Rangers pull away from the Devils

The model sees this as a dead even proposition. New Jersey and New York each have legitimate claims to the edge, and the projection system has split the difference: Rangers at 51% win probability. That's as close to a pick'em as you'll find.

The five-on-five numbers explain why. New Jersey holds a thin edge in volume: 50.0% expected goals and 51.0% Corsi to the Rangers' 48.0% and 49.0%. But volume doesn't determine outcomes; luck does, at least in the short term. The Devils are running cold at 0.971 PDO, a sign they're due for positive regression. The Rangers sit at 1.009, suggesting they're either hot or lucky—and regression cuts both ways.

The offensive edge leans toward New Jersey. Timo Meier leads with 1.08 expected goals per 60 minutes, 13 goals, and 15 assists, including 3 high-danger chances. Jack Hughes follows at 0.88 xG per 60 with 18 goals and 17 assists, though he's posted just 1 high-danger goal. His 11.5% shooting rate suggests that run is sustainable. For New York, Adam Sykora carries 0.91 xG per 60 with only 3 goals and 1 assist but a striking 20.0% shooting rate—a number that will likely regress. Gabe Perreault adds secondary scoring at 0.79 xG per 60, posting 8 goals, 11 assists, and 2 high-danger chances with a more measured 13.1% shooting rate.

Neither team is converting high-danger chances; both sit at 0% for high-danger and rebound scoring. The margin, then, may rest on which team capitalizes first on opportunities—not which team creates them.

Defensively, both clubs carry vulnerabilities. Johnathan Kovacevic projects as New Jersey's weak link; Vincent Iorio is New York's. Whichever back line breaks first could swing the game.

The projection system's indifference makes sense. A 51% Rangers edge is barely better than a coin flip. Both teams generate offense at nearly identical rates, both carry defensive liabilities, and both have regression vectors pointing in different directions. The Devils have the volume edge and upside if PDO normalizes. The Rangers have the recent results and momentum of a hot streak. In a game this evenly matched, opportunism and execution matter more than underlying talent—and those are things the model can't predict.