The Sharks looked like the better team—59% of shot attempts, near-parity in expected goals (2.82–3.00)—and then lost 5-0. San Jose generated dangerous chances and crashed the net; Buffalo got lucky and left chances for the opposing goalie to steal. The difference: goaltending.

A. Nedeljkovic was catastrophic, stopping 11 of 16 shots (0.688 SV%), 2.0 goals worse than expected. A. Lyon's answer: 23 of 23 (1.000 SV%), 2.82 above expected.

The Sharks' high-danger chances (7 total) never materialized. Buffalo's did: 2 goals from spots that should not have beaten the goalie, plus 5 goals on 21 rush attempts—lethal efficiency on the transition.

R. Dahlin recorded 1 goal and 1 assist in 19:15. L. Stanley added 2 assists in 18:06. S. Carrick scored twice in 13:27, with his deflection from the doorstep the game's fulcrum.

San Jose created 4 rebound opportunities but couldn't beat Lyon. The Sharks generated the volume and quality; the goaltenders decided the margin.

The model projects a 2.5-2.6 goal game with the Sharks at 50% win probability—you don't get much closer to a pick'em than this.

Buffalo holds a slim structural edge. The Sabres own 50.0% xG and 49.0% Corsi at 5v5, suggesting they're building chances and controlling play slightly better than San Jose, which posts 48.0% and 47.0% respectively. These gaps are razor-thin, the kind of margins that usually don't move the needle.

But PDO flips the script. Buffalo's 1.020 means the Sabres are running hot—converting at an elevated rate and catching breaks on defense that their underlying numbers don't justify. San Jose's 0.998 sits at expectation. Regression, for Buffalo, is not a possibility but an inevitability.

Jason Zucker is the Sabres' primary finisher. He generates 1.07 xG per 60 and has scored 14 goals on 15.1% shooting, with 6 high-danger tallies. Josh Doan supplies secondary punch with 1.01 xG/60, 15 goals, and 18 assists, his 11.5% shooting rate far more sustainable than Zucker's heater. Together they're beating their expected output—a warning sign for a hot streak approaching reality.

San Jose counters with Pavol Regenda, the Sharks' most dangerous volume generator at 1.10 xG per 60. He's potted just 5 goals, his 16.1% shooting suggesting room for the puck to cooperate more often going forward. Igor Chernyshov adds 6 goals and 8 assists on 0.91 xG/60, his 15.0% shooting indicating he's slightly outperforming the quality of his looks.

Neither team produces high-danger or rebound goals at meaningful scale—evidence of suffocating defensive structures and scarce second chances. The Sabres' weak link is Jack Quinn. Chernyshov carries that defensive burden for San Jose. Neither gives ground easily, which compounds the low-scoring nature the models project.

This is margin-of-error hockey. Buffalo has volume and fortune; San Jose has cleaner process. The model splits the difference at 50-50 because, at this level of competition, it's the only honest answer. Expect a tight, grinding affair decided by goaltending and whichever team's hot streak holds for one more night.