
Background
In my most recent article, I described the process of calculating regularized adjusted plus-minus (RAPM) over a 25 year span, specifically from the 1996-97 season to the 2020-21 season.1 You can check out a full interactive graph of the results here and a filterable table in the article. In summary, though, the results suggested that LeBron James led the league in on-court impact since 1997 with Kevin Garnett, Chris Paul, Stephen Curry, and Tim Duncan rounding out the top five among eligible2 players.
While high-end sustained impact is certainly commendable, I wanted to use the same data to look into shorter stretches of time. Peak versus longevity is a common point of debate in historic basketball discussion. A guy like LeBron James has an all-time great peak along with incredible longevity, but what a player like Shaquille O’Neal? His longevity wasn’t bad by any means, but his peak was certainly even more impressive. A full analysis should look for the best peaks in our timeframe from 1997 to 2021.
Thus, I split the data into five year spans between the 1996-97 season and the 2020-21 season and I calculated the RAPM for each of these spans (21 in total) using the same methodology as the most recent article (that means playoff possessions are double weighted). On to the results!
Results
There’s a lot of data to look at here and it’s impossible to get to it all. If you’d like to dive into it yourself, you can check out a full interactive graph here. There will also be a searchable table at the bottom of the article.
I previously mentioned top five players in RAPM since 1997: LeBron, Garnett, Paul, Curry, and Duncan. Well, these five players are the only ones who make multiple appearances in the top 30 five year RAPM peaks since 1997. Here they are.
The top right is where you want to be – elite impact on high volume. As expected, that’s an area in which LeBron thrives.
It’s a bit hard to appreciate other players when these five guys are hogging all of the spotlight. Thus, let’s take a look at a graph of each player’s single highest peak (by highest RAPM). Obviously longevity matters so this isn’t a perfect measure, but I think it’ll allow us to check out some elite peaks from players that haven’t been mentioned.
From 2012 to 2016, LeBron James won two MVPs and three Finals MVPs. He reached the NBA Finals five times (with three wins), had one of the greatest single-season peaks in NBA history (2013), and he led the greatest Finals comeback in NBA history (2016). He also apparently had the best five year RAPM peak on record. Surprise.
From 2003 to 2007, Garnett won one MVP and was named to the All-NBA First Team twice. His Timberwolves even missed the playoffs thrice in this span with a sole WCF appearance in 2004. It’s one of the least “accomplished” five year peaks on this list. It’s a real shame that Garnett was not surrounded by better supporting casts at his peak like other players.
From 2001 to 2005, Tim Duncan won two MVPs, two Finals MVPs, and two NBA titles. He was named to five All-NBA First Teams and four All-Defensive First Teams. Duncan’s 2003 season was particularly impressive, as the league MVP carried the San Antonio Spurs to a Finals win despite lacking for help at times. Duncan’s dominant Game 6 in the 2003 Finals showcased his versatility and defensive dominance, putting up an insane line of 21/20/10 with eight blocks.
After a spot on the All-NBA Second Team in 2014, Steph Curry went on a historic four year stretch from 2015 to 2018 which included two MVPs, four Finals appearances, and three Finals wins. While I did not split this version of RAPM into offensive and defensive components, one would expect Steph’s offensive RAPM to top the charts.
Dwyane Wade’s five year stretch from 2006 to 2010 is actually higher than you might expect – right above CP3’s 2012-16 peak (albeit by a negligible margin). Wade started off this five year stretch with a dominant 2006 season, averaging 27 PPG in the regular season and carrying the Heat to a Finals win over the Mavericks. Wade averaged 35/8/4 in the 2006 NBA Finals and brought home Finals MVP. While that was his only championship during this run, Wade was named to his first All-NBA First Team in 2009 and did it again in 2010 after injury-riddled seasons in ’07 and ’08.
Chris Paul was named to three All-NBA First Teams and two All-NBA Second Teams during this stretch from 2012 to 2016, but never even reached the Western Conference Finals. While Paul was individually dominant, the Lob City Clippers could never find their mojo in the playoffs, often due to untimely injuries to either Paul or Griffin – or even both.
For Michael Jordan, his 1997-2001 peak is actually limited to just the 1997 and 1998 seasons because he retired following the 1998 Finals. Nonetheless, his two year stretch included two Finals wins, a league MVP, and two Finals MVPs. Not bad.
I’m pretty surprised that Dirk’s RAPM peak is considered a stretch from 2010 to 2014 (it’s basically tied with his 2008-12 stretch, but still). I’m assuming his legendary 2011 run has an impact on that. Still, Dirk was undoubtedly a great player who has arguably become underrated in recent years.
The early 2000s featured incredibly high level of play from both Duncan and Garnett, arguably the two greatest power forwards in the history of the game. Then LeBron took the throne as the best player in the league, a title he may have held longer than any player in the history of the sport. That’s ridiculous dominance.
I didn’t set a minimum limit for possessions played (maybe I should’ve) which is apparent when you see Jordan’s small red point at the start. But the second leading player from both ’97-’01 and ’98-’02 was Shaquille O’Neal, so shoutout to his dominant peak with the Los Angeles Lakers.
What about Kobe? Kobe’s best RAPM peak was from the 2006 season to the 2010 season. In this span, Kobe led the Lakers to two Finals wins and three Finals appearances along with an MVP win in 2008. Kobe’s 4.26 RAPM was the fifth highest during this span: well behind LeBron, Wade, and Garnett, but only narrowly behind Duncan.
Finally, here’s a full table of the results.
A few last observations I’d like to make:
As always, remember that this is not meant to be a player ranking – it’s more of a fun mathematics & statistics project.
I had initially planned on wrapping up this series with an article dedicated to 1997-2021 playoff RAPM. However, I didn’t think it warranted its own article or much discussion because of the high variance in the results. You can check out an interactive graph of the results here if you’re curious, though.
- The data includes regular season and postseason matches, but with very limited data from the 1996-97 regular season specifically. The rest of the data (beginning with the 1997 postseason and ending with the 2021 Finals) is almost entirely complete. No data is available prior to the 1996-97 season, which is why the cutoff was chosen.
- Without any requirement for possessions played, Joel Embiid’s RAPM is actually the second-highest behind only LeBron. Of course, his sample size is far lower with less than 42,000 possessions compared to an average of over 200,000 among the five aforementioned players.
This is some amazing work. Congratulations.
Any reason why you wouldn’t post single season peaks? Even with prior information, is the data too noisy to be reliable on evaluating single season RAPM’s?
Thanks!
Thanks for the kind words. Yes, more informative priors would be necessary to reduce the significant noise of single-season RAPM. Metrics designed for single-season evaluation like LEBRON, RAPTOR, EPM, etc are a better option there.