Montreal Canadiens 3, Tampa Bay Lightning 2 - The scoreboard showed a one-goal game, but the story told two different versions of what happened at Amalie Arena on April 29. Tampa Bay controlled territory, generated chances, and played the kind of possession hockey that won them two Stanley Cups. Montreal finished their opportunities, trusted Jakub Dobes to save the rest, and walked out with a 3-2 victory that puts them one win away from eliminating a team that finished with the same point total in the Atlantic Division.
The Canadiens lead the first-round series 3-2 heading to Montreal for Game 6, and the path that brought them here is the same one that's carried them all season: win games you shouldn't by finishing better than the opponent possesses.
Tampa Bay fired 27 shots on goal. Per MoneyPuck, they controlled 54.1% of expected goals in all situations and 54.0% at five-on-five - numbers that typically correlate with winning. Nikita Kucherov attempted five shots in 18:43. Darren Raddysh added five more from the point in 21:26. Brandon Hagel generated quality looks in over 20 minutes. The Lightning did everything their model requires except score more goals than Montreal.
The Canadiens needed 24 shots to score three times. Brendan Gallagher buried one at 3:00 of the first period in just 6:48 of ice time after sitting out four straight games. Kirby Dach made it 2-1 at 7:00 of the second. Alexandre Texier scored the winner at 1:06 of the third period on a wrist shot from the left circle that deflected off Andrei Vasilevskiy's glove, the product of Lane Hutson's stretch pass that revealed Tampa's defensive positioning.
Then Dobes stopped everything else. The rookie made 25 saves on 27 shots, posting a .926 save percentage that doesn't capture the sequence that mattered most: the final minute, when Tampa threw everything at him and he turned away each attempt to preserve Montreal's lead. As Phil Mishkin of the Tampa Bay Lightning site wrote, this was "Montreal's best game in the series," with the Canadiens playing "quicker, more determined" hockey and executing at a "higher level for most of the game" despite the close score. (Phil Mishkin / NHL.com)
Tampa Bay has now followed the same pattern through five games: they win the even-numbered contests (2, 4), Montreal takes the odd ones (1, 3, 5). The difference isn't luck or variance anymore. It's philosophy.
The Numbers That Don't Matter Enough
The Lightning's season-long metrics tell the story of a team that should win more than it does. They posted 54.0% of expected goals at five-on-five all season, per MoneyPuck, generating 284.4 expected goals in all situations while allowing 241.5 - a plus-42.9 differential that ranks among the league's best. They finished 50-26-6 with 106 points, which sounds excellent until you realize Montreal earned the exact same point total (48-24-10, 106 points) while posting a minus-5.1 expected goals differential.
Tampa Bay out-possessed Montreal again on April 29. They controlled more of the ice, created more volume, and limited Montreal's high-danger opportunities. Vasilevskiy stopped 21 of 24 shots for an .875 save percentage - below his standard but not catastrophic for a goaltender who has spent a decade as the last line of defense when possession breaks down.
But possession broke down three times, and each time Montreal finished. One goal came off a scramble play that rewarded Montreal's willingness to crash the net rather than cycle for a better look. The goal at 7:00 gave the Canadiens a lead they wouldn't surrender. The winner early in the third came from transition - the kind of quick-strike hockey that doesn't show up in expected goals models until the puck is already in the net.
Jake Guentzel and Dillon James answered for Tampa Bay, but the Lightning never led. They chased the game without appearing desperate, controlled play without dominating, and left the ice with another loss that the underlying numbers say shouldn't have happened.
What Tampa Built and Why It's Failing Now
The Lightning constructed their roster around a specific theory: out-possess opponents long enough and talent will break through. It's why they prioritize puck-moving defensemen like Raddysh, who logged 21:26 while contributing five shots from the point. It's why they run Kucherov with Guentzel, trusting that skill plus volume will eventually overwhelm goaltenders. It's why they invested in Vasilevskiy as the anchor who saves the chances that leak through.
That model delivered championships in 2020 and 2021. It carried them to 106 points this season despite finishing with the same point total as a Montreal team that posted worse underlying numbers. But the model assumes that quality and quantity will eventually win out, and Montreal has spent 82 regular-season games plus five playoff contests proving that assumption wrong.
The Canadiens don't need to match Tampa's possession metrics. They posted 49.5% expected goals in all situations during the regular season, per MoneyPuck, allowing 273.4 expected goals while generating 268.3. They are, by process, a below-average possession team. But they finished with 106 points because they do two things better than Tampa: finish their chances and get goaltending when it matters.
Dobes' performance in the final minute exemplifies the difference. Tampa pushed, created traffic, forced scrambles. The rookie made the saves. Montreal doesn't need to dominate possession when they can score on 12.5% of their shots (three goals on 24 attempts) and trust Dobes to handle the rest. That's not sustainable over a full season according to analytics theory, but Montreal has made it work for 87 games now.
The Personnel Problem Nobody Predicted
Victor Hedman remains out on personal leave, robbing Tampa Bay of their most important defenseman and the player who anchors their possession game. Without Hedman, the Lightning lean harder on Raddysh and Ryan McDonagh, who logged heavy minutes but can't replicate Hedman's ability to kill opposing cycles and transition cleanly out of the defensive zone.
The absence matters more in tight games than blowouts. Montreal's third-period goal came from transition - Hutson's stretch pass finding a Montreal player in space, exploiting the kind of gap coverage that Hedman typically erases. Tampa's defensive structure held for most of the game, but one mistake was enough.
Montreal, meanwhile, welcomed Gallagher back into the lineup after sitting the first four games earlier in the series. Nick Suzuki added an assist in 18:15, Hutson chipped in the helper that set up the winner in 22:44, and Kaiden Guhle logged 20:45 from the blue line. Montreal's depth is winning minutes they don't dominate.
What Happens When Process Stops Producing Results
Tampa Bay faces a choice that has no good answer: trust the model that delivered championships or abandon the philosophy that's failing in real time.
Doubling down means accepting more games like this, where the Lightning control play and lose anyway. The coaching staff can point to expected goals share and shot attempts and territorial control, all of which suggest Tampa should be winning. But the score is the only number that advances or eliminates teams, and Tampa is running out of games to prove their process will eventually produce the right results.
Abandoning the model means discarding the identity that made them champions. It means questioning whether possession-based hockey still works when a well-structured opponent can finish their limited chances and get goaltending that erases territorial advantages. It means rebuilding a philosophy mid-series against an opponent who has already figured out how to beat them.
Neither option is appealing. Neither option guarantees success. And with Montreal heading home for Game 6 holding a 3-2 series lead, Tampa Bay is running out of time to decide which gamble offers the better chance of survival.
The Lightning still have the talent to win two straight games and advance. Kucherov, Guentzel, Hedman (if he returns), and Vasilevskiy have done it before. But they need to figure out why their formula stopped working - why out-possessing opponents no longer translates to victories, why generating quality chances no longer guarantees goals, why doing everything the analytics community says correlates with winning produces losses instead.
Montreal answered those questions months ago. They don't care about expected goals share or possession metrics or territorial control. They care about finishing their chances, trusting Dobes to make saves when it matters, and walking out with two points. They've done it 48 times in the regular season and three times in five playoff games. Tampa Bay has two games left to prove their way works better, or Montreal will prove it doesn't matter at all.