RPM seems like a fun stat.
It shows Jae Crowder as the 23rd best player in the league, Jared Sullinger as the 36th best player in the league, Amir Johnson as the 41st best player in the league... yeah ok.... According to that stat, all those guys are better than Thomas... and Boston apparently has 5 guys in the top 50 (Kelly Olynyk as the 48th best player in the league).... neat.
It does show Marcus Smart as the 313th best player in the league though... so there's that.
FYI, last year Jrue Holiday (while playing injured) had a RPM of 3.22 (10th amongst point guard)... Thomas had a RPM of 1.45 (20th). Holiday was better than him the year before as well.
See, this is the issue with the RPM stat. It's not the fact that the stat is inaccurate, it's the fact that people do not know how to interpret it.
The RPM stat doesn't show that Crowder is the 23rd best player in the league. It shows that he has the 23rd greatest positive impact on his team when he is on the court.
It indicates that Boston are +3.85 Points better (relative to their opponent) when Crowder is on the court. I.e. the Celtics are a better team when Crowder is on the court.
That doesn't mean Crowder is a superstar player - it just means that when he gets on the court he makes a major positive impact. For anybody who has watched most of the Celtics games this season, I think it's easy to confirm (by the eye test) that this is accurate. We
do play better when Crowder is on the court.
That is also true of Isaiah Thomas, Kelly Olynyk, Amir Johnson. We almost always play better as a team when those guys are on the court.
The valuable thing about RPM is that it's one of the only statistic that takes into account a players outright impact on a team - not only the parts that are easily measurable (points, rebounds, percentages, etc) but also the intangibles that are hard to measure.
For example there are some players out there who might not put up a lot of points, grab a lot of rebounds, or score especially efficiently. But they might be outstanding vocal leaders or great floor generals. Guys who set amazing screens. Guys who don't have great steal or block numbers, but are great at playing positional team defence. Guys who are so intimidating in the post that their mere presence scares opponents from attacking the basket.
There are a lot of ways guys can impact a game that can't be captured with basic stats. Amir Johnson is a perfect example. His statistics are utterly average and they have been his entire career, but he has been highly renowned his entire career as a guy who impacts games far more than his stats show. If you watch his games as a Celtic, you can see that. If you look at his RPM stats, they reflect that.
Amir Johnson has only had one block yesterday, and yet Marc Gasol (a guy universally seen as one of the top centers in the NBA) was completely incapable of making a basket while Amir Johnson was defending him. Watching the game, it was plain to see that he had a huge impact when he was on the court.
You say Boston has 5 players in the top 50, as if that is surprising. Why would that be surprising at all? Boston is a team that does not have a single All-Star, and yet has been competing for a playoff spot in the East all season long. How could they possibly be as competitive as they are, with no star power, unless they had a handful of guys who impact the game more than their stats show?
The RPM stat might be legitimately questionable when used for small sample sizes - like for example, players who only play 10 games in a season, or players who only play 7 MPG.
But when you're looking at a guys who play significant minutes over the course of a season, it becomes statistically relevant.
I think guys like Kelly Olynyk and Kevin Garnett are perfect examples of why the RPM stat is valuable.
Olynyk is a very unselfish player, who is often criticised for not being aggressive enough as a scorer. While this is a fair criticism, Olynyk is also a player who has incredibly high basketball IQ, has very high work ethic, and is very active in all other aspects of the game. When Olynyk is on the court, most of the good things that happen seem to be a direct or indirect result of something he has done. For example he'll pump-fake on the perimeter then drive past his defender, the defense will collapse on him, he'll kick it out to a shooter, who then makes a second pass to another shooter, who hits the open three. All made possible because of that initial fake-and-drive by Olynyk, but he gets no recognition for it on the box score. Plays like this happen all the time when Olynyk is on the court, which is why Boston almost always plays much better when he is playing, even if his actual box score statistics are rarely anything to write home about.
Explains why Olynyk has had a positive RPM this year (+2.11), and last year (+1.71) but not in his rookie year (-1.33) where he was clearly often very lost on the defensive end.
Likewise it explains why a guy like Kevin Garnett can still have an elite DRPM (+4.49) this year, despite the fact that he is clearly a shadow of his former self in terms of his outright box-score production. KG's individual defence was always just part of the reason for why he had so much impact on that end of the court. He's always been famously known for his vocal leadership on that end of the court - always telling guys where to be on defence, etc. Again that is something that regular stats will never capture, but a stat like DRPM can.
That, is why I am so in love with the stat. I like getting insight into a players overall contribution/impact to a team's ability to win games, rather than just seeing the obvious things that are clearly visible on face value.