Eight Games. New Coach. No Time to Practice.

2026-03-30 • NHLForecasts

Vegas fired Bruce Cassidy on Sunday. With eight games left in the regular season. The Golden Knights are 32-26-16, sitting third in the Pacific, holding a 4-point cushion over LA with a playoff spot that isn't guaranteed.

This is either the most desperate move in franchise history, or it's exactly the kind of cold-blooded calculation that has defined this organization since the day it existed.

Either way, John Tortorella is now running the bench in Vegas.

What Broke

Cassidy's body of work with VGK is genuinely impressive — 178-99-43 in the regular season, a Stanley Cup in year one, three straight playoff appearances. But somewhere in 2025-26, something came undone. His final stretch: 3-5-2 over his last 10 games. Since January 19th, the team went 8-15-4.

The numbers from our model tell the same story. In VGK's last 10 games heading into the firing, they were winning just 30% of them with a goal differential of -0.7 per game. That's not a slump — that's a team that has stopped competing.

GM Kelly McCrimmon put it plainly: "I think somewhere along the way we lost our spirit and lost our energy as a team."

That's coach-speak for "we lost the room." And at that point, you only have one option.

The Tortorella Gamble

John Tortorella is 9th all-time in NHL coaching wins (770). He's won a Stanley Cup (Tampa, 2004), claimed two Jack Adams Awards, and built a reputation as one of the most demanding and tactically disciplined coaches in the game. He's also coming off three rough seasons in Philadelphia where, by the end, his own assistant replaced him.

So this isn't without risk. But McCrimmon's framing of why Tortorella made sense is worth taking seriously: "When we're at our best, we play hard, we play fast, we get on teams early. I think that lines up with John's philosophy as a coach."

That's not spin — it's actually analytically accurate. VGK's underlying metrics this season are legitimately strong. They're 4th in the NHL in high-danger goals and generating chances at an elite rate. Our model has consistently rated them as a good team: when playing at home, Vegas has been predicted to win at a 58.3% clip this year — and they've actually won 58.1% of the time. Near-perfect calibration. The team's ceiling has never been the problem.

The floor is the problem. The worst 5-on-5 save percentage in the league (88.45%) has been bleeding them out, and a losing mentality had set in. That's what Tortorella is supposed to fix: defensive structure, accountability, no more passive bleeding that turns close games into losses.

What the Data Shows

Here's where it gets analytically interesting. VGK has been an oddly bi-modal team this season.

In games where they generate momentum and take over — what we track as Flood games, when a team strings together a run of consecutive unanswered goals — they've won every single one: 21 Flood games, 21 wins. When Vegas is the dominant team in a game, they close it out.

But in the other 29 games, they've gone 11-18. They cannot win close, grinding games. When the pace slows, when it becomes a defensive battle, when they need goaltending to steal one — they lose.

Tortorella builds teams that win exactly those games. His trademark is defensive structure, shot suppression, and a relentless forecheck that prevents opponents from settling into a comfortable rhythm. That's the missing piece.

Whether he can install any of that in 8 games is a different question.

The Schedule Doesn't Care

The remaining schedule is not kind.

Tonight they host Vancouver — our model gives them a 65.6% win probability, a good spot to open the Torts era. Wednesday they're home to Calgary (62.3%). Fine. After that, four straight road games:

Three of those four opponents are legitimate playoff teams. Vegas is an underdog in all four. Averaged across the full remaining eight games, their expected win probability sits at 52.7% — barely better than a coin flip per game.

The model puts VGK's playoff odds at 70%. That's solid but not comfortable. Four points over LA with 8 games left and a brutal road stretch coming means there's real margin for this to slip.

The two-game homestand to close — versus Winnipeg (56.7%) and Seattle (65.6%) — could matter enormously. Getting to those games in a reasonable position means surviving the road trip first.

The Big Question

Mid-season coaching changes produce about as many disasters as miracles. What they almost never produce is a complete system overhaul in eight games. The players don't have time to learn new reads, new defensive positioning, new forecheck schemes. Tortorella knows this.

What a coaching change can do immediately is reset accountability. Players who had mentally checked out re-engage. The pace of practice changes. The atmosphere shifts. Whether that translates to wins is genuinely uncertain.

What makes this Vegas move more defensible than it looks at first glance is that the underlying talent is real. Mitch Marner has 71 points. Rasmus Andersson gives them a legitimate top-pair defenseman. Adin Hill has actually been playing well since early March — his .924 SV% in that stretch ranked 4th in the NHL by goals saved above average. The pieces are there.

The question Tortorella has to answer isn't whether Vegas can play better hockey. It's whether he can make them want to in time for the games that actually count.

Eight games. Four of them on the road against playoff teams. Playoff odds at 70% and edging lower.

It's the most compelling VGK analytics story of the year, and it landed in our laps 48 hours ago.

Win probabilities and team metrics from our prediction model. Flood momentum data tracks multi-goal runs within individual games across the 2025-26 season. Coaching history and team stats via ESPN and NHL.com.

← Back to Articles