The underlying numbers said this game should be close. Expected goals split 3.70–3.50 in favor of San Jose; high-danger chances went 6–6. The Sharks won 6–3, which meant they outperformed their xG by 2.30 goals while Utah underperformed — a gap that reflects conversion, not variance alone.

San Jose's advantage was efficiency where it mattered most. Utah had more shot attempts overall (34 on goal to 30), but the Sharks were more efficient off the rush: 3 goals on 26 chances versus Utah's 3 goals on 29. Both teams crashed the net for rebounds with equal intensity. San Jose created 5 rebound opportunities and converted zero; Utah created 4 and converted zero. The offensive pattern was identical. The scorer was not.

Y. Askarov stopped 31 of 34 shots (0.912 save percentage)—exactly the standard required to win. K. Vejmelka allowed 1.89 more goals than expected in 9 shots, which explains why Utah fell short despite the xG closeness.

The turning point: Lawson Crouse's wrist shot from the high slot at 6:30 in the third period.

Depth scoring decided the margin. T. Toffoli posted 2 goals and 2 assists in 15:17, W. Smith had 2 goals and an assist in 19:03, and M. Celebrini added 3 assists in 18:14.

Utah is the better team and the model confirms it — barely. The projection puts Utah at 2.6 goals to San Jose's 2.5, a 52% Mammoth win probability that qualifies as a true coin flip. Scratch any confidence about direction; both teams have equal claim to the result.

Utah's advantage shows in the underlying play. They control 52.0% of expected goals at 5v5 and 53.0% of Corsis, which suggests territorial dominance — San Jose gets hemmed in more often than not. That territorial tilt matters when you need a result, but San Jose's performance over their last 10 games tells a different story: they're minus-7 in goal differential, which is underwater enough to keep this close.

Both goaltenders are on back-to-backs, so expect either backup deployments or starters running on fumes. Monitor whether A. Nedeljkovic and K. Vejmelka play or get swapped out.

Utah's threat comes primarily from Dylan Guenther, who carries 1.02 expected goals per 60 minutes and has compiled 28 goals and 15 assists with a 16.8% shooting rate. His 3 high-danger goals suggest he finds the best spots. Daniil But projects at 1.01 xG/60 but has only 2 goals on 2 assists — his 5.6% shooting rate is a red flag for variance catching up.

San Jose's Pavol Regenda poses the sharper threat: 1.10 xG/60 and 5 goals with a 16.1% shooting rate tells you he's both getting prime opportunities and finishing them. He's marked likely inactive for this matchup, which changes the calculus entirely. Without him, San Jose loses their cleanest offensive engine. Igor Chernyshov, at 0.91 xG/60 with 6 goals and 8 assists on a 15.0% rate, becomes the remaining danger.

The model isn't wrong to call it even. Utah should win the shot volume and the expected-goals count by small margins, but San Jose's recent collapse wasn't severe enough to break them — just enough to make them beatable. With Regenda potentially unavailable, Utah's structural edge tips the scales, but not enough to clear a spread.