Stanley Cup Final 2026: The Numbers Behind Hurricanes vs. Golden Knights

2026-06-02 • NHLForecasts

The matchup is locked: the Carolina Hurricanes against the Vegas Golden Knights, with the puck dropping in Raleigh on Tuesday, June 2. Our season-long model treats this as one of the closest Finals it can produce — but the way it arrives there is more interesting than the final number. Carolina and Vegas have reached the same destination by almost opposite routes, and the model rewards each of them for a different thing.

Below is the whole series through the lens of our forecasting system: win probability, the shape of the series, and the underlying expected-goals profile that drives it all. One note up front, because it changes the story: our numbers do not agree with the popular framing that Vegas is riding hot goaltending. By expected goals, the opposite is true.

The series number

Run the model forward from the current bracket and it lands on Carolina 53% · Vegas 47% to lift the Cup (52.8% to be precise). That is roughly a one-in-twenty edge — a lean, not a verdict. Carolina earns its sliver of an advantage from process and home ice; Vegas earns its near-parity from a defensive structure that has quietly been one of the best in the field.

Two roads to the Final

Carolina was the more dominant team in the bracket, dropping a single game all spring. Vegas took the harder road and saved its loudest statement for last.

Sweeping the Presidents' Trophy-winning Avalanche in the Western Final is the single most impressive series any team produced this postseason. Carolina's body of work is steadier; Vegas's peak is higher.

Game 1 projection — Tue, June 2 @ Carolina

• Win probability: Carolina 64.0% · Vegas 36.0%
• Expected goals: Carolina 3.35 — Vegas 2.45
• Median projected total: 5.5 goals

Home ice is doing real work here. The model makes Carolina a clear favorite in Raleigh and — notably — never installs Vegas as a favorite in any game of the series. On Vegas ice it narrows to something close to a coin flip rather than flipping outright. The one scheduling wrinkle in the Knights' favor: they closed out Colorado before Carolina finished with Montréal, so they enter Game 1 the better-rested side. You can see the live matchup page here.

How long does it go?

This profiles as a long series. The model gives roughly a 63% chance it reaches six or seven games, and the single most likely outcome is Carolina in seven. A sweep in either direction is a combined ~12% — possible, not probable.

The expected-goals fingerprint

This is where the styles separate. The table below is full-season expected-goals data from our shot model — per-game rates, share of expected goals, high-danger chances created and conceded, and finishing relative to expectation. (We use rates and share here rather than raw shot counts, which are attempt-basis and read high to a hockey eye.)

Carolina is a volume-and-territory team. They generate 51.5 shots a night, own 57.1% of the expected goals in their games, and bury opponents in the offensive zone. The one persistent flaw: they finish at just 0.90× expectation — their chance creation has outrun their conversion all season. They have been winning on sheer wave-after-wave pressure, not red-hot shooting. If their finishing ever catches up to their process, this is the scariest team in the league.

Vegas is the counterpuncher. They take fewer shots (43.9/game) but at slightly higher quality, and the standout number is on the defensive side: 209 high-danger chances conceded all season, against Carolina's 273. Vegas is content to be out-shot because it suffocates the dangerous looks — exactly the profile that frustrated a possession monster like Colorado. The whole series compresses into one question: can Vegas's structure keep blunting high-danger chances long enough for its slightly better finishing to matter?

Between the pipes — what the model actually sees

Goaltending is usually billed as the swing factor in a Final, so it is worth being precise about it — and here the model pushes back on the easy narrative. The relevant comparison is the two playoff starters: Frederik Andersen for Carolina and Carter Hart for Vegas. Goals saved above expected (GSAx) measures saves against the difficulty of the shots faced, which is a far better read than save percentage alone.

The honest read: the goaltending edge in this series belongs to Carolina, not Vegas. Andersen has been a genuine difference-maker — a .928 save percentage, three shutouts, and +2.8 goals saved above expected, all reinforcing Carolina's league-best team save metric (+21.9 GSAx on the season). Hart's .920 looks fine on the surface and his 12–4 record looks better, but layer in shot difficulty and he sits at −8.2 GSAx — eight-plus goals worse than an average goalie would have allowed on the same shots, with zero shutouts. Vegas is here in spite of its goaltending, carried by that elite chance-suppression defense, not because a hot netminder is stealing games.

That reframes the matchup. The popular story — "both nets are scorching, whoever blinks loses" — doesn't survive contact with the expected-goals data. If anything, Vegas's path to the Cup probably runs through Hart finally outperforming his season, and Carolina's biggest structural edge is the one nobody is talking about: the better goalie.

Where we sit versus the market

Sportsbooks have opened Carolina around −155 to −160 to win the Cup (roughly 58% implied), with Vegas near +130. Interestingly, the books and our model agree almost perfectly on the shape of the series — every series-length price lands within a point of our distribution. The only real disagreement is on the winner: the market is a touch more sold on Carolina (~58%) than we are (53%).

So if there is a contrarian lean in our numbers, it is a modest one toward Vegas on the series moneyline — "respectful skepticism of the favorite," not a fire-alarm edge. As always, the value (if any) lives on the series price, not on longshot exact-score tickets. Shop your number.

Bottom line

Carolina is the better team and a deserved slight favorite, and the model's single most likely outcome is Carolina in seven. But this is close to a coin flip, and the swing factors are not the obvious ones. Carolina's question is whether its finishing finally rises to meet its dominance. Vegas's question is whether its chance-suppression defense — and a goalie who has been below expectation, not above it — can hold the line for two more weeks. If Carolina's pucks start going in, the Canes lift the Cup comfortably. If Vegas keeps starving the slot and Hart finally plays to his record, the Knights have every tool to steal it.

Our pick: Carolina in 7 — with Vegas live for the upset.

Full schedule

*if necessary

← Back to Articles