APBRmetrics

Hey guys, been a while.

So various topics over on RealGM have had me considering something and I've spent some time hunting through the older posts and the b-ref blog, etc, leading me to the question:

How much difference does an individual player's scoring efficiency (ITO of TS%) really make to a team? Traditionally, you guys seem to mostly work with ORTG, which would seem a more comprehensive picture of player offensive impact, so maybe that's a better way to go, but has anyone done anything with usage-weighted TS differential against league average TS to see what the value is? Is that even a meaningful angle of analysis?

I saw in one of the other posts (Questions and Comments on ASPM) Mike G said that points per possession will always have value relative to opponent PPP, so is there a way to apply that at a player level and say this team had an ORTG of X and a DRTG of Y and his TS% contributed thus to his individual ORTG of Z, so the shift in value if his TS% had been some given other number would be the final value, or something like that? Is this just treading over individual ORTG?

I saw some stuff by EvanZ (PSAMS) and DMok (ASPM) and a few other more comprehensive examples; are those the only alternatives? I don't mean to diminish that work, but it seems to be a little deeper than what I'm looking for, though I might be thinking too simply. Is that the direction you need to go to in order to properly value individual player scoring efficiency? I saw Evan breaking it down by shot location relative to what an average player (at the same position) would do, I guess that's the more in-depth version of where I want to go.

Thoughts, links, better searches than what I've managed so people don't have to reiterate themselves?

Thanks in advance!

For the sake of clarity, I guess I'm talking mostly about primary offensive players. Major scorers and major playmakers. So players like Kobe, Rondo, etc.

How much of a difference does it make if your primary scorer is someone like Melo, who's typically in the 0 to +2.5% range compared to league average TS, versus a player who is in that +3.5 or better range?

What impact is there if you're a volume playmaker but an inefficient scorer using 10+ FGA/g? ElGee had an article about Rondo, for example, noting the general low-impact nature of his assists that kind of drove that comment, so I'm wondering what his scoring efficiency (or lack thereof) does to his ability to drive a quality team offense. I bring this up because these are the thoughts that spawned my post, so I figured I'd show you the whole thought process.

Cheers!

Here's the short version

1) 70% of ORTG (offensive rating) is explained by TS%. So, ORTG is mostly your points per shot, with other things added in.
2) If *all* you look at is TS%, you will probably think that many good offensive players are very bad and vice versa. This is thanks to the whole usage/efficiency tradeoff thing. Players that rarely shoot are typically catch-and-shooters (or catch-and-dunkers like Tyson Chandler) etc.
3) So less usage means less scoring. You can't win a game by making 100% of 1 shot per game, and so on.
4) There is a well-documented gain in team efficiency by being a volume shooter, regardless of efficiency (TS%). There are a plethora of reasons for this - you are going to get double-teams by using tons of possessions, which allows for kicks etc; your volume is partially due to necessity (better shooters get the ball at the end of the shot clock but likely have to take a more difficult shot, but a better one than other teammates might get). I could probably name ten more.

Even including turnovers/assists/rebounding from ORTG, you can still come to some logical conclusions about TS%. check out my formula that converts Usage & Efficiency to a usable number (an estimate of offensive impact per 100 possessions based on 8 years of +/- data). Play around with this!

Offensive Impact per 100 = (ORTG - LeagueAvg)*Usage%*1.035 + (Usage% - 20)*0.23 -0.1835

So to get back to your main question, it is quite possible for a volume shooter to be negative on offense (Monta Ellis), some average-efficiency volume shooters are still worth an extra point or two in efficiency margin per game (Allen Iverson, Derrick Rose, Russell Westbrook), and of course the best is to have high numbers in both (KD, LeBron, DWade, most of the great scorers).


Edit: Forgot the first Usage% term