Deadspin: CP# defines limitations of adv stats Topic

In broad terms, the Chris Paul Problem—there is one!—is as follows: Pretty much everybody agrees that Paul is, and has been for the past decade, one of the NBA’s best players. He made nine straight All-Star games from 2008 to 2016; he has made eight All-NBA teams and nine All-Defense teams; he is the league’s active leader in both assists and steals per game. He is a superstar.

Basketball analytics say he is even more than that. According to Basketball Reference, Paul has the sixth-highest career Player Efficiency Rating (PER) of any player ever; Win Shares Per 48 Minutes (WS/48) essentially cannot tell the difference between him (.2504) and Michael Jordan (.2505); Box Plus-Minus (BPM) says only Jordan and LeBron James are his superiors. His career 122.69 Offensive Rating (OffRtg) is the highest of all time. The judgment of the sport’s holistic metrics is that he is one of the best basketball players who ever lived.

Advertisement

And yet! Paul’s teams, first in New Orleans and then in Los Angeles, never went very far in the playoffs—he has never so much as reached the conference finals—and mostly didn’t even put up all that ferocious a fight. That’s not supposed to happen in the NBA, a star-dominated league in which the team with the best player usually wins, and in which other teams with players as good as Chris Paul is supposed to be have pretty much never lost as consistently as his teams have.

A little over a month ago, Paul used his leverage as one of the NBA’s biggest stars to engineer a blockbuster sign-and-trade sending him from the Clippers to Houston, where he’ll join fellow superstar guard James Harden on a Rockets team that won 55 games, the West’s third playoff seed, and a first-round playoff series this past season. At least as far as basketball’s holistic metrics are concerned, a strong regular-season team just added one of the best players of all time to its roster. When a historically great player—what the analytics say Chris Paul is—changes teams, it should change history, but nobody really expects this to. This is the Chris Paul Problem.

8/1/2017 2:27 PM
http://deadspin.com/the-chris-paul-problem-is-an-analytics-problem-1797409440?utm_campaign=socialflow_deadspin_facebook&utm_source=deadspin_facebook&utm_medium=socialflow
8/1/2017 2:27 PM
This is pretty disappointing. I thought it was going to be about the advanced stats the sim uses in its engine and how they are a bad idea.

Blk%, for example, doesn't mean someone is particularly good at blocking shots. It usually means that other people on the team aren't that good at it.

Even the rebounding %s aren't that accurate. Kevin Love is not a great rebounder, at least not to the extent his numbers in Minnesota said. Same with Dwight Howard and even Russell Westbrook. The numbers are - to an extent - a reflection that other people on the team aren't great rebounders or, as in the case with the three individuals I named, not even trying to get the rebound themselves. I don't know how many times I've seen one of Howard's teammates or Westbrook's teammates stare at a ball bounce off the ground and actually move out of the way so that Howard or Westbrook could get the rebound. Same thing when Love was in Minnesota. He goes to Cleveland and you see guys not deferring to him so that he can pad his stats; you see people like Tristan Thompson working hard to get every board he can. That doesn't mean these guys weren't or aren't great rebounders, it just means they aren't Moses Malone, Dennis Rodman or God help me, Bill Russell (whose fabricated numbers are lolable)
8/1/2017 4:40 PM
(but yeah, I see what the article means)
8/1/2017 4:40 PM
yes maybe I should have posted this in the NBA general thread
8/2/2017 11:34 PM
But...Love's dreb% is still north of 27%. On a team with much more athletic teammates than in Minnie.
Hard to argue he is not a very good rebounder.
8/3/2017 6:30 AM
Offensive rating lol. It lists Steph as #25 all time, while Deandre is #5. Stockton is like #4. Let CP3 have Karl Malone and he gets to the Finals.
8/3/2017 9:55 AM
Ash, imo blk% isn't really influenced by teammates -but I could see reb% being influenced by teammates-.
8/3/2017 1:52 PM
Posted by ncmusician_7 on 8/3/2017 1:52:00 PM (view original):
Ash, imo blk% isn't really influenced by teammates -but I could see reb% being influenced by teammates-.
A guard that averages 1 block per game on a team with no other shot blockers will have a much higher blk% than another guard averaging 1 block per game on a team with someone the likes of Eaton, Bol, Hakeem, Admiral, Mutombo, Zo, etc because of the way blk% is figured. Much, much higher.
8/4/2017 12:21 AM
Posted by ashamael on 8/4/2017 12:21:00 AM (view original):
Posted by ncmusician_7 on 8/3/2017 1:52:00 PM (view original):
Ash, imo blk% isn't really influenced by teammates -but I could see reb% being influenced by teammates-.
A guard that averages 1 block per game on a team with no other shot blockers will have a much higher blk% than another guard averaging 1 block per game on a team with someone the likes of Eaton, Bol, Hakeem, Admiral, Mutombo, Zo, etc because of the way blk% is figured. Much, much higher.
According to basketball-reference.com, blk% is blocks by the player vs number of opponent 2 point shots. In ur scenario, if the 1 blk/g is per 100 two point shots, then it's 1% regardless of what the teammates do.
8/4/2017 8:32 AM
Yeah the above is inaccurate. blk% is independent of teammates.
8/4/2017 10:56 AM
Posted by ncmusician_7 on 8/4/2017 8:33:00 AM (view original):
Posted by ashamael on 8/4/2017 12:21:00 AM (view original):
Posted by ncmusician_7 on 8/3/2017 1:52:00 PM (view original):
Ash, imo blk% isn't really influenced by teammates -but I could see reb% being influenced by teammates-.
A guard that averages 1 block per game on a team with no other shot blockers will have a much higher blk% than another guard averaging 1 block per game on a team with someone the likes of Eaton, Bol, Hakeem, Admiral, Mutombo, Zo, etc because of the way blk% is figured. Much, much higher.
According to basketball-reference.com, blk% is blocks by the player vs number of opponent 2 point shots. In ur scenario, if the 1 blk/g is per 100 two point shots, then it's 1% regardless of what the teammates do.
True.

btw, WIS doesn't do it that way, but the way they do it is still in line with the same reasoning.
8/5/2017 4:51 PM
Deadspin: CP# defines limitations of adv stats Topic

Search Criteria

Terms of Use Customer Support Privacy Statement

© 1999-2024 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.