That's a good point. The Lakers are a brand (like every team), and they are going to be more successful in the post-Kobe years if they can land another once-in-a-generation player like him. The Lakers is a tradition. Whereas nobody's signing with the 76ers (or my Zers) "to be remembered."

Fucking hell, does the NBA have a salary cap?

Tracking data shows James Harden still a bad defender. Original paper. I've seen other stats that said Houston's defense was due to fall off because they're giving up a lot of open looks (particularly from long-range), but teams were just missing.

Edit: Never mind. I think this data is from 2013-2014.

Can anyone with an Insider account post these two for me?

http://insider.espn.go.com/nba/story/_/ ... m%2F%22%7D

http://insider.espn.go.com/nba/story/_/ ... point-line

at some point recently (within the last month or so?), NBA.com put up Synergy's stats and, even better, its all in sortable mother fucking tables. i'm about to straight up believe in God.

Some interesting ways of looking at offenses in this post.

For the time being, I want you to ignore that I'm posting another Sixers' article and actually ignore everything this article says about the Sixers (though read it if you're interested, of course). Looking at some of these metrics, I wonder if they would correlate to winning, if you removed the Sixers from consideration. I'll post the stats here.

Secs/Poss
1. GSW 13.4
2. BOS 14.3
3. PHO 14.4
4. OKC - 14.5
5. MIL - 14.6
6. HOU - 14.6
7. DEN - 14.6
8. LAC - 14.7
9. PHI - 14.8
10. DAL - 14.8

I have no idea what difference the half second that separates 2nd from 10th means (seriously, #11 is at 14.9; the slowest team is Utah at 16.2), but Golden State absolutely stand out for being nearly a second faster than the next fastest team.

Seconds Between Passes
1. UTA - 4.12
2. NYK - 4.23
3. GSW - 4.3
4. SAS - 4.32
5. ATL - 4.48
6. PHI - 4.53
7. BOS - 4.57
8. DAL - 4.7

The Knicks play the triangle, so they should score pretty well here. The Warriors, Spurs, Hawks, and Mavs are teams that are expected on this list. I didn't expect the Jazz, but I think that team has a good future if things bounce right.

Half-Court Pace (half court movement estimated)
1. SAS - 42
2. PHI - 28
3. GSW - 17
4. DET - 15
5. UTA - 13
6. NYK - 12
7. CHA - 11
8. CHI - 7

I expected the Spurs to be #1, but holy crap. Frankly, this is higher than I expected the Sixers to be, but their coach is a Pop disciple. The rest is fairly close together and of mixed quality, but the Warriors scoring well here does not surprise.

XeFG% (expected eFG% based on shot selection)
1. HOU - 51.5
2. ATL - 50.8
3. PHO - 50.8
4. DET - 50.5
5. GSW - 50.5
6. NOP - 50.5
7. DAL - 50.5
8. PHI - 50.4
9. CLE - 50.4
10. UTA - 50.4

It's no surprise Houston leads the league in this since their entire offense is based on the statistically good shot. If you click on the link and see the difference between XeFG% and actual, it should come as no surprise that the Warriors and the Hawks have a lot of good shooters, the Pistons have bad shooters, and the Sixers have atrocious shooters. (Dallas shoots 1.7% better than expected and New Orleans shoots 1.2% worse).

On a Sixers tangent, the offenses they're trying to emulate are the Warriors, Hawks, and Spurs. The Sixers, Hawks, and Spurs have coaches from the same family. The Warriors are just really freaking good.

Best rim defenders in the league: http://nyloncalculus.com/stats/rim-protection/

There's not too much playing with stats here. It's just a calculation of the percentage of the opposing teams' shots that are contested at the rim by a player and the opponent's FG% on said shots, compared to the league average.


Here are all the players saving at least a point per game (top 11)

Rudy Gobert UTA 2.43
Andrew Bogut GSW 2.08
Roy Hibbert IND 1.84
Serge Ibaka OKC 1.76
Hassan Whiteside MIA 1.57
Nerlens Noel PHI 1.51
John Henson MIL 1.32
Tim Duncan SAS 29.0 1.21
Pau Gasol CHI 34.3 1.14
Alex Len PHX 22.0 1.13
Robin Lopez POR 1.12
Jonas Valanciunas TOR 1.10

#1 should be no surprise. In practice, coaches will hold up pads near the basket to simulate a shot blocker around the rim. Players have started to refer to these pads as "Goberts."

-------------------------------------------------------------------------------------

I did a silly exercise, but I took ESPN's RPM, ORPM, and DRPM to find out the median player per position (I tried to pick the biggest name if players were close).

Point Guard
RPM: Michael Carter-Williams
ORPM: Trey Burke
DRPM: Reggie Jackson

Shooting Guard
RPM: Dante Exum
ORPM: Jodie Meeks
DRPM: Dion Waiters

Small Forward
RPM: Nick Young
ORPM: Solomon Hill
DRPM: Wesley Johnson

Power Forward
RPM: Marcus Morris
ORPM: David West
DRPM: Brandon Bass

Center
RPM: Gorgui Dieng
ORPM: Gorgui Dieng
DRPM: Al Horford

Player
RPM: Dante Cunningham
ORPM: Michael Kidd-Gilchrist
DRPM: Blake Griffin

Quartiles:
PG: Jerryd Bayless, Carter-Williams, Isiah Thomas, Curry
SG: Dion Waiters, Dante Exum, JR Smith, Harden
SF: Kostas Papanikolaou, Nick Young, Nicolas Batum, Leonard
PF: Thomas Robinson, Marcus Morris, Nick Collison, Davis
C: Festus Ezeli, Gorgui Dieng, Tristan Thompson, Cousins.

-----------------------------------------------------------------

I really wish ESPN (or someone who wanted to copy it) would release their Gravity/Distraction/Respect Ratings of players. I want to see who moves a defense. Up through New Year's, this was the list:

1. Stephen Curry, Golden State Warriors | Respect rating: 97.9
Gravity score: 97.3 | Distraction score: 98.4

2. Kyle Korver, Atlanta Hawks | Respect rating: 96.0
Gravity score: 92.3 | Distraction score: 99.7

3. Klay Thompson, Golden State Warriors | Respect rating: 94.4
Gravity score: 89.6 | Distraction score: 99.1

4. Mike Conley, Memphis Grizzlies | Respect rating: 87.5
Gravity score: 79.3 | Distraction score: 95.7

5. Gordon Hayward, Utah Jazz | Respect rating: 84.0
Gravity score: 72.1 | Distraction score: 95.9

6. James Harden, Houston Rockets | Respect rating: 83.3
Gravity score: 67.2 | Distraction score: 99.3

7. J.R. Smith, New York Knicks | Respect rating: 83.0
Gravity score: 89.0 | Distraction score: 76.9

8. Jamal Crawford, Los Angeles Clippers | Respect rating: 81.6
Gravity score: 67.9 | Distraction score: 95.2

9. Manu Ginobili, San Antonio Spurs | Respect rating: 81.4
Gravity score: 65.4 | Distraction score: 97.3

10. Dwyane Wade, Miami Heat | Respect rating: 79.7
Gravity score: 82.2 | Distraction score: 77.1
-Not on Wade--teams have found out if you play him too loosely, he cuts to the basket and burns teams that way, so they have to play him like he's a 3-point shooter, to his frustration.

Notables: Kobe Bryant (13th); Derrick Rose (14th); Carmelo Anthony (31st); Russell Westbrook (39th); Chris Paul (45th); Dirk Nowitzki (47th); LaMarcus Aldridge (72nd); LeBron James (73rd); Tony Parker (77th); Chris Bosh (86th); Kyrie Irving (97th); Lance Stephenson (99th); Anthony Davis (114th); Monta Ellis (168th); Kevin Love (177th); Rajon Rondo (182nd); Tony Allen (188th); Chandler Parsons (211st); Elfrid Payton (223rd).

-----------------------------------------------------------------

Some more useful discussions on advanced basketball stats: http://apbr.org/metrics/viewtopic.php?f ... a94d271ded

-----------------------------------------------------------------
Wiggins one of the worst ROY winners (by rookie season) of all-time. There are some pretty good players who did poorly, though... http://pastebin.com/raw.php?i=kZBKbQgZ

I think I've found a potentially interesting market inefficiency based on this article on 3-point shooters who can suddenly shoot (or can't) when changing teams. Someone should figure out what makes someone a good shooter in one role and not in another and see if there are certain factors. Is it open 3-point percentage? Is it covered 3-point percentage? We have stats now for shooting in almost every conceivable situation. Someone who isn't me should crunch the data.

An old article that shows (in hindsight) possible problems with +/-, especially with young players. The Kevin Durant conundrum.

Yeah hockey's all about possession. Corsi% and Fenwick% are the two big numbers and they use shots directed at the net while on the ice as a proxy for possession (which they are pretty good at). Fenwick (I believe) adjusts for blocked shots and corsi doesn't account for them. On a team wide level both correlate excellently with sustained success and playoff success. The most useful measure here is performance in close situations at 5 on 5 since on powerplays/penalty kills you'll obviously have a big deficit and outside of 1 goal games there is always a catchup element observed in shots attempted by the trailing team that doesn't tell you as much about ability. On the individual level I'm of the view that they still need some work but they can tell us some stuff about the players. This can also tell you about quality of competition faced by various players.

Another very useful stat is zone start percentage. When you have a lot of offensive zone start time its easier to drive possession (since you're already in the offensive zone attempting shots) whereas if you start in the defensive zone a lot and still drive possession you've had to effectively turn the puck over and get shots on net.

Another stat that is good for determining how much success if based on luck is PDO which is save percentage plus shooting percentage on a team wide basis. Generally speaking a total of 100% is average and significant deviations to either side result from poor or good luck (though of course certain players demonstrate repeatable skills in saving and shooting accuracy).

There have been some attempts to measure scoring chances and their quality but they are in their infancy.

Hockey there is a lot of context since it's the most continuous of the four major American sports and there is a lot of on the fly changes and other issues that makes analysis of individual contribution hard.

damn somehow i missed the posts here from back in April. really great stuff, pgm. even for someone like me who is pretty locked in to stats, that data is new to me.


and that Kevin Durant article is really good. definitely worth considering when thinking about Wiggins (although Wiggins isn't Durant, either). young bigs do particularly poorly in plus/minus usually. it wouldn't shock me if Okafor was a "bad" player in terms of plus/minus even if he was killing it in raw production.


my way of analyzing players has become a sort of mix of various factors:
overall production (and minutes) vs overall efficiency vs adjusted plus/minus vs individual/specific skills (synergy stats, sportsvu stuff, etc) vs eye test. all taken with context, obviously.

we've seen too many players excel at one of these while failing in another and then switch teams to see the others catch up. context matters so much to the first three (production, efficiency and plus/minus). and even to skill stats: for example, its a lot easier to have great post-up per possession numbers if you have good shooting around you. i just think if you determine someone is a bad or great player based on one or two of those things while ignoring the others, you are going to get burned when the situation changes for that player.

i do think plus/minus, over a long enough sample period, for veteran players is a really good indication whether or not their role is the best fit for them.

Haha. Thanks. I would post pretty much every time I read something new.

Btw, go to Amin Elhassan's Twitter feed and go back to today (7/13/15) about four hours ago (2:00 p.m. EDT). There's an international coaching clinic with Ettore Messina and Mike D'Antoni. Great Xs & Os, stuff on analytics and more. Even some shots at conventional wisdom--e.g. D'Antoni was always asked how much he learned as Nate Archibald's backup, which he replied nothing because Archibald was a better player. But on stats, the target eFG% was 50%. This might be a better number--they wanted at least one point per possession. Post ups that only get 0.7 ppp are as useful as a jumpshot that gets 1+.



I tell people that all rookies are bad. Laker fans think Russell and Randle make them a playoff team. Young players make mistakes.

I think there's a good chance Okafor will be a "bad" player in BPM. Though Nerlens Noel was a positive player. There were three rookies who were positive last year--Noel, Mirotic, and Marcus Smart. If you look at the complete list, there's a mix of some good and some not-so-good players.



Yeah, you have to look at everything.

One thing I would love to see is a list of the most effective plays in basketball. Obviousy something like Kyle Korver shooting an open 3 has an expected value of close to three points, but that doesn't say how he got the three. That's kind of the next step is to really break that down. I'm sure coaches know the expected values of their plays.

you have more faith in some of them than i do. there is no doubt that every coach in the NBA is intelligent and knows basketball. but i'm sure there are plenty of guys in the NBA who wouldn't know this sort of thing or want to devote the time to it. there are still franchises that don't even give a shit about analytics.

You might be right. I just assumed there was some young film guy who wanted to get ahead and started breaking down their big plays. They do have the tracking data, so thy know the basics (pick and rolls, isos, post ups).

Turns out basketball-reference's BPM isn't actually a plus/minus, but an estimate.

I'm kind of bummed out by this. I will now take the numbers with more of a grain of salt. It's still probably a useful way of looking at players--at least offensively--but I need to take them for what they are.