I track an NHL prediction model across all 32 teams. The Islanders are dead last in predictability at 38.5%. A coin flip would literally be more accurate.
Unlike Buffalo (GM firing) or Florida (Barkov ACL), there isn't one clean explanation. It's just weird.
The overtime thing is historic. 10-0 in OT, first team in NHL history to do that. That's 20 extra standings points. Last season you went 5-9 in overtime. If each OT is roughly a coin flip, going 10-0 is about a 1 in 1,000 probability. No model predicts that.
Sorokin has 7 shutouts tying the franchise record. But on the nights he doesn't get the shutout, games are tight enough to go to OT. Then you win. Every time.
The model predicted a 54.5% win rate. Actual is 64.3%. The gap is almost entirely overtime magic. The honest answer is the model doesn't get the Islanders and I don't think any model would.
My full breakdown with some charts that wont upload to Reddit is here if you're interested: Full breakdown of all 32 teams: PuckCast
5 comments
Perhaps it’s the mood in the locker room that is the biggest predictor for this team (besides the 3v3 which we field our superstars who perform better with open ice – due to easier puck movement)
We can already guess how the team is going to play just based on location and time of day.
So glad statistics back up what this team puts us through
I don’t know much but I know this. That place was as Zadorov called it “ a library “ last night. Even when the Isles were up 1-0 the only people you heard was the drum section and the random guy screaming for everyone to wake up. This building better change and quick. When the team is in this type of spot the fans can will them to get going. It’s not like we haven’t been there before so wake up people. Or it could just be people aren’t as invested in this team like they used to be and I’m just kidding myself.
A coin flip going 10-0 is equally likely to going 5-5.