Can any math expert explain what Jokic is doing to get such high defensive advanced stats?

[dhsilv2]

I think your premise is wrong. The advanced stats I prefer to use, RPM, has embiid the 12th best and jokic the 65th best.

But that doesn’t fit narrative that people are pushing that advance defensive metrics are pushing jokic in the MVP. Race.

There is no narrative that Jokic is a better defender that I'm aware of. i saw a few silly posts LAST season about it, but it was mostly just that. Silly. People are pushing that Jokic is the best player in overall metrics, yes. Most of the metrics people discuss in 2023 are +/- derived and go through regression, no raw +/- or even APM. People aren't generally using box based metrics beyond the simple eye test to make sure they are fitting with the real advanced stats. And generally speaking they line up well enough. But you don't break down box metrics into offense and defense if you expect to get good results. They overwhelmingly start breaking when you start getting into the top tier players.

[ITYSL]

I dunno, he might. I'm always up for a little contrarian argumentation.

Embiid does take a lot of plays off, too. Might be some elaborate stat argument based on that. Don't think it would end well but I'd hear the case out.

The only one I can think of is that the Sixers are one of the worst defensive transition teams (5th worse in PPP allowed), which may be partially explained by Embiid sometimes not hustling back on defense. I'm sure there are other factors of that stat which I'm unaware of.

[HotelVitale]

I dunno, he might. I'm always up for a little contrarian argumentation.

Embiid does take a lot of plays off, too. Might be some elaborate stat argument based on that. Don't think it would end well but I'd hear the case out.

The only one I can think of is that the Sixers are one of the worst defensive transition teams (5th worse in PPP allowed), which may be partially explained by Embiid sometimes not hustling back on defense. I'm sure there are other factors of that stat which I'm unaware of.

I doubt an argument would bear that conclusion out, but I do get annoyed that Embiid isn't smarter with his energy on defense. He'll take a play off in a way that ends up in an easy dunk, or make a senseless foul on a good FT shooter to stop a play needlessly, and then two plays later find himself motivated enough to make 5 counter-moves to spin into a 15 ft fadeaway. That he could have got without any of that effort.

He's definitely been better with that the second half of this season but it's still a thing. And something I don't see with Jokic, who's generally very aware of how to be efficient with his movements at all times.

[JonFromVA]

I dunno, he might. I'm always up for a little contrarian argumentation.

Embiid does take a lot of plays off, too. Might be some elaborate stat argument based on that. Don't think it would end well but I'd hear the case out.

The only one I can think of is that the Sixers are one of the worst defensive transition teams (5th worse in PPP allowed), which may be partially explained by Embiid sometimes not hustling back on defense. I'm sure there are other factors of that stat which I'm unaware of.

The most obvious instance of this last night involved Harden not getting back and Maxey being content to stay with his man and not turn and run back to help.

Offensive efficiency, offensive style, and rebounding also play a role in creating fast breaks for the other team. Not to mention just having a solid defender with some length as the guy who's in position to help stop a fast break should be a huge advantage. Some players can defend a 3v1 and others are just hopeless in an open court 1v1.

[ITYSL]

I dunno, he might. I'm always up for a little contrarian argumentation.

Embiid does take a lot of plays off, too. Might be some elaborate stat argument based on that. Don't think it would end well but I'd hear the case out.

The only one I can think of is that the Sixers are one of the worst defensive transition teams (5th worse in PPP allowed), which may be partially explained by Embiid sometimes not hustling back on defense. I'm sure there are other factors of that stat which I'm unaware of.

The most obvious instance of this last night involved Harden not getting back and Maxey being content to stay with his man and not turn and run back to help.

Offensive efficiency, offensive style, and rebounding also play a role in creating fast breaks for the other team. Not to mention just having a solid defender with some length as the guy who's in position to help stop a fast break should be a huge advantage. Some players can defend a 3v1 and others are just hopeless in an open court 1v1.

+1

It's also just the nature of the beast, with your center being your main scoring option. It makes natural sense that on misses and even makes, a center will take a bit longer to get back for rim protection. They're bigger and slower. And that extra time can mean all the difference. No other center is in the top 30 in FGA/100. (If you count Giannis as a center, well, he's an anomaly and there's a reason why he's called the Greek Freak.)

[An Unbiased Fan]

I've posted about this on the PC board many times. +/- stats were meant to be roster evaluation stats, they don't correlate to individuals. Jokic benefits from how he's used in that roster rotation, nothing more. He's not a great defender

[FluLikeSymptoms]

We all know defensive stats suck even more than offensive ones. I don’t know why we’re settling.

We're not settling. There just isn't a perfect way to predict a player's defensive value. Especially not when you're trying to extrapolate a stat that was collected over an entire season where a player is going to have ups and downs in terms of effort and playing through injury to a theoretical predictor of how they would perform in a single game or playoff series.

It is settling. We decided 20 years ago that we needed advanced stats because we did, some have been developed, they’re not good enough. And the idea of, “it’s the best we’ve got so what can we do” is lazy and ridiculous. Jokic is a walking indictment of defensive analytics but no one will throw them out. He’s fine defensively. He’s the Steph Curry of Cs.

[kobe_vs_jordan]

We all know defensive stats suck even more than offensive ones. I don’t know why we’re settling.

We're not settling. There just isn't a perfect way to predict a player's defensive value. Especially not when you're trying to extrapolate a stat that was collected over an entire season where a player is going to have ups and downs in terms of effort and playing through injury to a theoretical predictor of how they would perform in a single game or playoff series.

Think there some decent ones they just not commonly used. Primary field goal percentage against in a paint is great stat for a big man stat. Deflections stats does a decent job are capturing active defenders without overcompensating for gambling like steals.

For wings , points per possession is pretty good. Wings that can plan man defense usually rotate well. For bigs, man defense doesn't always translate to rotation defense.

None of the stats will tell you who is the best but they will identify the good and the bad.

[Hobo4President]

DBPM is a purely box score derived stat that doesn't accurately measure a player's defensive contributions and is weighted by position and compared to teammates. Jokic has high DBPM because DBPM is Total BPM - OBPM. BPM weights for centres more liberally than OBPM in the stats that Jokic is unusually high on such as assists and defensive rebounds. The box score stats are also compared against his teammates, and since his DRB% and AST% are both top two in the league it suggests he outpaces his teammates on those stats by a significant amount. All this is basically saying that when Jokic gets 10 assists and 12DRB his OBPM increases but his BPM increases much more which drives up his DBPM.

Jokic isn't the only centre with a high DBPM for this exact reason - Sabonis is 3rd in the league in DRB% and 19th in AST% and as such is 13th in the league in DBPM.

You can read through https://www.basketball-reference.com/about/bpm2.html and see the coefficients and an explanation of the calculations. They even say that BPM is generally pretty good for measuring offense but should be taken with a grain of salt in regards to defense.

"What does this mean? Box Plus/Minus is good at measuring offense and solid overall, but the defensive numbers in particular should not be considered definitive. Look at the defensive values as a guide, but don't hesitate to discount them when a player is well known as a good or bad defender."

[JonFromVA]

I've posted about this on the PC board many times. +/- stats were meant to be roster evaluation stats, they don't correlate to individuals. Jokic benefits from how he's used in that roster rotation, nothing more. He's not a great defender

+/- stats do not support the idea that Jokic is a great defender, rather it's the box score derived so-called "advanced stats" which do.

Plus-minus data is just about the only reliable way to begin to evaluate individual defense and impact in general.

The models, even the good ones like RAPM are just trying to tackle the tough problem of trying to tease individual performance from a lineup based stat; but since they don't give us anything more than a number ... we are IMO better off doing it on our own.

[G35]

I am far from being a maths expert, but the general consensus is that advanced stats for individual defense are not very useful.

D-RAPM, purely looking at +/- impact of a player on the defensive end would probably be the best bet, but it's way too noisy for single-season data. A lot of other advanced stats using +/- combine it with box score counting stats, but dont know if those are actually better. DBPM, from basketball-reference, is by far the worst, since it's just using box score counting stats like rebounds and steals.

I think it's hard to really capture individual defensive impact as defense is really more team-oriented compared to offense.

edit:
having a quick look at pure +/-, embiid seems to have a slightly bigger defensive impact than jokic this season:
http://www.pbpstats.com/on-off/nba/team?Season=2022-23&SeasonType=Regular%2BSeason&TeamId=1610612755&PlayerId=203954
http://www.pbpstats.com/on-off/nba/team?Season=2022-23&SeasonType=Regular%2BSeason&TeamId=1610612743&PlayerId=203999

-4.22 vs -3.78 Pts per 100 Possessions - Defense when on the floor

We all know defensive stats suck even more than offensive ones. I don’t know why we’re settling.

We're not settling. There just isn't a perfect way to predict a player's defensive value. Especially not when you're trying to extrapolate a stat that was collected over an entire season where a player is going to have ups and downs in terms of effort and playing through injury to a theoretical predictor of how they would perform in a single game or playoff series.

This is an indictment on how Kevin Garnett is evaluated.....

[JonFromVA]

DBPM is a purely box score derived stat that doesn't accurately measure a player's defensive contributions. Jokic has high DBPM because DBPM is Total BPM - OBPM. BPM weights by position more liberally than OBPM in the stats that Jokic is unusually high on such as assists and defensive rebounds. The box score stats are also compared against his teammates, his high DRB and high assists in comparison to his teammates increases his BPM more than his OBPM. All this is basically saying that when Jokic gets 10 assists his OBPM increases but his BPM increases even more which drives up his DBPM.

You can read through https://www.basketball-reference.com/about/bpm2.html and see the coefficients and an explanation of the calculations. They even say that BPM is generally pretty good for measuring offense but should be taken with a grain of salt in regards to defense.

"What does this mean? Box Plus/Minus is good at measuring offense and solid overall, but the defensive numbers in particular should not be considered definitive. Look at the defensive values as a guide, but don't hesitate to discount them when a player is well known as a good or bad defender."

The bpm2.html links has some neat stuff in it, including a plot of BPM .vs. RAPM that shows individual points and the overall r-value for the fit. You only need to click on individual players to see how far BPM can drift from predicting the actual RAPM value.

It may work out to be not too crappy on average, but any specific player may be an outlier, and I get the impression an r-value of 0.427 for DBPM2 suggests exactly what the author tells us (when in doubt do not trust this stat). The r-value for OBPM2 by comparison is 0.655 which is still not great.

All this before even deciding if we think RAPM is reliably telling us something useful or if this model is actually holding up after all the changes that have happened.

[Bum Adebayo]

For defense, I prefer to look at playoffs. RS is too long, stars rest more on defense and take plays off, Embiid obviously does too, as do all offensive stars.

[Hobo4President]

For defense, I prefer to look at playoffs. RS is too long, stars rest more on defense and take plays off, Embiid obviously does too, as do all offensive stars.

The playoffs have their own problems too. The playoffs have benefits in that teams are actually trying their hardest and we see a player's performance under pressure. However, the playoffs are very match up dependent, we typically only see players matched up against teams in their own conference, they play a smaller amount of teams and we also have a smaller sample size. If the playoffs were a round-robin format where the top team from each conference make the finals it'd probably be more accurate but still has major issues. Even a double elimination bracket would help.

[Pachinko_]

I think the Jokic defense situation with the advanced stats is a trap that the basketball Gods have set up to expose those who never actually watch any games or simply don't pay attention to defense, because if you do you can't possibly rank Joker *that* high.

I mean if you have a keen eye you will notice that the guy generally has good positioning and probably puts his hands on the ball a bit more frequently than most Cs and gets some deflections and rebounds that looked lost, but other than that he's simply too slow, too earthbound and not even particularly big for an NBA Center. I'd say average defender. Personally I like him at that end because he's smart and tough and not scared of contact but at the same time I'd like my big to be more of a shot blocker and rim protector. Just more of everything really (except rebounds where Joker is already awesome with his quick hands and wide body)

[brutalitops]

Being very good offensively limits fast break, and makes defense better, I don't have the numbers to back it, just a guess.

This. Denver's transition defense is also tragic. When he's off the court, they score less and also have such a worse defense because they can't set up.

Jokic is above average (Not going to say good) in set defense. And average in transition. I

[Scalabrine]

It doesn't make any sense at all.

Embiid is averaging 10.2 rebounds (8.4 defensive), 1.1 steals, 1.7 blocks. The 6ers as a whole have the 9th best defensive rating in the league at about 113.2 points allowed per 100 possessions. Their defensive rebounding is 7th in the league in terms of %.

Jokic is averaging 11.9 rebounds (9.5 defensive) 1.2 steals, .7 blocks. The Nuggets as a whole have the 16th best defensive rating in the league at about 114.4 points allowed per 100 possessions. Their defensive rebounding is 9th in the league in terms of %.

By purely WATCHING the game, no one would pick Jokic as a better player than Embiid and yet all the metrics paint Jokic as this elite defensive presence. I watched Embiid have 4 incredible blocks last night and watched Jokic be a revolving door in the paint vs. the Raptors not too long ago.

I think this has to be related to the team composition both deal with. Jokic plays with a team that is worse overall defensively but does so much rebounding and improves them so much in the on/off numbers that the elite offense and scoring that he is delivering is making their defense look better because his team is out pacing an opponent.

I can't think of any other reason why. Outside of maybe Bruce Brown and KCP who are solid, the guys that Jokic has are not good defenders at all, but with Jokic they win off of offense.

I watch a lot of both players and I think Jokic is the better player.

You think Jokic is the better defender than Embiid?

Nope. I said that I think Jokic is the better player in response to the bolded above.

[ShootersShoot]

I dunno, he might. I'm always up for a little contrarian argumentation.

Embiid does take a lot of plays off, too. Might be some elaborate stat argument based on that. Don't think it would end well but I'd hear the case out.

The only one I can think of is that the Sixers are one of the worst defensive transition teams (5th worse in PPP allowed), which may be partially explained by Embiid sometimes not hustling back on defense. I'm sure there are other factors of that stat which I'm unaware of.

I dont think you can pin bad transition defense on the center..they are usually the slowest guys out there and the least agile. You want embiid to be the first guy back on the break or something?

[payton2kemp]

If defense was included in MVP discussion Jokic would never win.

[Modulate]

It doesn't make any sense at all.

Embiid is averaging 10.2 rebounds (8.4 defensive), 1.1 steals, 1.7 blocks. The 6ers as a whole have the 9th best defensive rating in the league at about 113.2 points allowed per 100 possessions. Their defensive rebounding is 7th in the league in terms of %.

Jokic is averaging 11.9 rebounds (9.5 defensive) 1.2 steals, .7 blocks. The Nuggets as a whole have the 16th best defensive rating in the league at about 114.4 points allowed per 100 possessions. Their defensive rebounding is 9th in the league in terms of %.

By purely WATCHING the game, no one would pick Jokic as a better player than Embiid and yet all the metrics paint Jokic as this elite defensive presence. I watched Embiid have 4 incredible blocks last night and watched Jokic be a revolving door in the paint vs. the Raptors not too long ago.

I think this has to be related to the team composition both deal with. Jokic plays with a team that is worse overall defensively but does so much rebounding and improves them so much in the on/off numbers that the elite offense and scoring that he is delivering is making their defense look better because his team is out pacing an opponent.

I can't think of any other reason why. Outside of maybe Bruce Brown and KCP who are solid, the guys that Jokic has are not good defenders at all, but with Jokic they win off of offense.

Jokic is leading the league with 45 kicked ball violations.
2nd place has 17.

https://deadspin.com/nikola-jokic-denver-nuggets-kicked-ball-violation-nba-1850226012

Well that explains everything. Dude is playing soccer on defense.