Colorado dominated the expected-goals battle 59–41 but lost 2–1, a margin that understates how thoroughly the Avalanche controlled play. This was a goaltending heist.

The Panthers mustered just 2.47 xG despite winning. Colorado created 3.58—more than enough to expect a comfortable victory—yet D. Tarasov's performance negated that advantage entirely. He saved 2.58 goals above expected, a two-goal swing that determined the outcome. The Avalanche's problem wasn't opportunity creation: they crashed for five rebound chances. The problem was execution, and D. Tarasov was impenetrable.

The scoreline felt inevitable in reverse. Florida won 47% of shot attempts (25–28) and benefited from the goaltending lottery; Colorado's underperformance on xG (2.58 goals below expected) reflects not poor finishing in isolation but an elite save percentage wall. S. Wedgewood stopped 23 of 25 (0.920 SV%), a competent effort that was simply outmatched.

High-danger chances finished even at 2–2, the only parity in this matchup. Everything else belonged to Colorado.

The inflection was A. Ekblad's slap shot from the top of the circles late in the second period (18:12), a moment of clinical finishing that shouldn't have mattered given Colorado's control but did because games are decided by goals, not chances. The Avalanche would dominate the rest of play without consequence. That is the definition of a steal.

D. Tarasov's excellence cannot be overstated. A 0.964 SV% is not merely above average—it was the difference between a convincing Avalanche win and a Panthers escape. B. Nelson's assist (25:15) and A. Ekblad's finish aside, the real story was the goaltending divergence.

Colorado leaves aggrieved and correctly so. The stat sheet indicts Colorado only for failing to convert. The scoreboard indicts Colorado for losing when they should have won.

The model projects a 2.8-2.5 Colorado win with the Avalanche at 55% probability—about as close to a pick'em as you'll find.

Colorado's case rests on process. The Avalanche are 9-0-1 in their last 10 and command the underlying play at 5v5, posting a 57.0% expected-goals share against Florida's 49.0%, plus a 57.0% Corsi advantage at 52.0%. They tilt the field territorially. The risk: Colorado's PDO of 1.021 is running hot. Expect regression; Florida's 0.989 may already reflect that gravity.

The rest column favors the visitors. Colorado arrives with three days between games; the Panthers have one.

Goaltending adds uncertainty. S. Bobrovsky is on a back-to-back—watch for either a backup start or a fatigued starter.

When Colorado takes the ice, Artturi Lehkonen and Ross Colton drive the attack. Lehkonen generates 1.11 expected goals per 60 minutes with 18 goals and 22 assists on the season, including six high-danger goals at a 14.9% shooting rate. Colton sits at 1.10 xG/60 with 7 goals and 15 assists; his 4.9% shooting rate suggests inefficiency despite quality chances.

Florida will counter with Sam Bennett and Brad Marchand. Bennett has 18 goals and 19 assists at 0.97 xG/60, matching Lehkonen's goal total but trailing in volume. Marchand is the finisher: 0.93 xG/60, 17 goals, 9 assists, and 19.5% shooting—nearly five points above Lehkonen. That clinical efficiency could offset Colorado's territorial edge.

At 55%, the Avalanche hold a genuine edge, but it's narrow enough that rest advantage and Marchand's finishing keep Florida in it.