The Mammoth dominated the underlying numbers and lost anyway. They generated 4.13 expected goals to the Islanders' 3.65, held 58% of shot attempts, and created seven high-danger chances to New York's four. Final score: Islanders 3, Mammoth 2, overtime.

Utah underperformed their xG by 2.13 goals—the chasm between what their shooting created and what they actually scored. D. Rittich posted 28 saves on 30 shots (0.933 SV%), stopping 2.13 goals above expected and stealing a game Utah's dominance should have won. K. Vejmelka's 19 saves on 21 shots (0.905 SV%) held the line, but it wasn't enough.

Matthew Schaefer's slap shot from the high slot at 02:06 of overtime decided it. New York's advantage came down to efficiency. Both teams converted twice on the rush—Utah on 30 chances, New York on 25—but the Islanders crashed the net harder, generating four rebound opportunities and capitalizing on one. Utah's seven high-danger chances nearly doubled New York's four; Rittich simply didn't allow them to matter.

Utah holds a 56% win probability over New York, with the model projecting a 2.4-2.8 game. Two edges explain the lean: territorial control and rest.

The Mammoth dominate 5v5 play. They generate 52.0% expected goals to New York's 48.0%, and control shot attempts at 53.0% Corsi to 49.0%. These margins suggest Utah dictates play.

The rest disparity amplifies that advantage. Utah arrives on one day of rest; the Islanders on three. Fresh legs mean faster decisions and sharper conditioning in the third period.

One watch: K. Vejmelka plays on a back-to-back for Utah, which could mean a backup start or tired starter.

For New York, Anders Lee (1.20 xG/60, 9G 20A, 6.4% shooting) and Bo Horvat (0.87 xG/60, 18G 10A, 12.8% shooting) anchor the offense. Lee generates at an elite rate but his 6.4% shooting suggests finish variance. Horvat's 12.8% rate indicates sustainable conversion.

Utah counters with Dylan Guenther (1.02 xG/60, 28G 15A, 16.8% shooting) and Daniil But (1.01 xG/60, 2G 2A, 5.6% shooting). Guenther's 16.8% shooting compounds Utah's shot volume advantage. Daniil But, despite 2 goals and 2 assists, produces at 1.01 xG/60—generating chances awaiting conversion.

Both teams sit at 0% high-danger goal rate and 0% rebound goal rate. Unconverted opportunities remain vulnerable. Whoever converts marginal plays first gains leverage.

New York's defensive weak link is Alexander Romanov; Utah should attack him early. Utah's Maveric Lamoureux is where New York should probe. Utah's path: control possession, exploit rest, and let Guenther finish. New York must pressure Lamoureux and cycle deep in Utah's zone—but they're behind in volume and recovery time. That's why the model calls for the Mammoth.

Data: MoneyPuck 5v5 models, NHL API