New York won 4-0, but Buffalo controlled the game. The Sabres owned 55% of expected goals yet lost without scoring—a heist powered by goaltending.

I. Shesterkin (NYR) stopped 2.92 goals above expected, posting a 1.000 SV% against all 37 shots. A. Lyon (BUF) allowed 0.62 more than expected, posting a 0.906 SV%. That's a three-goal difference in goaltending.

Buffalo's problem wasn't chance creation. They built five rebound opportunities and held a 53% share of shot attempts. The problem was execution. Against high-danger looks, Buffalo managed one to New York's two. The Rangers scored twice on just twenty rush attempts, overperforming their xG by 1.62. One team converted; one couldn't.

Shesterkin made sure of it.

The projection model cannot separate these teams. Buffalo sits at 53% win probability with an expected score of 2.4–2.6 goals per side—about as close to a pick'em as you'll see.

The Sabres hold multiple small edges. They are 3–2–1 over their last 10 games while the Rangers sit underwater at -7 goal differential in the same stretch. At 5v5, Buffalo's underlying metrics are marginally sharper: 50.0% expected goals share against New York's 48.0%, with both teams at 49.0% Corsi. The Sabres' PDO of 1.020 runs hotter than New York's 1.009, which suggests some regression is baked in for Buffalo.

Rest and goaltending amplify the tilt. Buffalo arrives on six days off; New York skates on one day. More pointedly, I. Shesterkin faces a back-to-back; expect either a backup in net or a tired starter.

The forward matchups are evenly matched. Adam Sykora drives New York's attack: 0.91 xG/60, 3 goals, 1 assist, with an exceptional 20.0% shooting rate. Gabe Perreault adds 0.79 xG/60, 8 goals, 11 assists on 13.1% shooting. Buffalo counters with Jason Zucker's 1.07 xG/60, 14 goals, 14 assists, and 6 high-danger chances, supported by Josh Doan's 1.01 xG/60, 15 goals, 18 assists, and 2 high-danger goals.

Both teams show a structural oddity: neither has scored a single high-danger or rebound goal, a pattern suggesting either a data anomaly or a genuine vulnerability. New York's defensive liability concentrates on Vincent Iorio, while Buffalo's weak link is Jack Quinn. If this game tightens, those matchups will decide it.

The models have it right: this is a coin flip.