Nashville won 3-2 despite generating just 42% of expected goals. Washington created 3.86 xG to Nashville's 2.74—a 58% share that normally carries games. Not this one.

The Capitals' collapse came in high-danger areas. Washington had six premium chances and converted zero; Nashville had two. J. Annunen stopped all the dangerous looks, recording a 0.938 save percentage (30 of 32 shots) with 1.86 goals saved above expected—the margin that decided this game.

Nashville's power play supplied the edge. The Predators scored twice on 14 power-play shots (1.22 xG), converting the chances Washington conceded on special teams. Washington's even-strength superiority—controlling the puck and the chances—vanished from the scoreboard.

On the rush, Washington stayed competitive (two goals on 23 chances to Nashville's three on 26). But in the moments that mattered, the Capitals had nothing. Ethen Frank's wrist shot from the high slot (09:03 of the third period) proved decisive.

This is about as close to a coin flip as advanced models can project. The forecasts land on Washington 2.6–2.7 goals with Nashville holding 51% win probability—the narrowest of edges, if one exists at all.

The underlying talent gap is nearly invisible. Washington edges the expected-goals share at 50.0% to Nashville's 49.0%, but the Predators flip the script on shot generation, logging a 50.0% Corsi advantage to the Capitals' 49.0%. One team dominates play quality; the other takes more shots. Both claims are defensible.

The wildcard is luck. Washington's PDO of 1.018 screams positive variance—goals are flowing at an unsustainable rate. Nashville sits at 0.992, suggesting regression upward. If the Capitals' shooting luck dries even slightly, this game tightens further.

Goaltending adds another layer of uncertainty. L. Thompson handles Nashville's workload on a back-to-back, which typically signals either a relief appearance or a starter running on fumes. Either scenario tilts toward more goals.

For Washington, Alex Ovechkin anchors the offense with a 0.97 xG/60 and 19 goals already—at 13.8% shooting, he's nearly at his historical efficiency. Anthony Beauvillier complements him at 0.89 xG/60 with 14 goals, hitting 10.4% on his attempts. Both excel in high-danger areas, with 2 high-danger goals each. Nashville's defense must collapse on them, or these two will find space.

Nashville counters with Filip Forsberg, whose 23 goals and 0.80 xG/60 establish him as the group's most dangerous scorer, shooting at 13.6% with 4 high-danger goals. Zachary L'Heureux presents a different profile: 4 goals on 0.85 xG/60 despite shooting at 16.0%, with only 1 high-danger goal.

The defensive liabilities are known. Washington leans on Declan Chisholm in defensive situations, a weak point. Nashville's Jonathan Marchessault carries similar liability. Both teams will test these matchups aggressively.

This genuinely is a toss-up. The models project it that way for good reason: neither team holds a meaningful edge on talent, and the variance factors—Washington's hot PDO, Nashville's cold one, and Thompson's workload—don't settle the question. That's what makes this game a true coin flip.