The Mammoth are the model's lean over the Maple Leafs — with PDO regression in play

The model leans toward the Mammoth, and the case rests on how the play has gone rather than on the scoreboard. The underlying numbers favour the Mammoth, and their luck has been running cold.

The model gives the Mammoth a 55% win probability, which makes this a lean, not a lock. Its projected score is 3.5 to 3.1 in the Mammoth's favour.

The strongest evidence is expected goals at five-on-five, a measure of the quality of chances each team creates. The Mammoth hold 57.5% of them. The Maple Leafs sit at 48.5%, below even.

Shot attempts, meaning every shot try including misses and blocks, tell the same story. The Mammoth take 55.5% of them, against 47.2% for the Maple Leafs. That points to a territorial tilt, with more of the play happening in the Maple Leafs' end.

The regression case is the luck side. PDO, the sum of a team's shooting and save percentages at five-on-five, sits at 0.977 for the Mammoth, a sign of cold luck. The draft expects positive regression, meaning results should improve even if the play does not change. The Maple Leafs are at 0.997.

In goal, K. Vejmelka sits at -5.8 goals saved above expected over 12 games. His save percentage is 87.2%. A. Stolarz, in 6 games, is at +1.1 goals saved above expected, with a save percentage of 91.0%. That gives A. Stolarz the edge, 6.8 goals above his counterpart this season.

Data: NHL API, Backcheck xG model