Doctor MJ wrote:BoatsNZones wrote:He leads by a 23% margin in RPM wins. That's fair to call it as a country-mile stat wise, no? Lebron has a miniscule lead on him on RPM per-game. Gobert is a solid 3rd there. Curry is 1st in offensive EPM as well. He's 2nd in VORP/BPM to Joker. He's right there. Jokic is my MVP (feel like I need to say it for the 35th time this thread), but Curry is catching up.
I have to say, as someone who has been using +/- stats seriously for 15 years, I really struggle to take RPM at all seriously. I had philosophical issues with it from the jump, but the way they just keep changing the formula, resulting in glaring changes, without ever explaining exactly what they're doing is a) a major problem in its own right and b) characteristic of Englemann's lack of basketball analysis common sense that he's always had.
The stat I'm using right now along these lines - and granted, that could change - is bball-index's LEBRON.
https://www.bball-index.com/2020-21-lebron-data/
By that stat, if you go by Wins Added, the leaderboard is:
1. Jokic
2. Gobert
3. Giannis
4. Steph
That ordering feels pretty reasonable to me.
Thank you for putting this into words. Englemann deserves credit for helping bring RAPM into the mainstream, but ESPN's RPM was a black box from the beginning. I can understand Englemann wanting to cash the ESPN checks, and I can understand ESPN not releasing the formula to keep their stat all shiny and clean (conveniently avoiding scrutiny for the formula), but not releasing the formula means it's not a stat you can trust.
I'm not nearly as well versed in RAPM as some, but I understand the basics of it, including that while it can do a good job of sorting players in a general way, an original premise was that it required multiple seasons of data to reduce noise properly and the specific spot on the list where players wind up shouldn't be used as Undisputed Truth when assessing players near each other on the list within a single season. It's another piece of information to add to the mix.
As for Englemann, I don't know him to say how he works, but your "common sense" comment resonates with me. It seems like he logically worked out a statistical system, and his main trust is in the logic of that system, but faced with certain weaknesses, such as, say, a role player being ten spots above an all-NBA player, his tendency is to address that from a purely statistical standpoint rather than through a basketball-educated lens.
LEBRON, on the other hand, appears to have a basketball-educated guiding influence baked in. They identify players in the listing by offensive archetype and include an adjustment for role in the formula. I haven't done the deep dive to see how that adjustment works, but based on the listing, it appears to have weeded out the role player outliers that Englemann still has. (I do wonder why they identify Curry as a primary ball handler when Doncic, Lillard, LeBron, Chris Paul, Harden, and Westbrook are all listed as shot creators, but probably not a big deal.)
As a Curry fan, it would be easy for me to point to the Englemann ESPN RPM since Curry tops that list by a sizeable margin when sorted for wins, but the LEBRON ranking makes more sense. I can make an argument (and feed my fandom) to push Curry above Giannis and Gobert, but there's no way I can rationally make an argument for anyone other than Jokic as the clear leader. LEBRON gets that right, while RPM doesn't.
