Quote:
Originally Posted by DeluxeMoustache
They have data for all players, plus puck
So they can start to actually measure things like traffic and see how significant they are
(so if you create a pyramid or ‘shot cone’ by drawing a line from puck position to the corners of the net, and count people in that area, it adds context with regards to the challenge for the goalie, which current models lack)
Thing is, they should be doing next level data analysis with machine learning - the conclusions they will be able to draw will likely be based on having the vision to articulate the problem plus the computing ability
|
They have a tracker in the puck, and yet still rely on low res low frame rate replays from bad angles to decide if a puck was in the net?
I think they have their priorities messed up here.