A month into the 2025–26 season, some home teams are making our prediction model look downright foolish. We're tracking the gap between what our forecasts expected and what's actually happening on the ice—and the early returns reveal which clubs are defying expectations, for better or worse.
The numbers below compare our model's projected home win probability against actual results since October 1st. We're only including teams with at least three home games to avoid pure noise, though even with that filter, small sample sizes mean these gaps will narrow as the season progresses.
The Model Was Too Optimistic
Some home teams are running ice-cold despite what the numbers suggested they'd deliver:
New York Rangers — Our model pegged them at 64.3% to win at home across six games. Their actual record? 0-for-6. That's a brutal 64.3 percentage point miss, and the kind of start that has MSG faithful wondering what went wrong.
Calgary Flames — Similar story in the Saddledome. Six home games, 52.2% expected win rate, zero actual wins. The Flames are proving that preseason optimism doesn't always translate to October and November results.
Columbus Blue Jackets — The model loved Columbus at home (76.5% expected), but the reality has been harsh: just one win in four tries (25.0%). Even accounting for a small sample, that's a jarring 51.5 point gap.
San Jose Sharks — Look, nobody expected the Sharks to dominate, but going winless at home through five games when you're projected at 35.9% still stings. Every game matters when you're rebuilding.
Los Angeles Kings — Four home games, one win, despite being slight favorites on average (53.8% expected). The Kings are underperforming at home by nearly 29 points.
The Model Was Too Pessimistic
On the flip side, some squads are proving doubters wrong and turning their home buildings into fortresses:
New Jersey Devils — Perfect 5-for-5 at home after being projected at 63.5%. The Devils are running 36.5 points better than expected, which suggests either they've found another gear or the Prudential Center crowd is worth more than the model credits.
Utah Hockey Club — In their inaugural season, Utah is an unblemished 6-for-6 at home despite projections of 65.6%. Call it new team energy, call it the Delta Center magic—whatever it is, it's working 34.4 points better than forecast.
Seattle Kraken — A perfect 4-for-4 at home when the model gave them a 72.5% chance suggests the Climate Pledge Arena advantage is real. The Kraken are outperforming by 27.5 points.
Florida Panthers — The defending champions are flexing at home with a 5-1 record (83.3%) against a 59.7% expectation. Winners win, and Florida's 23.6 point surplus says they haven't cooled off.
Detroit Red Wings — The Red Wings have rattled off six home wins in seven tries (85.7%) versus a 65.1% projection. That 20.6 point edge suggests Detroit might be better than many anticipated.
The Big Picture
These early-season outliers are fascinating but fleeting. Hockey's randomness means that a couple of bounces, a hot goalie, or a weird schedule quirk can create massive swings in small samples. As we get deeper into the season—20, 30, 40 home games per team—these gaps should shrink dramatically.
The real question is which of these trends represent genuine signal (the Devils really are better at home, the Rangers really are struggling) versus which are just statistical noise that will even out by February. That's what makes October hockey so compelling: we're watching the data tell us a story, but we don't yet know which chapters are real and which are just plot twists that won't matter by the finale.