Tampa Bay opens Round 1 at home as a 61.5% model favorite over Montreal. Its season record and recent results tell different stories.
The Season View
TBL finished 40-26-0 with a +0.83 goal differential per game. Vasilevskiy is healthy. The core that has been to three Cup Finals is mostly intact. That history gives context to a forecast built around recent performance.
The Model View
The model does not see pedigree. It sees the last thirty games.
| 30-game differential (TBL − MTL) | Value | Read |
|---|---|---|
diff_xg_diff_30 |
+0.23 | TBL slight process edge |
diff_shot_diff_30 |
+5.17 | TBL generates more shots |
diff_gd_30 |
−0.60 | MTL outscored TBL by 0.6/game |
diff_win_pct_30 |
−0.13 | MTL won 13pp more often |
diff_hd_diff_30 |
−0.53 | MTL better on high-danger chances |
TBL's process stayed clean into April — they were still creating the right shots at the right rate. But the results stopped landing. Meanwhile Montreal closed the season 7-3-0 in their last ten with a +0.60 goal differential per game. Over the same window, Tampa was sub-.500 with a negative goal differential.
Our v2 win model uses a 30-game window by design. Longer windows stay anchored to October form; shorter windows overreact to single games. Thirty games was the sweet spot in backtesting, and it is exactly the window that captures this kind of late-season divergence. The model is not broken — it is doing what it was built to do.
Which View Is Right?
Tampa's late slump could reflect temporary factors, such as rest management, minor injuries, or an unlucky stretch of finishing. If shot quality is more representative than recent goal differential, the season-long view deserves more weight.
The model's 61.5% forecast also reflects Montreal's strong closing form. Its usefulness depends on whether that improvement persists and whether Tampa's recent scoring problems reflect something more lasting than puck luck.
Shot quality and quantity favor TBL; recent results favor Montreal. The series will offer more evidence about those competing signals, but a handful of games cannot settle how accurate a probability estimate was.
What We Are Watching
Game 1 at home. Watch whether Tampa controls shot quality as well as the score. One win or loss cannot validate a probability forecast, but the way each team creates chances can help explain its performance.
Shot quality in transition. Both teams play structured, neutral-zone-heavy hockey. If Tampa's forecheck is generating the clean looks we saw all season, their top line will produce quickly. If Montreal's closing-stretch defensive structure travels, the series becomes a grinder that trends toward the underdog.
Vasilevskiy's workload. If he sees 30+ shots and posts the kind of game he was posting in March-April, the question about his health answers itself. If the shot count is low and TBL wins on goaltending, that is a less convincing signal than if they win by controlling play.
The Answer, For Now
The 61.5% forecast reflects the model's emphasis on recent performance. If Montreal maintains its closing form, that emphasis may prove useful. If Tampa's April was a temporary dip, a longer view may better describe the matchup.
Both teams have a plausible path through the series. That is the Tampa Bay question.
Get NHL forecasts in your inbox
Occasional emails with NHL forecasts, analysis, and model updates.