Track's proposed Dominance Metric
- Thread starter nerfstate
- Start date
obviously simple and intuitive but it sort ignores the incentives inherent in the NCAA's calculation of dominance. for example you often see wrestlers pressing in the final minute to secure a major decision, or they'll give up a stalling point to ensure a narrow victory. using this new metric, you can get a takedown to a pin in sudden victory and follow that up with 11 straight 2 point losses and you'd be considered average.
i personally like the cliff-edge property of the dominance score since the most important thing is victory. you win, you get 3 points. you can earn more, up to double, but then that's the max.
i'd like to see the correlation that this metric has with the dominance score, i'm assuming it would be pretty high. i'm also assuming the wrestlers who would benefit most from this are those who wrestle in worse (read: non Big Ten) conferences or who go to a lot of open tournaments.
it would have been nice if they linked us to a spreadsheet or leaderboard rather than only telling us zain and spencer's 2018 numbers.
i personally like the cliff-edge property of the dominance score since the most important thing is victory. you win, you get 3 points. you can earn more, up to double, but then that's the max.
i'd like to see the correlation that this metric has with the dominance score, i'm assuming it would be pretty high. i'm also assuming the wrestlers who would benefit most from this are those who wrestle in worse (read: non Big Ten) conferences or who go to a lot of open tournaments.
it would have been nice if they linked us to a spreadsheet or leaderboard rather than only telling us zain and spencer's 2018 numbers.
All of your points are excellent ones--especially the fact that the NCAA dominance score is calibrated to the current ruleset-dictated motivations. It would be pretty interesting to consider a new dual-scoring method too--I'm sure there is plenty of reason it is the way it is, but the idea of the final score of each bout contributing 10% of the dual score is interesting to look at.
I prefer what @LemonPie used back in 2019 when Nolf and Nickal were battling for the Hodge. His point differential per seven minutes is simple enough when the match goes full length, and when there's a pin he came up with an idea. Look at Nolf's tab and the first match is against Andrassy of Kent State. Nolf is up 12-5 when he pins him, so he is given an additional four points for the NearFall he would've earned.
Over the course of the season, you subtract points given up from points earned, divide that by the total number of seconds wrestled, and multiply by 420 to give you PD/7.
Over the course of the season, you subtract points given up from points earned, divide that by the total number of seconds wrestled, and multiply by 420 to give you PD/7.
The obvious hole in that method is the first-period fall.
David Taylor's 11-sec pin would be worth 6 pts in that metric, where he could've gotten 15+ via TF or via pin after building a big lead.
Put the problem on the chalkboard and I’ll figure it out after I’m done mopping these floors.
Take that ‘obvious hole’ and shut it before I break my foot off in it.
Sure, only 6 points work for David Taylor. But only 11 seconds work against him. For the match, his Point Differential would be 229.091, or the equivalent of a 15-0 tech in 27.5 seconds.
Is there a slight statistical bias in the favor of tech falls when compared with pins? Maybe, but probably only marginally. And for reference, Bo Nickal, with his dozen or so 1st period pins in 18-19, had a better Point Differential than Jason Nolf over the course of the season.
Speaking of Rivals Plus content, Lemon Pie is invoked and appears!
It’s just so incredibly insulting that I could spend hundreds of hours of my free time trying to introduce new metrics to the sport, and then some rube, who hasn’t paid any attention to my work, would suggest that I don’t understand the difference between a 6-0 decision and a 1st period pin.
“rube” - a word that should be used more
I think you missed the second paragraph. A takedown into a pin in 11 seconds would actually be a point differential of 229.1 per seven minutes.
[2 (+4) / 11] x 420
For context, after NCAAs in 2019, Nolf had a 14.76 PD/7 and Nickal had a 16.16 PD/7
EDIT: and apparently Lemon noticed this comment as well.
Hey, just because Ruby Tues went to sleep for 6 months and missed the Facundo talk, we don’t have to keep bringing it up again and again.
I did miss that paragraph, and that does rectify what I mistakenly thought was an issue.
Sloppy error that I might regret, except that the resulting overreaction was both entertaining and enlightening.
I just hope that someday I can be good enough that he calls me fat.
From the evidence, I'd say TrackWrestling should hire LemonPie. Or maybe he can start his own Rokfin channel ;-)
I do tend to think @psumacw is right that the official award shouldn't be given using the new metric without also looking at dual scoring but Lemon's is clearly the better one of the bunch--and he showed his work! This would also require groups like Track to record their data better, right? Didn't you have to jump through some hoops to extrapolate some of this, @LemonPie ?
I do tend to think @psumacw is right that the official award shouldn't be given using the new metric without also looking at dual scoring but Lemon's is clearly the better one of the bunch--and he showed his work! This would also require groups like Track to record their data better, right? Didn't you have to jump through some hoops to extrapolate some of this, @LemonPie ?
Similar threads
