Once Again The Hockey Analytics Community Is Proving That Their Own Data Sucks

This is the thing that really pisses me off about the people who swear by their analytics...it's arrogance. They plug in numbers and whatever their formula spits out they believe as incontrovertable evidence of a player's ability and production. It's unearned wisdom. Rather than taking the data, seeing how it says that Jonathan Toews isn't good defensively anymore and questioning the data and process itself, this pud goes and says that Toews's defensive game has "collapsed". That is preposterous. It's clear that this guy doesn't watch the Blackhawks or any team every night. You can't sit there and say that Toews isn't the two-way player he has always been. He's still very driven at both ends. The problem is that previously he'd spend most of his time playing with Marian Hossa on his right wing, and he'd have Keith, Seabrook, Hjalmarsson, and Oduya as his defensemen. That is simply not the case anymore thanks to Stan Bowman. I wouldn't be surprised if Stan hires this guy to be honest. 

Hockey is SO hard to track and analyze. I actually applaud the people who come up with data points and track players and try to make sense of this game that is so heavily dependant on teammates, opponents, situations, and circumstances. I love data and analytics and looking for patterns in them. I believe it to be extremely valuable. Which is why most of the data used by teams in the NHL is proprietary and the data available to the public for mass consumption is shit. Which is what this particular tweet is...it's shit. Toews did have a down swing that ran from like 2016-2018. The last two years he's been his old self. Driving the play. Generating chances, creating turnovers, and being a true #1C capable of leading a good team where they want to go. The problem is the supporting cast. The ever evolving rotation of fringe NHL players, rookies, and aging veterans. If you put 31 year-old Jonathan Toews on the 2010 or 2013 Blackhawks you'd barely notice he was a different guy. The play would look the same. He might look a little wiser around the eyes, and he might be a tick or so slower than he was back then, but he's still an unbelievably effective player and if your data suggests otherwise then you should throw your data in the trash and start over.