The debut and popularization of BPM

[Crow]

MPG? 18.

[Crow]

Big men may have or not have value from spacing from their own shooting or the impact of their presence on spacing for other's shooting. Short list but seems to make sense from top down that spacing is observed in RPM but maybe not in BPM if it is not 3pt shot driven? Is decent big man mid-range shooting a positive? I'd think so in some cases.

[ampersand5]

I apologize if I missed it in the thread, but can someone point out the major discrepancies between JE's SPM that he uses for RPM and BPM? I think RPM uses height, but I really have no idea.

Also, is it possible to see BPM be used for RPM in the future?

[Mike G]

BPM is now showing up for the playoffs at b-r.com
Among the top 100 in minutes (>112), the best and worst:

Code: Select all

BPM    player        tm    PER   ws/48
8.6   Tony Allen     Mem   16.1   .176
8.4   Blake Griffin  LAC   25.8   .199
8.3  al-Farouq Aminu Dal   20.3   .199
7.7   Tim Duncan     SAS   24.0   .228
7.4   Steph Curry    GSW   24.4   .216
7.4   Marcin Gortat  Was   22.6   .240
7.3   LeBron James   Cle   23.8   .147
7.3   Draymond Green GSW   19.3   .189
7.2   Chris Paul     LAC   24.5   .231
6.9   Jimmy Butler   Chi   21.8   .229
6.8   Patty Mills    SAS   23.2   .261
6.5   James Harden   Hou   25.4   .229
6.2   Marc Gasol     Mem   22.5   .203
6.1   Otto Porter    Was   16.8   .198
5.7   Andrew Bogut   GSW   16.3   .170


BPM    player        tm    PER   ws/48
-3.3  Timofey Mozgov Cle   17.9   .146
-3.4  Tony Snell     Chi    4.8   .036
-3.8  Kyle Lowry     Tor    7.9  -.115
-4.0  Jeff Green     Mem    8.4  -.036
-4.1  Corey Brewer   Hou    9.8   .010
-4.1  Nene`          Was    5.2  -.019
-4.9  Jamal Crawford LAC   10.1  -.027
-5.6 Dennis Schroder Atl   13.1   .023
-6.2  Ersan Ilyasova Mil    7.8  -.045
-6.5  Tony Parker    SAS    6.3  -.067

Not too many anomalies, even in limited minutes.
Tony Allen -- the Instigator

[Crow]

I have registered some questions / possible concerns about how BPM rates players at the extremes on certain stats including assists and defensive rebounds. Tonight I checked the lower performers on these stats. For now, take this simply as descriptive findings. For the 40% of league with 100 plus minutes this season and less than or equal to 9% assist rate, approximately 75% were rated below neutral on OBPM and BPM. For the 40% with less than 12% DReb%, approx. 85% were rated below neutral on DBPM and over 80% below neutral on BPM. It is really hard to be above neutral on these measures unless you exceed these thresholds. Overall only 30% of players over 100 minutes are above neutral on BPM. So it is a bit less likely to be above neutral on BPM if you are below 9% on assist rate than generally and a lot tougher if you are below 12% on defensive rebounding. I expect that higher minute players tend to do better on this metric as well as other metrics so the large group performance is not a total surprise when considering the talent curve. It would probably also be worthwhile to check the RPM estimates for these groups to add that relative performance comparison. Overall 30% are above neutral on RPM but are the percentages above neutral on RPM comparable to the BPM rates when failing to exceed these assist and DReb rate criteria? If one metrics are valued and used, it would be good know if they behave fairly similarly or notably different when players are above or below certain specific stat performance levels.

[Crow]

For the 40% at or below 9% Assist Rate, OBPM is on minutes weighted average just .03 higher than ORPM. No issue there. For the 40% at or below 12% DReb rate, DBPM is on average .14 below DRPM. That is still small but almost a 5 times bigger average difference. The discrepancy might not be coming from rebounding but instead other parts of defense, perhaps shot defense. FWIW.

[ampersand5]

Thanks Crow for bringing that up.

On one hand, its possible that what you are describing is just a byproduct of BPM. BPM does not know what makes a good basketball player, so in some instances, it might be displaying something that is not met by reality.

On the other hand, you have something extremely fascinating (why I think BPM is so incredible). What you outlined might be real and what you highlighted could be revealing something important about basketball. It might be the case that players/coaches are now going to learn from different styles of play that lead to suboptimal BPM numbers and alter their game accordingly.

[huevonkiller]

1. Same as regular season. Regressed with 4 games of 0 MPG.

2. I looked at quite a few more things, but didn't get anything that would help. Eventually I decided to put off any further experimentation to the off season and just do the bugfix for now.

It is possible that the stat is biased by LeBron being in sample. Hopefully some additional testing will show for certain whether BPM fits as well historically as in the sample. I don't think this is a big issue from what I have seen. Basically, BPM thinks LeBron was maybe slightly better than MJ (when you include his Wizard years) and is already approaching MJ in minutes, both in regular season and playoffs.

Well first let me congratulate you and BBR on your work, great job to you all.

Now as to what you said in your previous post, that's not quite what BPM is insinuating. BPM is also saying LeBron from ages 21-30 is slightly better in the regular season than Jordan from ages 22-30 (a harder feat). LeBron's career stats are already better than Jordan's if you just go by Chicago Bull era. LeBron produces nearly the same PER on less usage, which is why we're having this discussion.

Jordan plays more minutes in the regular season in his prime, LeBron plays more minutes in the playoffs through comparable points in their careers.

Here's another way to look at it:
LeBron from 21-30 was at 10.1 BPM/89.1 VORP vs Jordan's 10.0 BPM/78.5 VORP (at 22-30). They both are nearly equal in terms of PER once you adjust for usage as well. Jordan plays more in his prime, but LeBron conserves himself for the post-season.

The real difference comes in the playoffs. Jordan takes over offensively and his usage sky rockets, but his PER doesn't really get a nice boost given that extra usage. I'm assuming (having looked at average playoff defensive rating) this happens because the 80's and early 90s were actually not the Gold Standard for defense, contrary to the narrative. LeBron's nearly Jordan's equal in PER but on noticeably less usage (an impressive feat). LeBron has the Usage-PER-Ratio edge it looks like.

I suppose if one were to just look at Points Per Game or something like that, then yes it would be a mystery why they are both comparable. But looking at all the boxscore metrics, it is easy to see they are both on an equal tier. LeBron is 30, though, with essentially Jordan's career minutes at Jordan's rate.

[Mike G]

LeBron's nearly Jordan's equal in PER but on noticeably less usage (an impressive feat). LeBron has the Usage-PER-Ratio edge it looks like.
...

Are you saying rebounding and/or passing have more value than scoring? or what?

What remains of LeBron's career will determine if he's the greatest of all time. One big separation between Jordan and him is that MJ was 6/6 in his Finals' trips. LeBron is 2/5, will be either 2/6 or 3/6 after this season. A dynasty in Cleveland is iffy at best; but 4/8 might be about equivalent.

Jordan probably got more life out of his legs by retiring for a couple years, a couple of times. LeBron looks indestructible at times, though of late he tends to run over people more. And this is always inviting injury.

http://bkref.com/tiny/UDzl9
LeBron is within 120 RS minutes, and 190 playoff minutes, of Jordan's Bulls career.

[huevonkiller]

What remains of LeBron's career will determine if he's the greatest of all time.

He's just 30 years old, so he'll probably retire as the better player APBR-wise. And reality-wise imo.

One big separation between Jordan and him is that MJ was 6/6 in his Finals' trips. LeBron is 2/5, will be either 2/6 or 3/6 after this season. A dynasty in Cleveland is iffy at best; but 4/8 might be about equivalent.

Well to me it just means Jordan loses in the first round of the playoffs or Conference Finals, whereas LeBron in "off" years, is still able to propel his team to the Finals. Look at his "off" year in 2011, versus Jordan's in 1995. Look at the Heat this year versus the 1994 Bulls. LeBron is basically Jordan except he plays more minutes. I have no idea how many championships his team will win, I do not know what the future holds.

Fivethirtyeight recently said Jordan's titles only amount to 5 or somewhat less than 6 in the current Era, FWIW. I've never really judged either player on titles though.

Jordan probably got more life out of his legs by retiring for a couple years, a couple of times. LeBron looks indestructible at times, though of late he tends to run over people more. And this is always inviting injury.

http://bkref.com/tiny/UDzl9
LeBron is within 120 RS minutes, and 190 playoff minutes, of Jordan's Bulls career.

LeBron's ankles are huge, his body is just huge and sturdy. I think that's helped him avoid injury. Among other things.

[BasketDork]

So to clarify, the team adjustment is:
(((Tm Efficiency Differential × 120% - Sum of Team's BPM)÷5) ?