Zeke exists to take the fall for the imperfections of dubas
he’s basically jesus
I'll be sad when/if he finally figures out that expected goals is bullshit. but then he'll probably pretend he was saying that from the beginning.
I remember you saying much the same about Burke.dubas has been almost perfect, though, aside from goaltending decisions.
I'll be sad when/if he finally figures out that expected goals is bullshit. but then he'll probably pretend he was saying that from the beginning.
I remember you saying much the same about Burke.
It's crazy how often I spot mistakes on Naturalstatrick, and that's from barely even paying attention. Hell, just a few weeks ago it was widely reported that Engvall missed an entire period because of the data on that site. Which was wrong. If they can't even figure out who is on the ice, just think about the amount of errors there are with shot attempts and the other shit. I do think it provides a nice snapshot over a whole season as the sample size increases, but game by game it's hard for me to trust the data much. It's understandably nowhere near perfect.Publicly available stats aren’t great because of the amount of work required to validate / ensure quality.
I think most just pull from the event lots on nhl.com. One site (forget which) admitted they watch some games and then extrapolate.
I know somebody in a prominent position with an nhl teamwho got his start doing analytics work and the amount of time required to get good data is crazy.
I remember you saying much the same about Burke.
I can't wait till they draft Schneider. He's going to get a #alwaysdraftBradens poster in his bedroom and compare him to Slavin. And then I'll agree and he'll think I'm heckling him and get defensive.
I recall many expositions on why his teams were just great outside of the goaltending.
It's pretty well established that NHL data has serious measurement error. I presume this is why there's a strong push for tracking.
The problem with maligning all the data for this is: (a) not overly serious in the bigger scheme, and; (b) can be used to insist eye test > ...
Like, I'm suspicious of Mucho's xgf% rating in the BJs series but it's probably right.