Modifié 10 sept. 2021 à 12 h 50
Just felt like throwing my 2 cents in:
Analytics are a VERY USEFUL TOOL.
But they are flawed.
The “eye test” is a VERY USEFUL TOOL.
But it is flawed.
You can’t use solely analytics to judge prospects for example. It’s a idiotic thing to use on its own for prospects.
You can’t use just the eye test for NHL players for example.
It’s a idiotic thing to use on its own for players.
You must use both imo. But you also have to judge the situation or player which is where the eye test “evolves” imo.
Take it like this. I’m going to use a very controversial take here too. Erik Karlsson is a very hard player to justify using solely analytics.
Karlsson even during some of his best years had very poor analytics with the Sens (Not all of his years, but at least from the charts I recall seeing the other month). A lot of Sharks fans see a ok game and then go look at Karlsson stats (adv.) and go look he’s dog****. But I continually try to point out this flaw. Early in the 2020/21 season Karlsson had a PHENOMENAL game. He looked amazing and everything was clicking and I believe he got a couple points. His advanced stats for that same game WERE BRUTAL. Yeah he got some a couple points and just looked sooooo gooooood…. But his analytics didn’t change and if anything they regressed.
Now he’s the other side of that same coin. Vlasic doesn’t have god awful adv. stats. They aren’t great, they’re pretty poor, but none the less. Yet Vlasic is just terrible to watch. He’s slow, poor decision making, loses his man or position a ton. I can go on. Yet his analytics value him around the same as Karlsson.
Now I’m not saying Karlsson is amazing because he doesn’t always look good and his analytics suggest the same or that Vlasic is worth his estimated “2-3 million dollar value according to analytics” because often I’d rather see Vlasic be benched.
I’m using them as an example. Not all players play a style that looks good on paper (adv stats). And not all players stand out on the eye test. For example Slavin from the discussion is an amazing example. He’s got great stats, but even if they aren’t the best in the entire league, Slavin flourishes on the eye test and then also has good analytics to suggest that he’s not a fluke. Whereas Karlsson often looks great on the eye test for a majority of a game but he makes mistakes and his style of play firstly doesn’t mesh with a bad team, and his adv stats aren’t great because he’s usually trying to do too much. But that doesn’t mean he’s the worst guy in the league.
I think you should use the eye test and pair it with adv stats. And for some players that situation should change. Use context, and I think there are some fantastic analytics that should be considered always. Quality of Competition is a great one when you pair with Quality of Time on Ice. Use them with your 5v5 impact and mesh those with the eye test.
I don’t think analytics are bad, and I don’t think they’re amazing. But you shouldn’t throw one or the other out.