A philosophical question I've seen asked but never conclusively answered: should WR value be measured in a player's advantage over his teammates (e.g., Golladay has a +0.24 WOPR advantage over the next-closest receiver) OR should the metric stand for itself as an indicator of isolated value? In other words, should we expect (e.g.) Golladay to be more valuable than, say, Robby Anderson, even though both have similar WTPRR, BECAUSE of the team context? I guess it's kind of a team-centric VBD argument, where value is determined by how much player x outscores his teammates rather than just looking at the number and drawing conclusions. Just wondering if that would make these metrics even more predictive of future success. (That, or I'm salty over watching D.J. Moore get completely ignored last night in favor of a similarly targeted teammate; yeah, probably that!)
This is a great question, and unfortunately I don't have a conclusive answer. The conclusive thing I believe most strongly is there are no independently correct metrics to use without other context. At least not right now in football.
And then because value is so important to our game in terms of overlooked players on waivers, trade candidates, or DFS options, it's all relative to what the market is valuing. I used to be big on the expected points metrics and those types of things and now I think they are arguably overused because so much of the broad takes in the industry are on whether guys are over- or underperforming their roles, when their "roles" are defined as the volume they've seen to date but that is rarely perfectly sticky throughout a season. It's useful to know that stuff, but if everyone is taking past volume as a predictor of future volume and no one is trying to project how offenses might evolve and where future volume might go, suddenly there's an angle there.
My answer has close to nothing to do with what you're specifically asking about whether to consider team-adjusted stats or player-specific ones but is hopefully helpful. I guess applying it to that would be saying you need additional context in both cases.
Thanks, Ben. I'll admit I had to read your post a few times to really grasp that WTPRR is homing in on the attention a player receives based on role and playing time, while WOPR seems to be indicating these things as a share of overall team activity. It's probable that football is simply too complex to point at one thing and say "this is it!" At least not in the same way that baseball sabermetricians can point to things like WAR, OPS, et al. I agree that it's really important to have a forest and trees approach to these matters. If you've got Russell Wilson throwing you the ball, well, the question is less pertinent. The volume you get compared to your teammates' will matter less than if Kyle Allen is your QB; you're going to need those extra passes coming your way to offset the lesser quality of same. So maybe that's a good approach: if playing in a lesser offense with a lesser passer, measuring a player's input ("team-sensitive" WOPR/WTPRR) against that of his teammates is one key to whether or not he'll have a chance to consistently produce.
An absolute clinic and so fun to read. Thank you so much for the deep dive into the origin of some of these stats and logic that goes into building them, as much as the YTD analysis on what they mean in practice.
A philosophical question I've seen asked but never conclusively answered: should WR value be measured in a player's advantage over his teammates (e.g., Golladay has a +0.24 WOPR advantage over the next-closest receiver) OR should the metric stand for itself as an indicator of isolated value? In other words, should we expect (e.g.) Golladay to be more valuable than, say, Robby Anderson, even though both have similar WTPRR, BECAUSE of the team context? I guess it's kind of a team-centric VBD argument, where value is determined by how much player x outscores his teammates rather than just looking at the number and drawing conclusions. Just wondering if that would make these metrics even more predictive of future success. (That, or I'm salty over watching D.J. Moore get completely ignored last night in favor of a similarly targeted teammate; yeah, probably that!)
New subscriber; love the content!
This is a great question, and unfortunately I don't have a conclusive answer. The conclusive thing I believe most strongly is there are no independently correct metrics to use without other context. At least not right now in football.
And then because value is so important to our game in terms of overlooked players on waivers, trade candidates, or DFS options, it's all relative to what the market is valuing. I used to be big on the expected points metrics and those types of things and now I think they are arguably overused because so much of the broad takes in the industry are on whether guys are over- or underperforming their roles, when their "roles" are defined as the volume they've seen to date but that is rarely perfectly sticky throughout a season. It's useful to know that stuff, but if everyone is taking past volume as a predictor of future volume and no one is trying to project how offenses might evolve and where future volume might go, suddenly there's an angle there.
My answer has close to nothing to do with what you're specifically asking about whether to consider team-adjusted stats or player-specific ones but is hopefully helpful. I guess applying it to that would be saying you need additional context in both cases.
Thanks, Ben. I'll admit I had to read your post a few times to really grasp that WTPRR is homing in on the attention a player receives based on role and playing time, while WOPR seems to be indicating these things as a share of overall team activity. It's probable that football is simply too complex to point at one thing and say "this is it!" At least not in the same way that baseball sabermetricians can point to things like WAR, OPS, et al. I agree that it's really important to have a forest and trees approach to these matters. If you've got Russell Wilson throwing you the ball, well, the question is less pertinent. The volume you get compared to your teammates' will matter less than if Kyle Allen is your QB; you're going to need those extra passes coming your way to offset the lesser quality of same. So maybe that's a good approach: if playing in a lesser offense with a lesser passer, measuring a player's input ("team-sensitive" WOPR/WTPRR) against that of his teammates is one key to whether or not he'll have a chance to consistently produce.
An absolute clinic and so fun to read. Thank you so much for the deep dive into the origin of some of these stats and logic that goes into building them, as much as the YTD analysis on what they mean in practice.