Quote:
Originally Posted by EPiPeN11
i always love when ppl say "advanced stats" like they are really good stats when most of them are trash/super flawed. really the only ones that are really good at measuring what they are supposed to are the percentage ones aka TS%, eFG%, and TRB% (ORB% and DRB% too obv)
Even as the resident advanced stats skeptic, I think this is largely either hyperbolic or wrong. (Also, DRB% isn't a super great stat, imo.)
The problem is that we (both here and in the larger "intellectual" basketball community) don't have a very systematic understanding of how these stats work, what they measure, and most importantly, what they don't. We all know that box score based stats are not capturing a lot of stuff that is material to the game, but somehow forget this when these stats are rolled up into one uber stat.
The reason that APM is such an interesting stat is that it has the potential to capture "intangibles" but it's still both messy as **** and is more limited in application than a lot of it's proponents either realize, or admit. (Even aside from small sample size/large error problem, it's still mainly measuring a player's effectiveness
in the role he is playing.)