silverfish
got perma'd
I read up on Manny's xG model every now and then, since I'm just like 15 hours of work away from getting my own, whenever it is I actually sit down and do it, and I think I just came to realize that each shot type is its own model... In the footnotes:
Shot Quality And Expected Goals: Part I | Corsica
So instead of one massive input in the model, he's subsetting by shot type, and then again by rebound and non-rebound, and then forming their own unique regression.
I don't want to do that, that's too much work.
Each of the six shot types6 provided by the NHL forms its own category, and these are further subsetted by rebound and non-rebound. Only unblocked shots are used due to the unfortunate fact blocked shot coordinates are unavailable.7 These twelve shot sub-types are regressed independently according to the remaining variables. The rationale here is that each shot subset should respond to the variables differently. Namely, distance and angle do not influence slap shot quality in the same manner as they do, say, deflections. In addition, the relationship between goal expectancy and distance or angle are not assumed to be linear. That is to say, the model is not bound by the idea that shot quality changes at a constant rate along the scales of distance and angle. I found these modifications improved the model’s ability to assess goal expectancy.8
Shot Quality And Expected Goals: Part I | Corsica
So instead of one massive input in the model, he's subsetting by shot type, and then again by rebound and non-rebound, and then forming their own unique regression.
I don't want to do that, that's too much work.