Right, I was referring to the rankings specifically for the RAPM stat. My point wasn't that it's ALL wrong, but that if you give any weight to a stat that can jump to conclusions as silly as Amir Johnson > Love, the whole thing gets thrown off (even though Love and Johnson both got their rankings in RAPM written off as outliers).
RAPM is a multi-year stat. Given that Love was drafted in 2008, I'm not suprised his RAPM is low if a portion of it come from his first couple of years. +/- type stats are rarely kind to rookies and young players. Being on terrible teams with goofy rotations also tend to do goofy things to your +/- stats, even if you try to "adjust" for it. I took a look at his year to year stats, it seems it was positive this season, but rather bad in the 2 prior years. http://basketballvalue.com/teamplayers.php?team=MIN&year=2010-2011
A list with Yao Ming at #22 you can take seriously. Do you seriously believe Nene Hilario is better than Rose? The NFL list was much better because it cared little about numbers since the players were voting on it.
Numbers are too true. So if a player is playing hurt or injured and his play suffers because of it, but he's still gutting it out because he knows his team needs him the numbers don't take that into account. If a player has shot horribly for the game but in the final minutes leads his team to the W the numbers don't see that. The players do. They remember that stuff, I know I do. I remember when Rose decided that he was going to be Reggie Miller and drained a few 3s to beat the Rockets. Although the numbers don't say he's a good 3 point shooter, that didn't matter at the time. Who is taking Lamar Odom over Carmelo Anthony? Who is taking Greg Oden over Joakim Noah or Marc Gasol? Matt Bonner and Ryan Anderson over Scola? Really now? Not saying that all numbers are BS, but they should not be the final say.
I did but there is a reason I bolded part of your post because that specific part seemed to be saying 'Ha ha, look at Rose at 17! Overrated!' and that was the part I was responding to. I pretty much agree this list is trash.
Even if you feel the relative ranking of Love and Johnson was distorted by the inclusion of RAPM (which they weren't because those results were outliers), that doesn't mean the overall rankings aren't made more accurate on average (which would be the goal).
Yes I think he's overrated and by no means efficient or helping his team as big as some say he does, advanced stats actually prove that he's one of the most unefficient players on the Bulls roster. He takes a lot of shots and got his MVP because the team did very well and he had great highlights/numbers(which don't tell anything about efficiency). Of course he's a better player than 17, I think the list is garbage because it only focuses on questionable stats that mess up the whole ranking. I didn't want to diss Rose, just waited for the annoying fanboys you see this year that think he's the best PG in the game.
Deke is on the list. Is he coming out of retirement? lol Lol @ Rashad McCants being rated in the 300's.
Look at the top 50. For 42 of them, RAPM is an outlier. I understand that in theory more metrics = better average, but that's only true if the metrics are representative of the value that your average is trying to measure. If I included a metric that based their skill as a player on just their body mass index and age, that would hurt, not help, the ratings system. RAPM isn't THAT exaggeratedly bad, but when it only falls in the same RANGE as other metrics 8 out of 50 times, why should I trust it when it tells me that Nene is better than Chris Bosh or Paul Pierce (a judgment which DID get counted)?
Nene getting a higher rating than Bosh/Pierce isn't that crazy, IMO. But I hear you. RAPM is noisy (though significantly less noisy than some other variants of APM). The thing in its favor is that it can capture parts of the game that simply aren't measured by the other metrics being used. Incorporating RAPM is an attempt to come up with a more comprehensive rating which incorporates both the "measurables" and the "immeasurables", but in doing so yes there's a chance that precision is lost. It comes down to this -- would you rather have a rating that's fairly precise, but only takes into account specific things (the "measurables"), or a rating that's more all-encompassing but less precise? Either way, one shouldn't rely exclusively on it.
