one model refuses to dismiss are all below.
The Market
| Venue | Kalshi — a CFTC-regulated event-contract exchange (18+; availability varies by state, as of July 2026) |
| The Contract | One market per player, 28 listed: "If [player] wins the NHL Hart Memorial Trophy in the 2026-27 NHL season, then the market resolves to Yes." Markets expire June 30, 2027. There is no field contract, so an unlisted winner settles every market No. |
| Liquidity | Early and thin: since listing on June 29, only 9 of 28 candidates have traded at all (about 2,100 contracts total in 30 days). Last-trade prices below are stale by default; the four most-traded names sum past 100% on last prints alone. |
| Timestamps | Verdicts generated July 29, 2026; prices as of July 29, 2026 |
Every price on this page is a last-trade candlestick close, quoted with its staleness on display, and 19 of the 28 contracts have never traded, including Kucherov's. On a board this thin the panel's numbers are doing the work the order book cannot do yet. If reading an exchange board is new to you, our companion on how to bet NHL on Kalshi covers the mechanics, and the liquidity guide explains why thin books lie.
How The Panel Works
Eight models (ChatGPT GPT-5.5, Claude Fable, Claude Opus, Claude Sonnet, Gemini 3.1 Pro, GLM 5.2, Kimi K3, DeepSeek V4) each received the same fetched data card: all 28 candidates with their full 2025-26 stat lines and current 2026-27 teams from live roster data, final league standings, the last ten Hart votes from the NHL's records API, and the exchange's own settlement rule. No model saw a market price. Each seat priced the entire board as one probability distribution summing to 100%, including a bucket for an unlisted winner. After round one, every seat read the other seven's anonymized reasoning and could revise; six of the eight completed that revision round (two seats returned unusable output on revision and kept their round-one numbers). The headline blend is the equal-weight average of each seat's final distribution. These numbers are model estimates, not predictions of fact and not financial advice.
The Board
| Player | 2025-26 (current team) | Last Kalshi Trade | AI Blend |
|---|---|---|---|
| Connor McDavid | EDM, 138 pts (48G/90A) in 2025-26 | 23c | 18.1% |
| Nathan MacKinnon | COL, 127 pts (53G/74A) in 2025-26 | 26c | 18.0% |
| Nikita Kucherov | TBL, 130 pts (44G/86A) in 2025-26 | no trades yet | 11.3% |
| Macklin Celebrini | SJS, 115 pts (45G/70A) in 2025-26 | 27c | 7.8% |
| Leon Draisaitl | EDM, 97 pts (35G/62A) in 2025-26 | 18c | 5.4% |
| Nick Suzuki | MTL, 101 pts (29G/72A) in 2025-26 | no trades yet | 3.8% |
| Kirill Kaprizov | MIN, 89 pts (45G/44A) in 2025-26 | 30c | 3.7% |
| David Pastrnak | BOS, 100 pts (29G/71A) in 2025-26 | no trades yet | 3.5% |
| Cole Caufield | MTL, 88 pts (51G/37A) in 2025-26 | no trades yet | 3.2% |
| Cale Makar | COL, 79 pts (20G/59A) in 2025-26 | no trades yet | 2.4% |
| Jack Eichel | VGK, 90 pts (27G/63A) in 2025-26 | no trades yet | 2.2% |
| Jack Hughes | NJD, 77 pts (27G/50A) in 2025-26 | 34c | 1.6% |
| Auston Matthews | TOR, 53 pts (27G/26A) in 2025-26 | 10c | 1.4% |
| The Field (anyone unlisted) | - | not offered | 7.9% |
The 15 candidates below Matthews all blend under 1.3%: Rantanen (1.2%), Vasilevskiy (1.0%), Quinn Hughes and Kyle Connor (0.9%), Marner (0.8%), Hellebuyck and Bedard (0.7%), Nylander, Shesterkin and Matthew Tkachuk (0.6%), Brady Tkachuk (0.5%), Crosby (0.4%), Panarin (0.3%), Ovechkin and Dorofeyev (0.2%). Every seat's number for every one of them is in the full table further down.
The Two-Man Race, And Why It Is Actually Three
The top of the board is as close as a blend can get: 18.1% to 18.0%. What separates the two cases is which lesson of the 2025-26 vote a model weighted more.
The McDavid case is the base rate. He has finished winner, runner-up or finalist in seven of the last ten Hart votes, and he just led the league in scoring by eight points in a full 82-game season.
"The only thing that beat him in 2025-26 was the team-quality tiebreak the card flags, EDM at 93 standings points versus TBL at 106; that is a re-rollable coin, not a permanent condition, and if Edmonton lands back in the mid-100s his scoring lead makes him the default winner." — Claude Opus
The MacKinnon case is the template. Voters just rewarded elite production on a strong team over raw scoring on a weaker one, and nobody fits that template like a 127-point, plus-57 season on the league-best 121-point Avalanche.
"MacKinnon pairs the league's best team with an MVP-caliber statline and a proven Hart pedigree, replicating the exact team-over-scoring-title pattern that just decided the 2025-26 vote." — Claude Sonnet
Kimi K3, the panel's most McDavid-leaning seat at 22%, pushed back on that template in the revision round and held:
"[T]hat vote ranked MacKinnon third behind McDavid with the very team-context inputs its proponents cite; team quality was the tiebreak against McDavid, not a crown for MacKinnon, who was out-scored by both men ahead of him." — Kimi K3
And then there is the third man. Kucherov is the reigning MVP, put up 130 points on a 106-point team, and returns to the same situation. The panel gives him 11.3%, every seat between 9.7% and 14%, which makes him the most agreed-upon number on the board. The market gives him nothing at all: not one contract has traded. On a mature board that would be a screaming signal; here it is just an empty order book, which is its own kind of information about how early this market is.
Where The Panel Splits From The Market
The thin prices are where this board gets interesting, because a last trade on a few contracts is not a market opinion, and the panel treats it accordingly. Four gaps stand out, all fades:
- Connor Bedard: 41c last trade, 0.7% blend. The widest gap on the board. Bedard's 75 points in 69 games is real growth, but Chicago finished at 72 standings points, second-worst in the league, and no model sees a Hart path through that. The 41c printed on 11 contracts.
- Kirill Kaprizov: 30c, 3.7% blend. A 45-goal season on a 104-point team is a fine resume, but every seat priced him behind teammate-quality peers with stronger scoring cases. The panel's range is 2.7% to 6%; the price implies five times the blend.
- Macklin Celebrini: 27c, 7.8% blend. The market's most-traded name, and the panel's most interesting disagreement. Every model respects the trajectory; several priced the narrative aggressively. But the blend still sits at less than a third of the price, because a Hart run requires San Jose, at 86 points last season, to leap into the playoff picture.
- Alex Ovechkin: 50c, 0.2% blend. Not a real price: 19 contracts total. A 64-point, minus-5 season has no MVP case, and the panel says so unanimously. This row is the single best illustration of why last trades on thin books are not probabilities.
McDavid at 23c and MacKinnon at 26c are the prices closest to panel opinion: the McDavid print sits within five points of the blend, the MacKinnon print eight above it. Draisaitl at 18c against a 5.4% blend leans rich; Jack Hughes at 34c against 1.6% is another thin-print artifact (125 contracts). Treat every one of these as a note on the board's youth, not a graded edge: with books this empty, the next real order rewrites the price.
The Field: The Panel's Quietest Big Number
The revision round moved one number more than any other: the chance the winner is not on the list at all. DeepSeek opened at 23% on the field, arguing the 28 names omit anyone from Carolina (113 points, second-best in the league) and Buffalo (109, fourth). The panel pushed back on the size but conceded the structure: Gemini went from 1% to 14%, Opus from 8% to 11.1%, GLM from zero to 7.5%, while DeepSeek itself came down from 23% to 7.5% after crediting Kucherov's case more. The blend settled at 7.9%, which quietly makes "someone not listed" the fourth-largest outcome on the board, ahead of Celebrini.
"The candidate list completely omits the league's 2nd-best (Carolina, 113 pts) and 4th-best (Buffalo, 109 pts) teams. Given the proven voter preference for team success, an unlisted star from these rosters carries immense win probability." — Gemini 3.1 Pro
Remember the settlement mechanics: there is no field contract to buy. If an unlisted player wins, all 28 markets settle No. That makes the 7.9% a pure headwind on every Yes position, and it is priced into every blend number above.
Dark Horses The Panel Won't Dismiss
- Nick Suzuki (3.8%). The quiet climber: 101 points with a plus-37 on a 106-point Montreal team. Claude Sonnet put him at 7%, the panel's highest, on the strength of exactly the profile the 2025-26 vote rewarded: a team's clear engine at the top of the standings.
- Cole Caufield (3.2%). A 51-goal season already banked, on the same rising Canadiens core. Sonnet again led at 5%, treating the Suzuki-Caufield pairing as two draws from the same winning narrative.
- Cale Makar (2.4%). The positional long shot. ChatGPT went to 4%, noting a defenseman with 79 points on the league's best team only needs the voters' occasional appetite for history; no defenseman appears among the fetched last ten winners, which is both the obstacle and the story.
- Andrei Vasilevskiy (1.0%). The goalie flyer. A 39-15, .912 season on a contender keeps him alive at 2% for Kimi K3, with Hellebuyck's 2024-25 win as fetched proof the door opens when no skater runs away with it.
Every Seat's Number
| Player | ChatGPT (GPT-5.5) | Claude Fable | Claude Opus | Claude Sonnet | Gemini 3.1 Pro | GLM 5.2 | Kimi K3 | DeepSeek V4 | Blend |
|---|---|---|---|---|---|---|---|---|---|
| Connor McDavid | 18.0% | 19.4% | 21.1% | 14.0% | 16.0% | 15.1% | 22.0% | 19.5% | 18.1% |
| Nathan MacKinnon | 15.0% | 15.5% | 16.1% | 20.0% | 18.0% | 17.0% | 18.0% | 24.5% | 18.0% |
| Nikita Kucherov | 12.0% | 9.7% | 10.1% | 13.3% | 10.0% | 10.4% | 14.0% | 11.0% | 11.3% |
| Macklin Celebrini | 9.0% | 8.7% | 8.6% | 5.0% | 8.0% | 4.7% | 10.0% | 8.0% | 7.8% |
| Leon Draisaitl | 7.0% | 4.9% | 5.0% | 3.0% | 5.0% | 6.6% | 6.0% | 6.0% | 5.4% |
| Nick Suzuki | 1.5% | 2.9% | 2.5% | 7.0% | 4.0% | 3.8% | 4.0% | 5.0% | 3.8% |
| Kirill Kaprizov | 2.7% | 3.9% | 3.0% | 6.0% | 4.0% | 3.8% | 3.0% | 3.0% | 3.7% |
| David Pastrnak | 5.5% | 2.9% | 2.5% | 4.0% | 4.0% | 2.8% | 3.0% | 3.0% | 3.5% |
| Cole Caufield | 3.5% | 2.9% | 2.5% | 5.0% | 3.0% | 2.8% | 3.0% | 3.0% | 3.2% |
| Cale Makar | 4.0% | 1.9% | 1.8% | 2.5% | 2.0% | 2.8% | 2.0% | 2.0% | 2.4% |
| Jack Eichel | 1.6% | 2.9% | 2.0% | 3.5% | 2.0% | 2.8% | 2.0% | 1.0% | 2.2% |
| Jack Hughes | 3.0% | 1.9% | 2.0% | 1.5% | 1.0% | 1.9% | 1.0% | 0.5% | 1.6% |
| Auston Matthews | 2.8% | 1.9% | 1.2% | 1.0% | 2.0% | 1.9% | 0.0% | 0.5% | 1.4% |
| Mikko Rantanen | 1.8% | 1.9% | 1.5% | 2.5% | 0.0% | 0.9% | 0.0% | 1.0% | 1.2% |
| Andrei Vasilevskiy | 0.7% | 0.5% | 1.0% | 0.7% | 1.0% | 1.9% | 2.0% | 0.0% | 1.0% |
| Quinn Hughes | 1.0% | 0.5% | 0.8% | 0.4% | 1.0% | 1.9% | 1.0% | 1.0% | 0.9% |
| Kyle Connor | 1.4% | 1.0% | 0.8% | 2.5% | 1.0% | 0.0% | 0.0% | 0.5% | 0.9% |
| Mitch Marner | 0.5% | 1.0% | 0.0% | 1.2% | 1.0% | 1.9% | 0.0% | 1.0% | 0.8% |
| Connor Hellebuyck | 0.6% | 0.5% | 0.7% | 1.5% | 1.0% | 0.9% | 0.0% | 0.0% | 0.7% |
| Connor Bedard | 0.5% | 1.0% | 1.5% | 0.8% | 0.0% | 0.9% | 0.0% | 0.5% | 0.7% |
| William Nylander | 0.3% | 1.9% | 1.0% | 0.4% | 0.0% | 0.9% | 0.0% | 0.5% | 0.6% |
| Igor Shesterkin | 0.3% | 0.5% | 0.8% | 0.5% | 1.0% | 0.9% | 1.0% | 0.0% | 0.6% |
| Matthew Tkachuk | 0.4% | 1.0% | 0.6% | 0.2% | 1.0% | 0.9% | 0.0% | 0.5% | 0.6% |
| Brady Tkachuk | 0.3% | 1.0% | 0.6% | 0.3% | 0.0% | 0.9% | 0.0% | 0.5% | 0.5% |
| Sidney Crosby | 0.2% | 1.0% | 0.4% | 0.6% | 0.0% | 0.9% | 0.0% | 0.0% | 0.4% |
| Artemi Panarin | 0.5% | 0.5% | 0.3% | 0.3% | 0.0% | 0.9% | 0.0% | 0.0% | 0.3% |
| Alex Ovechkin | 0.1% | 0.3% | 0.2% | 0.1% | 0.0% | 0.9% | 0.0% | 0.0% | 0.2% |
| Pavel Dorofeyev | 0.1% | 0.2% | 0.1% | 0.2% | 0.0% | 0.9% | 0.0% | 0.0% | 0.2% |
| The Field | 5.7% | 7.8% | 11.1% | 2.0% | 14.0% | 7.5% | 8.0% | 7.5% | 7.9% |
Columns are each seat's final number: the revised round-two figure for the six seats that completed revision, round one for ChatGPT and Claude Sonnet. Rows may not sum to exactly 100% due to rounding.
A Worked Example: Reading These Prices Like A Bettor
A worked example makes the whole board concrete. Take the McDavid contract at its 23-cent last trade. A cent price reads as an implied probability, so 23c is roughly a 23% chance, which converts to about +335 in American odds; our Kalshi odds guide has the full conversion math. The panel's blend says 18.1%, about +452. That five-point gap is the market being modestly more optimistic than the machines, and on a two-sided exchange you could take either side of it. Now run the same exercise on Bedard: 41c implies about +144, while the panel's 0.7% implies longer than +14000. Nobody is really disagreeing about Bedard by two orders of magnitude; the 41c is 11 contracts of noise, which is why the liquidity column matters as much as the price column.
When sportsbooks hang their own 2026-27 Hart futures as the season approaches, the discipline is ordinary line shopping: put the book's number through a de-vig pass to strip the hold, set it next to the exchange price and the panel blend, and see which of the three tells a different story. The odds comparison on our live odds screen does the no-vig math on NHL game markets all season, and the same convert-then-compare habit is what makes an awards board readable. None of that is a recommendation to trade this market; it is how to read one that is still this young.
What Would Move This Board
Three things reprice everything. First, where the uncertainty markets resolve: Kalshi lists open next-team markets on McDavid, Matthews and Hellebuyck among others, and several seats explicitly discounted McDavid for the contract noise those markets represent; an early extension in Edmonton removes his biggest stated drag. Second, October standings: the entire top of the panel's board is conditioned on the 2025-26 lesson that team quality breaks ties, so the first month of Edmonton's and Colorado's seasons does more to this race than any point streak. The same panel's read on those team races is already on the board in our 32-team Stanley Cup verdict and the Eastern and Western Conference boards. Third, actual liquidity: 19 of these 28 contracts have never traded, and the first real order flow will make some of the gaps above vanish and reveal which ones were real. We will re-run this panel when the board has prices worth grading.
To be explicit about what this page is: the percentages are model estimates, not predictions of fact and not financial advice. Model panels have public, graded track records on our verdicts, and they are wrong in graded, measurable ways. Kalshi availability varies by state and over time; you must be 18 or older to trade event contracts, and every trading decision is your own.
Hottest Prediction Markets Right Now
- Fed decision in July? · $56M traded
- 2027 Pro Football Champion · $50M traded
- Rocket Classic Winner · $48M traded
- Pro Baseball Champion · $39M traded
- Bitcoin price at the end of 2026 · $29M traded
- 2027 Pro Basketball Champion · $18M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub.
Event contracts involve risk and are not appropriate for everyone. 18+. Availability varies by state. Trade responsibly.
