Yeah it’s pretty quirky, from the source at sports reference.
3 Likes
Here’s an excerpt from The Midrange Theory by Seth Partnow, the former director of basketball research for the Bucks, founder of Nylon Calculus, and one of Dean Oliver’s (the inventor of offensive and defensive ratings) colleagues about why individual defensive ratings are not a good stat. If anyone wants a more long-winded explanation, here it is (apologies in advance for clogging up this thread with a fairly off-topic discussion)
Since we are explicitly defining player ability in terms of moving the scoreboard in favorable directions, it is tempting to…simply look at how the team has performed on offense and defense with the player on the floor, and voilà. The lure is especially strong when trying to measure defense, where we are desperate for useful guidance.
Unfortunately, this pulls us toward a common mental trap, which goes something like this:
- We need a good measure of defense.
- Individual defensive rating is a measure of defense.
- Individual defensive rating is therefore a good measure of defense.
Too which I have one response:
DEFENSIVE RATING IS NOT AN INDIVIDUAL STAT.Let me explain.
In 2019–20 Enes Kanter had a solid year for the Boston Celtics. In his 17 or so minutes per game, the Celtics scored 113.8 points per 100 possessions while allowing 103.9, meaning Boston outscored opponents by just under 10 points per 100 possessions. Meaning Kanter had a 113.8 ORTG, a 103.9 DRTG, and a +9.9 net rating for the Cs, right?Wrong.
The difference might appear to be semantic, but “The team did X while Player Y was on the court” and “Player Y had a net rating of X” are substantially different. The former is clearly just describing what happened. The second is at least implying causation, that the Celtics won Kanter’s minutes by nearly 10 points/100 because of Kanter.
I get why the term “rating” entered the popular lexicon. When Dean Oliver wrote Basketball on Paper in 2004, the concepts of “per possession” or “per minute” stats were new and foreign to the basketball world. Describing things in terms of “points allowed/100” looked math-y and might have been a bit of a turnoff. So it was instead called “defensive rating” to help the medicine go down.
Today, per possession stats are the norm rather than the exception and we’re awash in rating systems. That name, “rating,” is an unhelpful anachronism, implying some sort of calcu “lation or evaluation where none exists. It’s just points allowed/100, an expression everybody would now take in stride.
All of which is the long way to say that offensive and defensive ratings are not individual statistics. As such, Kanter did not have a 103.9 DRTG. The Celtics had a 103.9 DRTG with Kanter on the floor.116 But the “ratings” are established well enough as labels that we’re stuck with them, and it becomes a matter of constant repetition of the mantra DEFENSIVE RATING IS NOT AN INDIVIDUAL STAT until that too becomes broadly accepted knowledge.
edit: one other thing I’ll add is that most DRTG calculation formulas are black boxes, which means no one knows what actually goes into computing DRTG except whoever calculates it.
3 Likes
I don’t disagree with Clark’s defense yesterday… in fact the box score had him for 1 block and watching the tape I had him for 3 blocks. That would have made his rating better too
4 Likes
Also the box score stats are only as good as the home team stat crew. And it’s well known in college basketball circles that a lot of places tend to skew stats more favorably for the home team. Sometimes that’s just because they’re more familiar with their own players than the other teams players.I did stats for my alma mater for a long time as well as officially in the NCAA tournament. And it was well known there were some schools that just had terrible stat keepers so we always kept stats while visiting as well. I can still remember one game where we had 24 assists and the official box score had us for 12
2 Likes
On Stats/Efficiency:
I am somewhat resigned to the fact that there are not great defensive stats (i.e., the stats are not that illustrative), but another comment on the offensive efficiency. I generally think, conversely, offensive efficiency stats are pretty illustrative, albeit (like any other stat) are susceptible of one-sided interpretation. I think when you combine OEff with the counting stats, and some look at the on/off stats, you do get a pretty good picture. But, as Dave points out, you do have to use the dreaded eye test, as well, to filter all these things.
(also, on Dave’s latest post, I only recently became aware that apparently lots of folks think Stockton got plenty of assists–pun intended–from the Jazz official scorer).
On the game in general:
How about Taine’s little mid-range bullsh** game? Interesting stuff. That was one of my favorite developments yesterday, since I already thought he was a pretty good perimeter shooter, and I think he can get to the rim decently well. He does need a bit more touch from the mid range… He should watch Armaan mid-range tape…
And I remain terribly confused that Taine is basically a Power Forward. There are many things this season that have me scratching my head, but Murray essentially competing for small ball 4 minutes with Igor continues to confuse me…
2 Likes
I also though KC had 3 blocks. Kept waiting for the announcers to mention it.
1 Like
Let’s not lose sight that Malachi was not available yesterday. Hopefully Taine’s minutes were reflective of something other than necessity.
2 Likes
I think his minutes were by necessity – but not Malachi’s absence. By memory, Taine basically subbed in for Jayden. I know it’s odd for us to wrap our heads around, but Taine is basically a small ball 4 this year.
7 Likes
yeah it’s wild but Taine seemed to be the 4 on offense even when he and Kody were in together. Continues to top the list of things I did not see coming. Not sure who guarded who during that time but against Syracuse the 3/4 isn’t much difference anyway. As a zone team, they get away with basically playing two stretch 4s in Boeheim and Swider.
It also seems to be a game by game matchup thing. More teams play small or have those stretch types that don’t want to go inside anyway, it’s sometimes better to defend them with a guard better equipped to stay in front or recover on pick and pops. In the case of yesterday, they could get away with it defensively while adding another shooter.
5 Likes
I think he mostly matched up with the taller/older/Cornell x-fer Boeheim. Jimmy?
2 Likes
Yeah this is bizarre. Surely Taine isn’t a 4 in our system moving forward??
No but I don’t think they trust him with the ball as much as they’d want right now to be a guard. He plays there sometimes but that Kody was the guard when they were the 3/4 combo yesterday is revealing.
Tell ya what though, anybody that had doubts that Leon Bond would be primarily a forward at UVA should pay attention. This is the way the game is going. They recognize that.
5 Likes
Yeah, Taine is pretty loose with the ball on the few clips that I’ve seen where he is a ballhandler (mostly B/W game, Fiba game, and old NZ clips). I hope he cleans that up enough to gain Tony’s trust.
One open question in my mind is how Taine would do guarding an ACC shooting guard. He mostly guards bigs, even if they’re perimeter oriented. E.g., seeing him guard that Clemson dude who is the rebounding king of Youngstown St definitely didn’t look like a natural matchup.
One hilarious thing I found out today was if a national pundit says anything on social media in anyway positive about UVA basketball, the Auburn trolls come out in force with their “double dribble” talk. Can’t imagine still being that bitter… But hilarious to witness.
10 Likes
Love it. Their fans were awful in Minny. I am from that area as yall know and Auburn is like the rest of the rest of fans
5 Likes
They conveniently forget that the defender grabbing Ty’s jersey caused the double dribble.
5 Likes
Very true. But don’t point that out to them… Just fires them up even more… haha
1 Like
I don’t recall anyone posting this, but here is link to Buddy Boeheim post-game comments where he’s very complementary of CTB’s coaching against the SU zone (4:00 minute mark).
I’d never heard him post-game but he certainly comes across like he’ll eventually join his father’s profession.
https://www.si.com/college/syracuse/basketball/buddy-boeheim-breaks-down-virginia-loss
12 Likes
those boeheim kids can really fill it up or is it just their dad drawing up plays for them? Jimmy only rested for 2 mins!
1 Like
They have the green light and it shines bright
1 Like