ThunderBolt wrote:slick_watts wrote:ThunderBolt wrote:So I’m curious, I watch Ferguson and he seems to be playing defense but the analytics aren’t completely supporting this yet. Why?
rapm and all prior informed metrics are designed to be skeptical of short term outliers and ferg himself only has barely over 1k minutes played to begin with.
most analytics we talk about, or at least all rapm-liked ones, are predictive rather than descriptive. which is why they will appear to be skeptical.
something like DBPM would be descriptive, and ferguson currently has a decent DBPM this season. DBPM is sorta shoddy but you get the point.
After how many minutes will RAPM no longer be skeptical?
many hundreds of minutes i am guessing.
it's worth noting that despite the success of the ferguson starters, the thunder defense is 2.5pp100 better on the year with him on the bench. he is also not particularly productive on the boxscore defensively and fouls a lot. the starting unit represents less than 1/3 of ferg's minutes. we're allowing > 10pp100 more to opponents in ferguson minutes other than the starters than the ferguson starters.
so it isn't so cut and dry. ferg appears to contribute positively with the starters defensively but this only encompasses a fraction of his total minutes, and the other minutes he plays isn't nearly as successful defensively. it's possible there may not be a meaningful defensive impact for rpm to ferret out.