|
|
Post by eric on Mar 14, 2015 18:20:19 GMT -6
PER is Player Efficiency Rating, developed by John Hollinger as a way of normalizing production per minute: so if one player scores 8 points per game in 12 minutes per game, we can see he's better than someone who scores 16 points per game in 40 minutes per game. It also normalizes for possessions per game, but it does not (directly) normalize for possessions USED per game. Is this a problem? Let us consider the 522 player-seasons from the last two full seasons to amass at least 1000 MP and see if there is a correlation with USG rate:  It's a problem. Let's try just subtrating USG rate from PER. We expect that our average will be 15 - 20 = -5, and in fact...  ...it is! (-5.19, anyway.) Look how much better this matches the normal distribution than unadjusted PER...  ...and statistical measures confirm this. The root mean squared error of unadjusted is 4.74, the same value for PER-USG is 3.38, a dramatic improvement. . Who cares? Well, consider the top five in PER so far this season. 31.4 Anthony Davis 29.8 Russell Westbrook 27.5 Stephen Curry 26.6 James Harden 26.1 LeBron James Now look at them in PER-USG. +3.6 Davis -8.6 Westbrook -1.2 Curry -4.3 Harden -7.0 James Westbrook suddenly doesn't seem so good, right? And maybe LeBron actually is in decline, but is keeping his per-game stats pumped with a slight bump in usage (as most every superstar before him has done). Bottom line, the MVP is obviously Curry, it's been Curry for months now, and any journalist who tells you Westbrook has better stats is a liar or ignorant. |
|
RWisoursavior
Former GM 
Posts: 1,808
Likes: 128
Joined: June 2014
|
Post by RWisoursavior on Mar 14, 2015 18:23:16 GMT -6
Give it to AD!
|
|
Bankz
Former GM
Posts: 7,254
Likes: 895
Dump Bucks: 18,475
Joined: April 2014
|
Post by Bankz on Mar 14, 2015 18:36:45 GMT -6
Seems like Anthony Davis should be MVP
|
|
RWisoursavior
Former GM
Posts: 1,808
Likes: 128
Joined: June 2014
|
Post by RWisoursavior on Mar 14, 2015 19:42:59 GMT -6
|
|
|
|
Post by eric on Mar 15, 2015 11:19:32 GMT -6
Re: Davis as MVP
He's already missed 12 games, and projects out to a ceiling of 2491 minutes played if he maintains his season average in every remaining Pelican game. This would be the lowest MP total of any MVP in the sportswriter era (1981-present). All but one have been above 2700 minutes, and he was on the overall 1 seed.
His team stinks, and almost certainly won't make the playoffs. (Hollinger odds has them at 30% now, and that doesn't factor in Durant coming back from injury.) A sportswriter MVP has never missed the playoffs. All but one have been a 3 seed or higher, and he played 3402 minutes.
His PER is historically great, but that tells you more about PER than it does about him. He trails Curry in all overall metrics obviously because Curry has played 10% more minutes, but he trails him in most rate metrics too: WS/48, BPM, WP/48. How do we know which advanced metric is best? That's an issue I'll go into greater detail in another post, but for a good fast read you should go WS, WP, BPM, PER in that order, and the first two are way more important than the others.
tl;dr: Davis doesn't have the best stats. He's not a great defender. Thus he's not the best player. And he shouldn't be MVP.
|
|
|
|
Post by eric on Mar 15, 2015 11:25:56 GMT -6
Simply fabulous. The movement stuff is the next frontier (first broached by Zach Lowe's Ghost Raptors), and it's great to see people exploring it. |
|
Bankz
Former GM
Posts: 7,254
Likes: 895
Dump Bucks: 18,475
Joined: April 2014
|
Post by Bankz on Mar 15, 2015 13:51:57 GMT -6
Re: Davis as MVP He's already missed 12 games, and projects out to a ceiling of 2491 minutes played if he maintains his season average in every remaining Pelican game. This would be the lowest MP total of any MVP in the sportswriter era (1981-present). All but one have been above 2700 minutes, and he was on the overall 1 seed. His team stinks, and almost certainly won't make the playoffs. (Hollinger odds has them at 30% now, and that doesn't factor in Durant coming back from injury.) A sportswriter MVP has never missed the playoffs. All but one have been a 3 seed or higher, and he played 3402 minutes. His PER is historically great, but that tells you more about PER than it does about him. He trails Curry in all overall metrics obviously because Curry has played 10% more minutes, but he trails him in most rate metrics too: WS/48, BPM, WP/48. How do we know which advanced metric is best? That's an issue I'll go into greater detail in another post, but for a good fast read you should go WS, WP, BPM, PER in that order, and the first two are way more important than the others. tl;dr: Davis doesn't have the best stats. He's not a great defender. Thus he's not the best player. And he shouldn't be MVP. According to your OP he has the best adjust PER and that was what you based the MVP on.... And to say Davis is a not a great defender and some how imply that Curry is.....  #Davis4MVP #ChicagoFinest #ComeHomeBaby |
|
|
|
Post by eric on Mar 15, 2015 15:36:44 GMT -6
According to your OP he has the best adjust PER Yes. and that was what you based the MVP on.... No. And to say Davis is a not a great defender and some how imply that Curry is..... They're both average defenders. :dislike: |
|
RWisoursavior
Former GM
Posts: 1,808
Likes: 128
Joined: June 2014
|
Post by RWisoursavior on Mar 15, 2015 19:59:52 GMT -6
Simply fabulous. The movement stuff is the next frontier (first broached by Zach Lowe's Ghost Raptors), and it's great to see people exploring it. Pretty amazing how flabbergasted Zach Lowe is to talk about watching ghost Lebron. |
|
|
|
Post by MJ on Mar 16, 2015 16:19:19 GMT -6
Want Pelicans to make the playoffs so AD has 4-5 games on a national stage. He's fucking amazing
|
|
