Defensive Statistics
- Thread starter TheKingSlayer
- Start date
TheKingSlayer
He was in the way!
Pominville Knows
Registered User
struckbyaparkedcar
Guilty of Being Right
I see a lot of structural problems with this thought experiment:
1) the fact that defensive usage statistics really only matter within teams (the player with a team-high QoC of .75 is more valuable than a player with a 5th best QoC of 1), so they can't be factored into the equation
2) the fact that we have no measure of "expected defensive output" or "defensive usage" or the different microactions required by different systems (being the Kings checking center might mean more takeaways, but fewer blocks, and such).
3) the fact that "defensive action" stats such as hits, blocks, and takeaways are also measure of not having the puck and therefore not particularly valuable, so they're out as well.
4) possession metrics are role-reliant and harm players who are asked to buoy overmatched units for extended periods while not getting scored on (see: Shea Weber's fancystats), even though we're trying to seek out these high defensive usage / high defensive output players with an all-encompassing stat like this.
1) the fact that defensive usage statistics really only matter within teams (the player with a team-high QoC of .75 is more valuable than a player with a 5th best QoC of 1), so they can't be factored into the equation
2) the fact that we have no measure of "expected defensive output" or "defensive usage" or the different microactions required by different systems (being the Kings checking center might mean more takeaways, but fewer blocks, and such).
3) the fact that "defensive action" stats such as hits, blocks, and takeaways are also measure of not having the puck and therefore not particularly valuable, so they're out as well.
4) possession metrics are role-reliant and harm players who are asked to buoy overmatched units for extended periods while not getting scored on (see: Shea Weber's fancystats), even though we're trying to seek out these high defensive usage / high defensive output players with an all-encompassing stat like this.
struckbyaparkedcar
Guilty of Being Right
For whatever its worth, my method has been to seek out high defensive usage players using zone starts and the different QoC metrics, and then compare their possession and goals against stats.
And usage categories and the differences between systems are a really big deal to me. It's silly how people have compared guys like Pietrangelo, Doughty and Weber to players like Subban and Karlsson using like stats for the past few years. Those players just don't get asked to do the same things.
And usage categories and the differences between systems are a really big deal to me. It's silly how people have compared guys like Pietrangelo, Doughty and Weber to players like Subban and Karlsson using like stats for the past few years. Those players just don't get asked to do the same things.
