2016 RAPM/RPM/etc. *Full 2016 RS + PS RPM & RAPM Updated 6/24*

[SideshowBob]

There is no way Lebron should have won the finals MVP last year, not with that efficiency. He had an extremely high usage rate but in reality this was like his 6th or 7th best playoff series, if that. I do think however that Curry should have won it instead of Igoudala.

Missing the forest for the trees

Just looked through NBAwowy for the 2015 Finals for the Cavaliers:

Series (569 Poss)

Lebron: 47.1% TS & 8.5% TO rate
Others: 47.4% TS & 16.2% TO rate

Cleveland w/Lebron ON the floor (274 MP, 527 Poss)

102.3 ORTG, 106.5 DRTG, -3.2 Net

29.4% ORB, 78.6% DRB

Non-Lebron players posted a 50.0% TS & 16.8% TO rate

Cleveland w/Lebron OFF the floor (24 MP, 42 Poss)

52.4 ORTG, 102.4 DRTG, -50.0 Net

4.0% ORB, 57.1% DRB (meaning GS put up 96% DRB rate and 42.9% ORB rate)

Players: 26.2% TS & 11.9% TO rate

[PearGreatness]

There is no way Lebron should have won the finals MVP last year, not with that efficiency. He had an extremely high usage rate but in reality this was like his 6th or 7th best playoff series, if that. I do think however that Curry should have won it instead of Igoudala.

I don't think he was the best player out there, but I do believe he was the most valuable. How awful would the Cavs have been without him?

I would have given the award to Curry though. Giving the fmvp to a player on the losing team just doesn't feel right.

[Johnny Firpo]

There is no way Lebron should have won the finals MVP last year, not with that efficiency. He had an extremely high usage rate but in reality this was like his 6th or 7th best playoff series, if that. I do think however that Curry should have won it instead of Igoudala.

Missing the forest for the trees

Just looked through NBAwowy for the 2015 Finals for the Cavaliers:

Series (569 Poss)

Lebron: 47.1% TS & 8.5% TO rate
Others: 47.4% TS & 16.2% TO rate

Cleveland w/Lebron ON the floor (274 MP, 527 Poss)

102.3 ORTG, 106.5 DRTG, -3.2 Net

29.4% ORB, 78.6% DRB

Non-Lebron players posted a 50.0% TS & 16.8% TO rate

Cleveland w/Lebron OFF the floor (24 MP, 42 Poss)

52.4 ORTG, 102.4 DRTG, -50.0 Net

4.0% ORB, 57.1% DRB (meaning GS put up 96% DRB rate and 42.9% ORB rate)

Players: 26.2% TS & 11.9% TO rate

Technically he might have been the most valuable but that's not what the award is for despite the name, and even if that was the case, it was because his team was so awful without Love and Irving because he simply wasn't playing that well by his own standards. He could have shot a lot better and actually win the series despite his team's shortcomings. His jumpshot was clearly nonexistent during the finals and while the Warriors played good defense, he missed tons of open jumpers.

[Johnny Firpo]

I don't think he was the best player out there, but I do believe he was the most valuable.

That doesn't matter even if true, you guys know well that the award's name is a lie. It's the "we think this is the player who played the best from the winning team" award.

[Doctor MJ]

I got it from [ https://www.dropbox.com/sh/teutg7zvxudqnlw/AAAUkNkDUG0KWeewPZbnwS2ja?dl=0 ].

But I think there may be a miscommunication. You're mentioning when Lebron left Cleveland for Miami. I'm looking at Lebron's return to Cleveland (2015 season) and how his efficiency is much lower than when he was in Miami (specifically 2014 and 2013 seasons).

Or are you suggesting there's something about Lebron's relative role in Cleveland versus Miami, that hasn't changed since he left and then returned, which is driving the difference?

I believe this is JE's multi-year (3-year with most weight on listed year, less weight on n-1, and less on n-2). Not sure why it's labeled as PI RAPM. The only PI sets we have are the ones on GotBuckets.com (up to 2014) and the one's linked in colts18's signature (up to 2013).

Thanks for clarifying.

But even using GotBuckets or colts18's document we see scoring efficiency not always making much of a difference in ORAPM for scorers (though I didn't standardize the results). Again, we could use Lebron as an example, and compare his 2013 and 2010 years. Or we could look to Kobe who somehow had higher ORAPM in 2006 than Nash (using colt's). Jordan in the 1998 PI RAPM is another example.

I just don't understand how Lebron saw a decrease in his ORAPM when his efficiency skyrocketed (outside of the two possibilities I mentioned earlier). Or how '06 Kobe and '98 Jordan neared the top of the league in ORAPM without being very efficient.

I understand that, if the alternative of a teammate shooting is just that much worse, volume scoring on slightly above average efficiency could provide a large lift. But it certainly wouldn't scale well with better shooting teammates. And my understanding was that RAPM accounts for teammates. That is, a scorer won't look better for playing with inefficient teammates.

I like RAPM as a stat but the results sometimes conflict with what I think is (deductively) necessary for high offensive impact. So, something has to give, but I don't know which.

Well, the word we tend to use for +/- stats is "impact", or sometimes "lift". It's about what your team can do with you that they can't do without you. I believe what you're pointing out is that LeBron's ORAPM in Miami was a bit soft - by his standards, not by any normal standards - and well, that was him playing on a team with Wade & Bosh which was never able to crossover into GOAT territory, but was pretty capable without him.

[CommonerCoffee]

C.) Therefore, the best overall teams, generally speaking, are those that have the largest separation between their ORTG/DRTG (GSW and SAS this season). The caveats above still apply - you want to be consistent. A team with +10 differential that plays like a +10 against all teams is better than a +10 team that plays like a +6 against some opponents and +14 against others.

Firstly, incredible post. You're my favorite poster on the boards and things like above are why.

I suspect you do already know this most of what I'm about to write but I feel like it would be good to say anyways. I feel like consistency is wildly overrated or at least its value drastically varies with the mean quality of the team and specifically how their performance varies. In the above situation if the +6 is against the bottom feeders of the league and the +14 is against the true contenders, then that team would be undoubtedly far superior to the uniformly +10 team. The discrepancy could even be explained by varying effort level. And even if not, given the format of the playoffs, it's far more valuable to be strong against strong teams.

Similarly, consistency should only be valued by teams with top expected performance levels, as they want to minimize their variance so they won't be losing to lesser teams. The second tier of teams would want the higher variance as that would give them the best chance to beat better teams.

Toy example: Team A performs at a normally distributed SRS level with mean +10 and some variance. Team B performs at mean +8 and some variance. And say at any given game, the team performing at the higher SRS level prevails. Lets say the variance of both team is 2. Then via simulation A wins 0.759958 of the time. Lets say Team A is consistent and operates at a lower variance of 1. Then Team A wins 0.814091. And if we assume Team B is inconsistent and performs at a higher variance of 3, then a Team A with variance 2 wins only 0.71016.

Of course we primarily talk about top teams, so it's natural that consistency has a positive connotation and your original point was concerning top teams after all. However, I've seen inconsistent and bad conflated here so many times I think this was still worth writing.

[SideshowBob]

RPM updated

Draymond and Lebron continue to establish a gap over Paul/Curry. Draymond now a tenth of a point ahead of Lebron and over +9, Lebron climbing incrementally on both Off/Def and is just under +9. Curry/Paul/Leonard all within a tenth of a point of each other but Curry has dropped below Paul now - as Dray gets more ON court time w/o Curry and the Warriors continue to dominate, the model is observing separation between the two.

[yoyoboy]

Anyone know when we'll get updated RAPM data? Since Curry's box-score stats were a lot better than LeBron's this season, yet LeBron has established a decent lead in RPM, I'd expect his RAPM lead to be even more significant. Of course, Green will undoubtably have a sizable margin over everyone.

[bondom34]

https://docs.google.com/spreadsheets/d/1_JkGXDXtu3cSLzplkQo4mDYOMLi-Sh2YfIb9nKeMNZI/edit

Crow posted an update.

[SideshowBob]

https://docs.google.com/spreadsheets/d/1_JkGXDXtu3cSLzplkQo4mDYOMLi-Sh2YfIb9nKeMNZI/edit

Crow posted an update.

Nah, that's the data from the end of the RS

JE updated NPI RAPM Single-Year and Multi-Year on APBR through the end of the regular season.

Single-Year - Link
Multi-Year - Link (less weight to older seasons, 3 seasons total)

---------------------------------------------------------------------------------

Single-Year

Top 10 (O | D)

1. Green +7.93 (4.51 | 3.42)
2. Curry +6.16 (4.65 | 1.51)
3. James +6.07 (4.21 | 1.86)
4. Leonard +5.17 (1.93 | 3.24)
5. Durant + 4.79 (3.99 | 0.79)
6. Paul +4.46 (2.94 | 1.53)
7. Thompson +4.26 (2.92 | 1.34)
8. Lowry +3.91 (2.58 | 1.33)
9. Korver +3.54 (1.20 | 2.34)
10. Westbrook +3.45 (3.15 | 0.31)

Top 5 Offense

1. Curry +4.65
2. Green +4.51
3. James +4.21
4. Durant +3.99
5. Westbrook +3.15



Top 5 Defense

1. Green +3.42
2. Leonard +3.24
3. Snell +3.22
4. Adams +2.91
5. Duncan +2.78

---------------------------------------------------------------------------------

Multi-Year (less weight to older seasons, 3 seasons total)

Top 10 (O | D)

1. James +7.65 (5.61 | 2.04)
2. Green +7.64 (3.25 | 4.39)
3. Curry +7.47 (5.98 | 1.49)
4. Paul +7.22 (4.77 | 2.45)
5. Durant +6.18 (4.77 | 1.41)
6. Leonard +6.04 (2.00 | 4.04)
7. Rubio +5.38 (2.92 | 2.46)
8. Harden +4.87 (5.04 | -0.17)
9. Korver +4.86 (2.76 | 2.10)
10. Ginobili +4.67 (2.232 | 2.35)

Top 5 Offense

1. Curry +5.98
2. James +5.61
3. Harden +5.04
4. Paul +4.77
5. Durant +4.77

Top 5 Defense

1. Green +4.39
2. Leonard +4.04
3. Cousins +3.35
4. Kidd-Gilchrist +3.35
5. Gobert +3.11

---------------------------------------------------------------------------------

Green/James barely edging out Curry for top spots in multi-year. Paul not far behind.

Curry tops in offense both sets, but still lags behind Green overall in both sets.

Curry/James nearly dead-even in single-year, in line with their single-year RPM on ESPN

Davis is a small positive in S-Y

Irving is a negative in both sets

You'll notice some differences once a full RS/PS set comes out (James caught up with Green, Curry down namely).

[kayess]

Given all the Draymond/Curry talk, if we were to "knock" or mentally adjust the scores in our head, it'd be adjusting Draymond's O/Curry's D, right?

It's hard to believe Curry's not head and shoulders above even the +5 guys in Durant/Kawhi, but oh well. Will most likely still be the GOAT RS to me, hopefully he can still give 2000 Shaq a run for the the GOAT peak

[SideshowBob]

I don't know anything about this guy though. It's entirely possible he's basically a kid desperately trying to be taken seriously by elevating himself above something else.

Well he's at the presser tonight and he's no kid. Looks mid-60s easily.

Haven't had the time to get back to him yet but his initial response was pretty civil

[SideshowBob]

RPM Updated 5/24

Lebron took the RPM lead as his offense and defense have both gradually been on the rise as the playoffs persist. I think its likely he's eating into Draymond's S-Y RAPM lead as well (JE should post a final set in June).

Russ's box-score dominance ((playoff BPM was at +12.3 before tonight's game) + OKC wrecking GS should coincide with him rising as well.

[Clyde Frazier]

I don't know anything about this guy though. It's entirely possible he's basically a kid desperately trying to be taken seriously by elevating himself above something else.

Well he's at the presser tonight and he's no kid. Looks mid-60s easily.

Haven't had the time to get back to him yet but his initial response was pretty civil

I knew "Cleveland Scene" sounded familiar when I was watching that! Wasn't sure if it was the same guy who wrote the abhorrent article you posted here. He made that awful joke about asking lue if he wanted to comment on the officiating like casey did. Have to say, you have a lot more patience than I do with people like that... commendable

[SideshowBob]

Updated 5-Man Unit RAPM via JE Twitter

SBDS has dropped a bit. Love/James/Smith/Thompson + either Kyrie/Delly now both top 5. OKC rising towards top spot.

[Doctor MJ]

Updated 5-Man Unit RAPM via JE Twitter

SBDS has dropped a bit. Love/James/Smith/Thompson + either Kyrie/Delly now both top 5. OKC rising towards top spot.

Makes me think we should have a "Lineup of the Year" award.

[SideshowBob]

RPM updated (through CF). Lebron now with an established lead over everyone (and only 1.2 points behind Leonard defensively).

Curry slight lead over Green again.

[kayess]

RPM updated (through CF). Lebron now with an established lead over everyone (and only 1.2 points behind Leonard defensively).

Curry slight lead over Green again.

NPI and PI RAPM should be updated pretty quickly after the finals, right?

[SideshowBob]

RPM updated (through CF). Lebron now with an established lead over everyone (and only 1.2 points behind Leonard defensively).

Curry slight lead over Green again.

NPI and PI RAPM should be updated pretty quickly after the finals, right?

Yeah probz. This is the most engaged with the playoffs I've seen him on twitter so he might be on the ball w/regards to timeliness.

[SideshowBob]

Poor performance of the death lineup is going to help boost the Cavs players.