Article on Ibaka
http://www.grantland.com/story/_/id/102 ... -congo-nba
Yes his midrange FG% is up. But his overall eFG% and TS% are at career lows. Ibaka's share of shots that are jump shots have risen 40% since rookie year. Hitting 50% of your mid-range shots and leading the league on that is good but taking a mid-range shots for him or most others is a worse decision on average than an inside shot if you can get it. Ibaka's at the rim FG% is his worst since rookie season. Still fine at 67% but down from 73% last season.
His OR% is at a career low as is his steal %. His block rate is barely above his rookie season career low. Yes his assist rate is up but it is still very low. His overall winshares per 48 minutes is below his career average and is the lowest since he was a rookie.
From another poster on a team blog: "Case of a player who is focusing on a new aspect of his game, letting the old aspects slip."
Team has been better with Ibaka at center last three regular seasons and he has been individually better there the last 2 but his share of time at center is near career low. Was much higher last playoffs. Basketball reference data seems conflicted about the team +/- then looking at summary and lineup data. Do they really know enough about Ibaka at Center to make a well informed decision?
Early 2013-14 stat-based observations
Re: Early 2013-14 stat-based observations
I've been looking at Ibaka as well. I'm not sure whether playing him "strictly" at center is the solution, but there does seem to be two HUGE factors effecting his success.
1.) When playing without Westbrook but with Durant, Ibaka has struggled. I can't recall the exact numbers, but when looking at NBAWOWY splits since last season, Ibaka's TS% drops by 10% or more without Westbrook.
2.) Kendrick Perkins. This season, when playing without Perkins, Ibaka shoots 10% higher (TS%). His percentage of shots within 3-feet jumps from 32% to 41%, and his fg% on those shots goes from 61% to 75%.
These observations seem to agree with common sense. Westbrook creates good looks for Ibaka by drawing defensive-attention, particularly when driving. Perkins is a space-eater with no ability to float on the perimeter for a jumper, forcing Ibaka to spend undo amounts of time 16 feet from the basket, just to be a player on offense.
http://hoopdon.weebly.com/
1.) When playing without Westbrook but with Durant, Ibaka has struggled. I can't recall the exact numbers, but when looking at NBAWOWY splits since last season, Ibaka's TS% drops by 10% or more without Westbrook.
2.) Kendrick Perkins. This season, when playing without Perkins, Ibaka shoots 10% higher (TS%). His percentage of shots within 3-feet jumps from 32% to 41%, and his fg% on those shots goes from 61% to 75%.
These observations seem to agree with common sense. Westbrook creates good looks for Ibaka by drawing defensive-attention, particularly when driving. Perkins is a space-eater with no ability to float on the perimeter for a jumper, forcing Ibaka to spend undo amounts of time 16 feet from the basket, just to be a player on offense.
http://hoopdon.weebly.com/
-
italia13calcio
- Posts: 100
- Joined: Sun Dec 08, 2013 2:54 am
Re: Early 2013-14 stat-based observations
Ibaka's been very high in my rankings, so I too have been looking at his game to figure out why he is so high. His midrange game, as Kirk pointed out, is very refined. For catch and shoot 2 pointers, he is near the top ten. He also has a good close game. Crow, I'm working with the SportVu data for close shots, not exactly the same as shots at the rim, and according to those numbers he's finishing at a 76% rate - which needless to say is very impressive and by my PAA metric puts him in the top ten of that category.
While his OR% may be at a career low, his TR% is up from last year. He is also averaging 9 rebounds a game, up over a reb per game from last year. It may be that as OKC asks him to do more, his efficiency slips, which is understandable. Still, his rebounding puts him near the top ten according to my metric. Although his blocks are down he is still a fantastic defender at the rim, allowing opponents to score only 42% - way below average.
This also got me thinking about a debate Bill and Jalen started on one of their more recent videos: should Ibaka or David Lee make the All-star game? With a loaded West, its no guarantee that either will, but Bill seemed adamant that Ibaka should while Jalen thought Lee would. Crow, given your view that Ibaka has become less efficient (worse?) this year, is it safe to say that you don't think he should?
While his OR% may be at a career low, his TR% is up from last year. He is also averaging 9 rebounds a game, up over a reb per game from last year. It may be that as OKC asks him to do more, his efficiency slips, which is understandable. Still, his rebounding puts him near the top ten according to my metric. Although his blocks are down he is still a fantastic defender at the rim, allowing opponents to score only 42% - way below average.
This also got me thinking about a debate Bill and Jalen started on one of their more recent videos: should Ibaka or David Lee make the All-star game? With a loaded West, its no guarantee that either will, but Bill seemed adamant that Ibaka should while Jalen thought Lee would. Crow, given your view that Ibaka has become less efficient (worse?) this year, is it safe to say that you don't think he should?
Re: Early 2013-14 stat-based observations
I hadn't looked at that but I guess my initial lean is towards Lee. But I see his eFG% and TS% are at careers too. Coaches probably won't split statistical hairs to decide it.
Re: Early 2013-14 stat-based observations
Actually I can't see either actually making it. Won't they take Dirk or Duncan or possibly Cousins instead?
Re: Early 2013-14 stat-based observations
I compared the average stats of the big minute PGs of the 6 best teams vs the average for the rest. The biggest percentage performance advantages for specific stats were for offensive and defense rebounding (+9 and +15% respectively), 3 pt FG% (+8.5%) and TO% (about +8%). Steals, eFG% and TS% edges were about +5%. Usage was a bit more than 3% higher. Assist rate was actually 2% lower (more and better secondary passers?). they were 3% older. They actually played a bit less minutes per game. Overall their winshares per 48 minutes was 57% higher. Their offensive rating was almost 5% higher while the defensive rating was a bit more than 4% higher. (Quite a big difference in the % difference on these metrics.)
Not sure if these areas of better performance are most important or just what it happens to be. But something that could be thought about.
Not sure if these areas of better performance are most important or just what it happens to be. But something that could be thought about.
-
italia13calcio
- Posts: 100
- Joined: Sun Dec 08, 2013 2:54 am
Re: Early 2013-14 stat-based observations
Yeah I don't know if either will make it either. Obviously my metric seems to think Ibaka should be a shoe in, but the West is loaded and all star berths aren't exactly decided by analytics. I just thought it was a little strange for Bill to seem so sure Ibaka would be in and the same with Jalen's view on Lee.
Re: Early 2013-14 stat-based observations
May say some things about Bill and Jalen and how they view "athletic" and "skilled" players, or substitute another "dividing" line for people..., or maybe they just wanted an argument.
Or maybe Jalen went with the player more known for offense (being oriented that way himself... at least in his mind or my perception of his mindset. His career offensive and defensive ratings actually are 103 108 respectively which makes him below average on offense and pretty average on defense for the time period). And Beantown Bill went with a Bostonian "defense wins championships" perspective. By offensive and defensive ratings there isn't much difference between these 2 players this season- absolute pt variation of 6 total on the 2 ratings. They have actually have never been as similar in past years as now by this method, FWIW.
Or maybe Jalen went with the player more known for offense (being oriented that way himself... at least in his mind or my perception of his mindset. His career offensive and defensive ratings actually are 103 108 respectively which makes him below average on offense and pretty average on defense for the time period). And Beantown Bill went with a Bostonian "defense wins championships" perspective. By offensive and defensive ratings there isn't much difference between these 2 players this season- absolute pt variation of 6 total on the 2 ratings. They have actually have never been as similar in past years as now by this method, FWIW.
Re: Early 2013-14 stat-based observations
DeMarcus Cousins leads all centers (and C-F and F-C) in PER (27.1). His ws/48 (.180) trails only Anthony Davis'.
That PER is 4th in the West (after Durant, Paul, and Love) and behind only LeBron from the East.
He's also 5th in TReb%, and his Ast% (17.8) is roughly equal to the sum of those other 4 (Drummond, Jordan, Bogut, Howard).
He's the first player since Shaq in 2000 to have per-36 rates of at least 26-12-3 (pts-reb-ast).
Before that, it's all the way back to Kareem in 1972; before that, Wilt, Baylor, Pettit.
http://bkref.com/tiny/D6NsS
That PER is 4th in the West (after Durant, Paul, and Love) and behind only LeBron from the East.
He's also 5th in TReb%, and his Ast% (17.8) is roughly equal to the sum of those other 4 (Drummond, Jordan, Bogut, Howard).
He's the first player since Shaq in 2000 to have per-36 rates of at least 26-12-3 (pts-reb-ast).
Before that, it's all the way back to Kareem in 1972; before that, Wilt, Baylor, Pettit.
http://bkref.com/tiny/D6NsS
Re: Early 2013-14 stat-based observations
Cousins has to be in the running for the empty stats teamMike G wrote:DeMarcus Cousins leads all centers (and C-F and F-C) in PER (27.1). His ws/48 (.180) trails only Anthony Davis'.
That PER is 4th in the West (after Durant, Paul, and Love) and behind only LeBron from the East.
He's also 5th in TReb%, and his Ast% (17.8) is roughly equal to the sum of those other 4 (Drummond, Jordan, Bogut, Howard).
He's the first player since Shaq in 2000 to have per-36 rates of at least 26-12-3 (pts-reb-ast).
Before that, it's all the way back to Kareem in 1972; before that, Wilt, Baylor, Pettit.
http://bkref.com/tiny/D6NsS
Re: Early 2013-14 stat-based observations
Who else goes on an 'empty stats' team?
Love? Lee? Randolph? Jefferson?
They're all bigs. Who among guards?
Is DMC being promoted as trade bait?
Now that I look at it, I see he played 41 minutes in a win over Miami, with 27-17-5.
Then he had 29-14-3 in a loss to the Spurs, in which Sac led by 6 after 3.
Then in a win over the Rockets he had 17-16-2, with 4 Stl. Dwight had 15-8, 1 block, in a lot more minutes.
Love? Lee? Randolph? Jefferson?
They're all bigs. Who among guards?
Is DMC being promoted as trade bait?
Now that I look at it, I see he played 41 minutes in a win over Miami, with 27-17-5.
Then he had 29-14-3 in a loss to the Spurs, in which Sac led by 6 after 3.
Then in a win over the Rockets he had 17-16-2, with 4 Stl. Dwight had 15-8, 1 block, in a lot more minutes.
Re: Early 2013-14 stat-based observations
"Who else goes on an 'empty stats' team?"
IMO it should be the guys with the largest gap between PER (or maybe offensive winshares per 48 minutes) and RAPM.
The only three players with a 20+ PER last season and a RAPM estimate below +2 were Al Jefferson (at+1.2), Chrish Bosh (at +1.9) and Cousins (at +1.9). This is still positive but somewhat disappointing and not good value on the first two.
David Lee was PER 19 but only +1.2 RAPM. Kemba Walker was almost PER 19 but only +0.4 on RAPM in his third season. Ryan Anderson was PER 18 but only +0.4 on RAPM. Shawn Marion PER 18 but -0.7 on RAPM for Cuban's team that once lead on APM usage but may not use it as much now. Zach Randolph, Ty Lawson, JR Smith, Lillard, Monta Ellis, Tristan Thompson, Gay, G Henderson, Nash and Vucevic were PER 16- 18 but fairly weak to just above average on RAPM with Vucevic at -0.4 and Henderson at -0.9. Boozer was PER 17 but -1.1 on RAPM. With the possible exception of Chicago all these teams are considered active in advanced analytics.
Among the worst on RAPM who played 20+ minutes, Ben Gordon was -6.4 on RAPM last season (very worst in league) but managed a 13 on PER.
Who traded for and signed both Jefferson and Gordon? Michael Jordan and company.
Afflalo was PER 13 but -2 on RAPM. Wesley Johnson PER 11 but -5 on RAPM.
If one wants to focus more on "empty scorers" then the comparison should be on points per game compared to RAPM.
Brandon Jennings scored almost 18 pts per game last season but had an overall RAPM estimate of -2.5. Which GM / front office is going to give him $8-10-12+ million per year? DeRozan also scored 18 pts per game but was -1.3 for the Raptors, a team which I heard does not use APM much or at all. They paid him nearly $10 million / yr for 4 seasons.
Joe Johnson was PER 14 and about +1 on RAPM but a pretty empty value given his $20 million salary last season. Not so terrible a performance / pay ratio but OJ Mayo was at best a blah move by Milwaukee's management. Surely there was better ways to spend $8 million a year.
IMO it should be the guys with the largest gap between PER (or maybe offensive winshares per 48 minutes) and RAPM.
The only three players with a 20+ PER last season and a RAPM estimate below +2 were Al Jefferson (at+1.2), Chrish Bosh (at +1.9) and Cousins (at +1.9). This is still positive but somewhat disappointing and not good value on the first two.
David Lee was PER 19 but only +1.2 RAPM. Kemba Walker was almost PER 19 but only +0.4 on RAPM in his third season. Ryan Anderson was PER 18 but only +0.4 on RAPM. Shawn Marion PER 18 but -0.7 on RAPM for Cuban's team that once lead on APM usage but may not use it as much now. Zach Randolph, Ty Lawson, JR Smith, Lillard, Monta Ellis, Tristan Thompson, Gay, G Henderson, Nash and Vucevic were PER 16- 18 but fairly weak to just above average on RAPM with Vucevic at -0.4 and Henderson at -0.9. Boozer was PER 17 but -1.1 on RAPM. With the possible exception of Chicago all these teams are considered active in advanced analytics.
Among the worst on RAPM who played 20+ minutes, Ben Gordon was -6.4 on RAPM last season (very worst in league) but managed a 13 on PER.
Who traded for and signed both Jefferson and Gordon? Michael Jordan and company.
Afflalo was PER 13 but -2 on RAPM. Wesley Johnson PER 11 but -5 on RAPM.
If one wants to focus more on "empty scorers" then the comparison should be on points per game compared to RAPM.
Brandon Jennings scored almost 18 pts per game last season but had an overall RAPM estimate of -2.5. Which GM / front office is going to give him $8-10-12+ million per year? DeRozan also scored 18 pts per game but was -1.3 for the Raptors, a team which I heard does not use APM much or at all. They paid him nearly $10 million / yr for 4 seasons.
Joe Johnson was PER 14 and about +1 on RAPM but a pretty empty value given his $20 million salary last season. Not so terrible a performance / pay ratio but OJ Mayo was at best a blah move by Milwaukee's management. Surely there was better ways to spend $8 million a year.
Re: Early 2013-14 stat-based observations
http://hoopshype.com/columns/david-nurs ... ubios-shot
"A glaring stat that reflects Rubio’s lack of confidence in his own shot is his clutch statistics (fourth quarter or overtime stats). Rubio is shooting 22.2 percent and his Timberwolves are -12.9 points per 48 minutes with him on the floor in clutch situations."
9th season as a "pro", 3rd in NBA. His shot should have shown more progress by now.
"A glaring stat that reflects Rubio’s lack of confidence in his own shot is his clutch statistics (fourth quarter or overtime stats). Rubio is shooting 22.2 percent and his Timberwolves are -12.9 points per 48 minutes with him on the floor in clutch situations."
9th season as a "pro", 3rd in NBA. His shot should have shown more progress by now.
-
Bobbofitos
- Posts: 306
- Joined: Sat Apr 16, 2011 7:40 am
- Location: Cambridge, MA
- Contact:
Re: Early 2013-14 stat-based observations
At a certain point you are what you eat.Crow wrote:http://hoopshype.com/columns/david-nurs ... ubios-shot
"A glaring stat that reflects Rubio’s lack of confidence in his own shot is his clutch statistics (fourth quarter or overtime stats). Rubio is shooting 22.2 percent and his Timberwolves are -12.9 points per 48 minutes with him on the floor in clutch situations."
9th season as a "pro", 3rd in NBA. His shot should have shown more progress by now.
Rubio is a very unique player. I don't think he's a good, unique player, though.
http://pointsperpossession.com/
@PPPBasketball
@PPPBasketball
-
talkingpractice
- Posts: 194
- Joined: Tue Oct 30, 2012 6:58 pm
- Location: The Alpha Quadrant
- Contact:
Re: Early 2013-14 stat-based observations
We track this every day, though replacing PER here with our FORPM model (random forest based spm/boxscore model, details on our site). Here's the top 5 in terms of FORPM minus RAPM for this season thus far (using NPI RAPM) ->Crow wrote:"Who else goes on an 'empty stats' team?"
IMO it should be the guys with the largest gap between PER (or maybe offensive winshares per 48 minutes) and RAPM.
PLAYER EMPTY
John Henson 6.261194862
LeBron James 5.739122948
Joakim Noah 4.673948936
DeMarcus Cousins 4.638314698
Andre Drummond 4.474820934
Slightly off topic.... Obviously, lots of people now will tell me that Lebron is an sss fluke. Now that the season is half over, I completely eschew that notion. He's blatantly coasting on defense, as evidenced by his approx -2 RAPMd this season thus far, and this time series of BLK% and STL% over his career (rookie year in top row, 2014 in bottom row) ->
STL% BLK%
2.2 1.3
2.8 1.1
2.0 1.5
2.1 1.3
2.4 2.1
2.4 2.4
2.2 2.0
2.1 1.3
2.6 1.7
2.4 1.9
1.8 0.8
We got so many accusations of overrating him last year, and now are getting the precise opposite comments. But his numbers are what they are. He decided to punt on defense this season, for the obvious reason(s). ftr I posted something similar to this on a different/worse forum earlier this week in their Lebron thread, and got banned within 24 hours for reasons I've yet to figure out, which I thought was quite phunny of them.
But anyway, back to the original comment by Colts re DeMarcus. The evidence backs that up completely. Big caveat here though (concurs with Mike G's point re his historic performance on offense)... his RAPM numbers are okay (+1.4), and the big deviation we see here is because his FORPM is intensely high (2nd in nba after KD..... and a tad above LBJ and CP3, who are themselves a mile ahead of the rest of the pack). He's near the historical pinnacle of the USG/Efficiency frontier. On defense, he's 3rd in the NBA in DRB%, and (gasp) 8th in STL%. Obv boxscore models don't account for the lazy defensive lapses, bad rim protection, and confused rotations. He has the 4th highest IPV in the league right now though, putting it all together.