Nithoniniel
Registered User
The thing is, you judge metrics by their effectiveness as a predictive or descriptive tool. So while I understand your argument, it's clear that the best models actually do a fair job of accounting for all these 'essential variables' or that effectiveness would be worse. Granted, a lot of us who like to look at newer metrics tend to overstate just how strong that effectiveness is to begin with.The eye test is much better than any advanced statistic. I have nothing against statistics, but the problem is, unlike baseball, the physics of hockey is much too complex to simplify using even "advanced" statistical models that are not capable of including all of the essential variables necessary to accurately compare players, and the eye test tells you everything you really need to know about players, so my advice is to put away the stat sheet, and just watch the game.
Furthermore, while any metric is flawed and doesn't portray reality with complete accuracy, the same is true for domain knowledge, that is the eye test. We are heavily biased in what we see and how we experience it, so at worst you are choosing between two bad options here. Often when that is the case, the best thing to do is to use both of them in tandem, so the flaws can balance each other out.