I dunno what their philosophy there is but PER is also based on MJ as a baseline. 

unless this is a myth

is it a myth 

Damn interested to know what they did to make this so

That’s pretty bad…

But, logically, doesn’t that mean that the total BPM stat is bad as a whole? If “whatever is left must be defense” clearly isn’t defense, then what is it supposed to be? The stat makers, as you say, assume it’s defense. But it clearly isn’t in this case, as Jokic’s defense should be a slight positive at best and neutral-to-bad at worst. That extra value should be represented by something; if it’s offensive value, it should be present in OBPM. If it’s an intangible thing, then it should be represented by something like an IBPM. Otherwise, it’s just floating.

Jokic had a BPM of 14.2 this year with an OBPM of 10.1. If the 4.1 gap between the two isn’t defense, then what is it? If it’s offense, then it should be reflected in OBPM, no? And if it isn’t supposed to be an estimate of offensive value, but it also isn’t defensive value, then what value is it?

BPM doesnt know anything, it’s an estimate/prediction using box score mainly and team performance as a proxy for RAPM which is super noisy for small/even single season samples.

He was all defense, Jokic?

Is it ragebaiting when it’s a mistake due to stupidity?

If I remember correctly, it also oddly includes an interaction between defensive rebounding and assists (of all things) that blew up Westbrook’s numbers. I’d think that obviously applies to Jokic.

Easy when you don’t contest shots and get all the missed layups. And before someone comes with most of his rebounds being contested, I’m almost 100% sure this is because the player missing the layup is close and that counts as a contested rebound.

DBPM (and many other advanced stats) should be used to look for outliers, and supposedly that’s how teams often use it. They basically look for players that are far better or worse than expected, and then investigate why. The context helps you learn if it’s noise or if they are legitimately better than expected.

None of the concerns associated with DBPM apply to BPM. BPM is the core quantity being predicted/regressed, OBPM/DBPM try to figure out how BPM should be split.

What? MVP is determined by a media vote, and I highly doubt there are many voters who vote based on BPM.

Because foreigners watch basketball too?

I’m not even sure what you are even trying to imply here, just because one guy fucked up doesn’t mean all foreign writers don’t know ball.

Also, it’s a weird thing to complain about when SAS and Kendrick Perkins both have votes.

Basketball is a global game with a global fan base and player base. Your best player is Canadian and you used a lottery pick on a Serbian.

Relevant flair

You yourself mentioned that a foreign writer, who probably looked at these stats

I did not mention that the foreign writer looked at DBPM, which they very obviously did not given the rest of their selections.

The Italian writer appeared to think he was voting on All Offense teams rather than All Defense.

So it is to some extent influencing these lazy media members.

Again, it very much is not. No one cares about DBPM.

You’re right, my bad. 79th percentile isn’t bad, but of course it’s still a massive difference compared to a top-10 defender

I mean, Gordon was injured most of the year and played hurt when he came back. His d-epm was better in previous years. It’s not the perfect stat, but way better than dbpm.

True, but the metric basically reflects the DPOY voting (top 3 in D-EPM are Wemby, Ausar, and Chet). The voters either used this metric to determine their votes, or the stat simply matched the eye-test.

Better to just parrot whatever the top voted comment on Reddit says

it’s still miles better than dbpm because it takes into account lineups and on/off data, where as dbpm is strictly an aggregate box score metric. Ranking defenders based on dbpm is about as accurate as ranking players based on PER.

for context he’s the 38th best defensive center

Based on what? Your own subjective opinion?

It’s actually a fairly good tool.

Do you watch last nights game?? Just forget the stats for a second what did your eyes tell you?

The whole way we evaluate players on slashlines does overrate rebounds

Everyone looks for PRA, half the time we add efficiency, 1 in 10 times we add turnovers

Yeah like when teams would send Kawhi when he was at his defensive best to the corner so they could play 4 on 4.

The numbers don’t lie though. It’s the people crafting the narrative with the numbers that are lying. If you were purely looking at stats you would believe that Whiteside was one of the premier interior defenders in the 19’ season because he led the league in blocks. But he got those numbers because he was hunting for blocks and would end up hurting the defense sometimes.

If you don’t add the eye test as well then you could pick any statistical category that favors your player and make anyone look good. It’s all arbitrary selections anyways.

He can’t contest well and can’t switch on anyone, but he calls out plays, gets in passing lanes, and forces deadballs when he is out of position. it makes other teams offense less efficient.

If he is so bad at defense why is that bad why don’t teams score on him more? Don’t argue that Aaron Gordon is prime Hakeem cleaning up his mess. At the end of the it is scoring more and stopping the other team from scoring, and he is great at one, and the other team is confusingly poor on the other.

Oh yeah, it was a noticeable thing sometimes how tired and beaten his defenders were for a while. Matthew Dellavedova averaged 6 points in 30 minutes per game that finals he guarded steph, but he had to play because they couldn’t tire Kyrie on him.

Steph is also a fascinating example of someone whose defense improved as their career went on, and kind of late. He was bad/poor early, got to average by the late 2010s. and depending on the year and the competitiveness, he was kind of a low end good defender in the 2020’s. Strengthening his ankles and his core so much helped, as wel as again schemeing.