I'm going to be honest with you, I think there is a disconnect between the Leafs xGA and their actual defensive ability. They don't give up many chances because they have the puck alot, and generally keep teams to the outside. However, when they do give up chances they are back-breakers and that is sometimes missed when you measure xG without any additional context. Sometimes a team only gives up a few high quality chances but those high quality chances end up in the back of the net, so how good are they really?
I track data and analytics in the OHL so I understand that a large sample set helps to reduce alot of noise, and certainly more sophisticated tracking models that incorporate pre-shot movement, rebounds and screens can further help to differentiate the good from the bad. However, in my experience there is still noise because things like box outs and tying up sticks in front help make a difference. For example, if two players get a shot off of a rebound in a HD area but one of them is all alone and the other has to fight through sticks and bodies they are tracked the exact same. In the public sphere there is this tendency to ignore these variations because they stabilize with increased data, but the problem is that the playoffs are a small sample and if there is a blind spot in a model it doesn't disappear with more data, the problem just multiplies at a rate that is smaller than the perceived norms.
All of this is a long winded way of saying that I see weaknesses in the Leafs defense that I don't think the stats are adequately capturing. There is no question that the Leafs goaltending compounds the issue, perhaps more than I am even accounting for, but I do not think the two are isolated.
And to be completely fair, I do not think I can say with certainty that the Panthers will be successful in the playoffs either because we really haven't seen any team quite like them, at least to my memory. They give up a ton of high quality against, but they also out chance and out shoot everyone.
