Quote:
Originally Posted by DeluxeMoustache
Nope.
Improving sv% by grouping ####s in to buckets of shot location and type arguably helps from a probability standpoint
There is nothing measured to reflect where a shot taken is aimed, and dealing with that is pretty much more important than anything
The end
|
Yeah sorry it's not that simple. You're not an expert on this as much as you like to pretend you are.
Same thing with Hanifin in the dallas series. He's been one of the worst d men for Vagas but because he scored a game winner everyone assumes he is playing well. Statistically he's been terrible and it matches my eye test.
If my eye tests tells me someone is playing bad and I look at the underlying stats and he's playing great then I can change my opinion. When someone you think is playing well and all underlying stats point to the opposite then maybe you can re consider.
Expected goals over the long haul is a great tool. Not every shot will have mathews shooting and not every shot will be backlund. It averages out and over a larger sample size it is a great tool.