more Heat, less Heat (schtevie, 2011)
Posted: Fri Apr 15, 2011 12:34 am
1 of 2
Author Message
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Tue Feb 15, 2011 6:17 pm Post subject: more Heat, less Heat Reply with quote
So, the season's well past half over, and I thought it might be worth focusing a bit of attention on the Heat. Why I write is that I am looking for the best explanation as to why the Heat have done such a bad job integrating the offensive talents of James, Wade, and Bosh. To me this is more than a bit of a mystery.
Here is the benchmark comparison exemplifying the (relative) failure. In Cleveland in 2008-09, the top two most used lineups featured James and Ilgauskas. Rounding these out were Mo Williams, Delonte West, and either Ben Wallace or Anderon Varejao. These had an Off Rtg of 6.6 and 7.1 over the league average, respectively.
Compare this to this year's most used lineup for the Heat. Added to the same James and Ilgauskas are Wade, Bosh, and Arroyo. They have an Off Rtg of 7.1 over the league average. (And the swapping of Chalmers for Arroyo shows much worse.)
Not to denigrate the offensive prowess of Wallace and Varejao, but this result simply should not be. Diminishing returns, indeed.
Is part of the explanation that the similarity really isn't. That the Heat have been significantly improving over the course of the season? Does anyone know of a source which shows a time plot of Off Rtg? The Heat Index has a recent suggestive plot of a secular decrease in long 2s, and perhaps that is part of the story.
What is the story?
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Tue Feb 15, 2011 8:56 pm Post subject: Reply with quote
An offensive rating of 114 is pretty good. The current Heat lineup with that rating is about 10th best in the league of the 40 used over 200 minutes. The two from the Cavs in 2008-9 were about 10th best for the top 50 most used lineups.
The current Heat lineup is notably worse than those Cavs lineups on net turnovers according to the 5 man data at 82 games. Not sure if or how much of that difference is coming on offense though.
The game level team efficiency data is available at this link and you could compile how you want or it is probably also in the motion chart section there too (or the motion chart at DSMok1's site).
http://hoopdata.com/teamgl.aspx?team=MIA
Back to top
View user's profile Send private message
greyberger
Joined: 27 Sep 2010
Posts: 50
PostPosted: Wed Feb 16, 2011 12:52 am Post subject: Reply with quote
The 2010 Cavs with Lebron on the court shot 54.8% eFG (+6.4% compared to off court) and made 20 FTs per 48 (+5 versus off).
The 2010 Heat with Wade shot 51.5% eFG (+6.8% versus off) and made 19 FTs per 48 (+6 versus off). The Toronto units with Bosh in also got about 5 more FTs made per 48, FWIW.
The Heat 'starter' lineups this year shoot ~52% and make ~23 FTs per 48. Is that better than the 2010 Heat, or worse than the 2010 Cavs? The team as a whole is about 6th best in eFG, 3rd best in offensive FTR. What's the right comparison here?
What would it take to exceed expectations? 60% eFG and the same free throw rate?
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Wed Feb 16, 2011 7:01 am Post subject: Reply with quote
An Off Rtg of 114 is pretty good. It's, of course, simply zero improvement on what was.
Is this not more than a little bit surprising? You swap out Williams, West (decent offensive players) and a millstone offensive PF/C and replace them with Arroyo, the best (?) offensive shooting guard, and one of the best offensive power forwards in the league, and the result is....nothing?
The numbers are a bit dated but compare the sums from Steve and Aaron's 2007-08 Offensive +/-s. A swap of -5.7 for +11.1, a net change of +16.8. One might have expected a slight improvement.
That there is but one ball to share suggests that there would be some diminishing returns, but what explains this?
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3597
Location: Hendersonville, NC
PostPosted: Wed Feb 16, 2011 8:05 am Post subject: Reply with quote
Superstars each have a team built around them. Put them on one team, and it's a different environment. The best teams have been a superstar and some strong supporting players. On occasion, 2 such players (Robinson and Duncan, Kareem and Magic) .
The nearest similar case might be the assembling of the 2008 Celtics. They did OK right off the bat, but those guys played distinct positions, unlike this bunch. LeBron's kind of bouncing around between PF and PG.
So now we have a sample of 1.5 such seasons of gathering +/- superstars.
And with the Celts, a few others emerged in significant roles: Rondo, Perkins, Posey, House .
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Wed Feb 16, 2011 9:09 am Post subject: Reply with quote
Mike, I like seeing the conventional wisdom overturned as much as next guy (probably more, actually) but I think you assign a wee bit much to the importance, make that the centrality, of role players. PR aside, why bother to go to all the trouble assembling a Dream Team (apologies for using this horrible term) for the Olympics, if a couple of superstars and some solid role players would have equal effect?
The predicate is that these three guys had histories of being very efficient at being able to score, distribute, and not turn the ball over. It wasn't an illusion that they were offensive superstars. (The Adj +/- numbers are but one reference.) Accordingly, in combination, the expectation wasn't and isn't (?) that there should be no net offensive improvement.
Back to top
View user's profile Send private message
Ilardi
Joined: 15 May 2008
Posts: 265
Location: Lawrence, KS
PostPosted: Wed Feb 16, 2011 11:34 am Post subject: Reply with quote
Schtevie,
I think there are two plausible explanations (not mutually exclusive) for the conundrum you've identified:
1) As you know, APM values simply reflect a player's impact in a given role; thus, when you add two or more superstars to the same lineup and thereby cause their respective roles to change substantially (as in the case of Wade and LeBron, both of whom were formerly in the role of single-handedly dominating the offense), you likely wind up with substantially diminished contributions from each;
2) Adjusting/developing an offense to capitalize efficiently on the presence of multiple superstars with overlapping roles may take more time than generally appreciated . . . thus, it would be highly informative to look at time-trend analyses of the Heat's offensive efficiency this year (especially on a lineup by lineup basis) to see if indeed there is a trend toward greater efficiency as the season has progressed.
Also, of course, role players really do enter into the equation: It may well be that advancing beyond an already-lofty efficiency of 114 with James/Wade/Bosh requires the presence of two highly specialized role players with skill sets that may or may not be present on the Heat's current roster.
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Wed Feb 16, 2011 12:50 pm Post subject: Reply with quote
What do the top 50 minute lineups over 114 on offensive efficiency from the last 2 seasons look like in general?
About 2/3rds have a bigtime assist PG.
About 2/3rds have a notable post scorer or threat.
About 2/3rds have 3+ decent 3 pt. shooters on the court at the same time.
2/3rds of the lineups were from 8 teams.
About 87% were playoff level teams.
It doesn't appear that many of these teams are fastbreak leaders.
The teams with 2 or more such lineups include Boston, Dallas, Denver, Houston, Clippers, Miami, Lakers and OKC. 6 of the 8 Coaches played PG in the NBA, Phil Jackson and Eric Spoelstra did not.
Arroyo is far from a bigtime assist PG this season. They do get assists from others but they are 27th on team assists per game.
Miami gets the least team shots at the rim per game in the league and Bosh is only getting 3 per game.
Arroyo, Wade, James are three 3 point shooters, but 2 are below average on 3pt FG%. As a team they are 8th on 3pt FG% but only 14th on attempts. They are #2 on long 2s (16-23 ft.) per game.
Looking at the top Heat lineup it appears that most of stints where it had really poor offensive efficiency the opponents had a shotblocker or at least a big / tough dude inside.
With the 08-09 Cavs, James had a bit more help on assists from Williams, and better 3 point shooting but again very low shots at the rim (28th). With Shaq in 09-10 only moved them up to 16th on shots at the rim (and team assists went up too). He probably needs a "bigger" post scorer or more frequent drivers, more good passing and more 3 point shooting. Getting to 114 on offensive efficiency without being stronger on these criteria, without a former NBA PG as coach, without team familiarity from last season may be a semi-accomplishment. (With the Cavs James also lacked a former NBA PG as Coach as did Bosh and Wade with their former teams.)
Last edited by Crow on Wed Feb 16, 2011 1:27 pm; edited 1 time in total
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 68
PostPosted: Wed Feb 16, 2011 1:26 pm Post subject: Reply with quote
Crow, what do you think about trey volume here? I know that's something you've followed closely (you follow everything closely!).
Noticed:
08-09 Cleveland: 8.8 treys per 100 possessions
10-11 Miami: 7.4 treys per 100 possessions
Maybe it's hard to crack a historic efficiency at just 6.9 treys per game, particularly if you're not getting an inside/outside mix (you mentioned Miami not getting many shots at the rim). Just a thought.
Might connect back to Schtevie's earlier thread about coaches not maximizing the trey influence. Might be a dead end. Figured I'd throw it out there because I had remembered that 08-09 team having some perimeter threats LeBron could kick out to. That 1.4 treys per 100 possessions is at least a noticeable difference between the two teams in the discussion...
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Wed Feb 16, 2011 1:32 pm Post subject: Reply with quote
Thanks Jeff. I agree with you about at least looking hard at the role of 3 point shooting on top efficiency teams and top performing teams in the playoffs.
'08-09 Cleveland was 3rd on made 3s per game. One leg of the inside / 3 pt / assist combo but just one leg. (Positing that this combo matters to see what is found.)
'08-09 champs Lakers were a bit below average on 3 pt makes but compensated with near top on team assists and top 10 on shots at the rim.
Last season the Lakers weren't top 10 at any of these regular season and don't appear special on at least 2 of these in the playoffs either but made it with experience, a top 4 overall playoff offense (aided by #1 offensive rebounding rate) and 3rd best playoff shot defense. They did it another way outside being strong on these 3 criteria. (They were high on 3 pt attempts in the playoffs but shot below average on them.)
In '07-08 the Celtics were top 10 on 2 legs (assists and 3s).
In '06-07 the Spurs essentially had all 3 (one 11th place).
This season the Lakers are balanced and nearly have all 3. Denver also almost have 3. Orlando, Dallas and Atlanta 2. Boston just 1. Miami almost 1. OKC and Chicago not close to even one.
3 of the last 4 champs were top 10 on 2. Whether it will be the case again this season will help build or weaken the case for the 3 criteria being pretty important.
Last edited by Crow on Wed Feb 16, 2011 2:31 pm; edited 4 times in total
Back to top
View user's profile Send private message
horsecow
Joined: 01 Dec 2009
Posts: 12
Location: Iowa City, IA
PostPosted: Wed Feb 16, 2011 2:08 pm Post subject: Reply with quote
I wonder if free throw shooting has something to do with this. This year, all of the big three have gotten to the line less:
FTA
LeBron down from ~10.0 to 8.8
Wade down from ~9.5 to 9.0
Bosh down from ~8.4 to 5.7
= down from ~27.9 FTA to 23.5 FTA
To equal the same number of points, their collective FT shooting would have to improve considerably, but it's actually gotten worse (especially Wade's). Free throw shooting is the one area where having a superstar teammate doesn't make completing the shot any easier, so if these three are getting to the line less (and the team's improved offensive efficiency in other areas isn't compensating), the team is scoring less.
Why they are getting to the line less is another question, though. Lack of floor spacing for Bron/Wade drives? A "superstar foul" quota by league refs?
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Wed Feb 16, 2011 2:20 pm Post subject: Reply with quote
Free throws could be a 4th leg to a strong offense... and turnovers and offensive rebounding could 5th and 6th if you want to stretch back out beyond field shooting and the related assists. Maybe some teams compensate for not having 2 of the 3 legs by having the 4th or 5th or 6th. '09-10 Lakers were 0 for 3 on the 3 initially name criteria but go 2 for 6. Enough I guess. How much the mix of ingredients matters in strong offensive efficiency and what combo has the best track record would take more research / review and writing. There are some old threads where I and a few others looked at champ and top contenders by the 4 Factors and overall efficiencies. And I now see there was a 2010 JQAS article on this very topic.
If you look at the 6 offensive criteria, Lakers nearly have 4, Denver almost 5. Orlando 3, Dallas and Atlanta still just 2 Boston almost 2. Miami almost 2. OKC 1 and Chicago almost 1.
Of course defense matters too.
I think it was useful to quickly compare the Heat's offense with these others but back to focus on the Heat themselves if you wish...
Back to top
View user's profile Send private message
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Mon Feb 21, 2011 1:47 pm Post subject: Reply with quote
I am a bit bewildered, perhaps even bothered, by the general sense - if this is true - that there is nothing really surprising about the Heat's offensive performance this year: An Offensive Rating of 114 is a very good start; Shallow learning curve and all that; Nothing in the historical record suggesting that they should have been special. Eh.
I put the possibility aside, for now, that the average performance over the year is not representative, that the Heat offense today is actually far more potent now than at the beginning of the season. I do so because I don't think it's true. Having intermittently followed the relevant statistic over the season, I have no sense that there has been a trend showing improved offensive efficiency over time. I could be wrong. (Perhaps Tom H will choose to anchor the discussion at the eponymous blog with this paramount Heat Index.)
If this impression is correct, however, there is then the issue of explaining and contextualizing the Heat's offensive performance. And my point is that what has transpired in this regard is completely at odds with what was anticipated. The degree of diminishing returns was (correctly) the focus of preseason discussion; the possibility of no returns (correctly) not being part.
In addition to their reputations of having high (at least, not low) basketball IQs, Wade, James, and Bosh going into this season each had a clear record of being a superior ball handler, scorer, and passer. The Heat also acquired 3 point shooting specialists. The direct implication of these facts was that opposing defenses would be stretched and in turn each component of the Heat offense would be expected to see less pressure. That there is one ball to be shared only implied that the gross statistical totals (Pts, PPG, Assists, etc.) of one or more players would be expected to decrease. The overall efficiency of the offense however was necessarily expected to increase. But there is no compelling evidence that it has (compared to a reasoned baseline of expectations) and much that it hasn't.
There is the highly suggestive evidence provided by the benchmark Cleveland Ilgauskas/James starting lineup of 2008-09. Expanding on this, the minute weighted average Offensive Rating of all Ilgauskas/James lineups, in 2008-09, was 114.3, 6.0 points above the NBA average. By contrast, this year's I/J average lineups have yielded 113.1 points per 100 possessions, 6.1 points above this years NBA average. Again, no change. Additionally, the average of all I/J/W/B lineups this year (there have been five) have been 114.2, 7.3 points above the NBA average. Compare that to the top ten I/J lineups (in terms of time on court) from 2008-09. These clocked in at 115.6, 7.3 points above that year's NBA average. Again, no change. So, it isn't just a starting lineup phenomenon.
What then accounts for the dysfunction in Miami, holding Ilgauskas constant? Interestingly and surprisingly (I suppose) it has been the addition of Wade and extraction of James from the Ilgauskas lineups that has most improved offensive performance (though this should be taken with a few grains of salt as it is an observation based upon a small sample of infrequently used lineups).
For all lineups (693.47 minutes worth) that minimally include I/J/ and B (that is when Wade is not necessarily on the floor), these yield an Off Rtg of 113.1. Then of these, in the 53.38 minutes when Wade isn't on the floor, Off Rtg plummets to 99.0. Similarly, the I/J/W then "not Bosh" numbers are 113.7 and 109.0 (in 707.26 and 67.17 minutes respectively). So, Bosh is apparently positive. Then finally, there is the benefit of getting rid of James. The I/W/B and "not James" numbers are 114.8 and 125.1 (in 677.69 and 37.6 minutes respectively).
The conclusion that the Heat have produced, um, sub-par offensive performance this year does not hinge on Ilgauskas however. The basic and general evidence can be found in the year on change in the effectiveness of the three stars. A list of some unanticipated results:
(1) The Assist %ages of James, Wade, and Bosh have all decreased (by 7.1 %age points on average). Well above average passers, with anticipated greater opportunities to make assists, suddenly unable to pass? How does that happen?
(2) The Turnover %ages of James and Wade have increased (a bit) despite decreased individual usage. With what story of offensive organization is that compatible?
(3) But more comprehensively, there have been significant decreases in the offensive efficiency of James and Bosh, offset only in part by a smaller improvement by Wade.
Consider a play efficiency measure, inclusive of the effect of turnovers, call it TS*%, equaling (Pts/2)/(FGA + 0.44*FTA + TO). Compared to last year, Wade's TS*% has gone from 0.494 to 0.504, a 1.05 %age point improvement. By contrast, James' has gone from 0.530 to 0.496, a drop of 3.43% points. And Bosh's has dropped 1.81 %age points, from 0.529 to 0.511.
These results, it should be repeated, are incompatible with inferences from basic offensive theory (assuming increased scoring potential, of course): individual scoring opportunities would necessarily collectively drop, but the realized efficiency should have increased.
James and Bosh haven't been injured, have they? Looking at the year on shot clock comparisons at 82games shows no difference to suggest that this might be a mitigating factor. (And by the way, let me take this opportunity to, yet again, thank Aaron and Roland for the wonderful resources that they provide to the community. It would be a much poorer world....) The coach seems as aware as any, regarding statistical issues, yet he somehow hasn't been able to combine these extraordinary resources in a way that yields a (non-trivial) net improvement.
What is left to explain the result (again, assuming no underlying time trend in the data) an appeal to randomness?
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Mon Feb 21, 2011 2:21 pm Post subject: Reply with quote
Restating and expanding points made above:
1. The Heat are below average in the league on offense from the PG position and not very directive with passing from that traditional hub. And very low on team assists. Maybe 2 heavy ball-handling wings with weak PGs is not that overpowering in pressure on the defense.
2. Just average on team attempts from the 3 point zone. Miller being back helps some but hasn't make them consistently elite on 3 pt frequency yet. Hard to keep Jones or Miller on the court a lot with Wade and James playing the same positions most of the time. Wade and James are averaging 4.6 3 pt attempts in 77 combined minutes per game over the last 10 down from about 6.5 for the season. I'd think that is low for starting wing combos. In fact, using hoopdata, it appears that starting wing combos at the same minute level would average about eight 3 point attempts, so what they get from House, Miller and Jones is needed to help get them back to average.
3. 30th at shot attempts in the paint (and presumably very low even in lineups with the big 3 and Z since they are dead last). If I was the defending them I'd guard the paint pretty hard and live with them trying to win with mid-range or 3 point shots.
All things that make them easier to defend I'd think. And having a Coach who didn't play in the NBA, especially at PG may limit their greatness on offense too. I expected some issues from points 1 and 2 but I didn't expect the 3rd point would be as bad a problem with Bosh and 2 drivers.
Last edited by Crow on Mon Feb 21, 2011 4:09 pm; edited 3 times in total
Back to top
View user's profile Send private message
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Mon Feb 21, 2011 2:49 pm Post subject: Reply with quote
Crow wrote:
If being below average in the league on offense from the PG position and not very directive with passing from that traditional hub, being average on team attempts from the 3 point zone and being 30th at shot attempts in the paint (and presumably very low even in lineups with the big 3 and Z since they are dead last), all things that make them easier to defend I'd think, and having a Coach who didn't play in the NBA, especially at PG, doesn't help your search enough I don't think I have anything extra to add at the moment.
Not sure what to make of the "Coach who didn't play (PG) in the NBA" correlation. Why should we believe that that should have a functional relationship with outcomes, offensive or defensive?
And it was always expected that offensive distribution through the PG wasn't going to be a feature of the offense. If you are importing three really good passer/scorers, you neither need nor want that function to be performed by another. Surely (?) if the ball was going more through Arroyo or Chalmers, things would be worse, not better.
Finally, that the offense is not getting enough high quality shots (3s and shots in the paint) is an alternative expression of the offense sucks (compared to potential) phenomenon not an independent explanation. They surely (?) have the raw material that gives them the potential to score from far as well as near.
I do note that Tom H has a nice bit today about how Bosh has regressed and could improve on the latter account.
recovered page 2
Author Message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Mon Feb 21, 2011 2:53 pm Post subject: Reply with quote
It is more a starting point than conclusive answer, but I noted above that of the 8 teams with 2 or more big minute lineups over 114 on offensive efficiency this season 6 of the 8 Coaches played PG primarily in the NBA. Eric Spoelstra did not. About 40% of coaches played PG in the NBA so this group has almost double the % representation on teams with 2 or more big minute lineups over 114 on offensive efficiency than expected. It could be looked at deeper.
Knowing from firsthand experience how things really are on the NBA court and what works best might be worth something in designing and managing an offense. It is not a strange suggestion, in my mind, to think that not having that experience could either hamper very high offensive performance or at least the coaching not be helping as much as it might with such experience.
If a Coach without NBA PG experience doesn't get the Heat a title in a year or two I'd expect we will see a new Coach with at least NBA experience and perhaps with NBA PG experience.
If a Coach with either version of NBA experience were to take over the Heat in the future you could get a team specific comparison of the height of the top lineup and overall offensive efficiency achieved by the Coach without such experience to with.
Last edited by Crow on Mon Feb 21, 2011 10:51 pm; edited 7 times in total
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3627
Location: Hendersonville, NC
PostPosted: Mon Feb 21, 2011 3:00 pm Post subject: Reply with quote
I'll suggest that at age 26, both LeBron and Bosh can not be expected to get any better and may well be in decline. Wade at 29 is decidedly over the hill.
Chalmers is 24, Joel Anthony is 28, and everybody else on the team is 30 or older. Z, Juwan, and Damp all have actual gray in their beards.
While playoff experience will prove to be important in the playoffs, the major players here have had a lot of (undue?) playoff pressure in their young lives. Probably a trade-off there.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Mon Feb 21, 2011 3:34 pm Post subject: Reply with quote
Last season the Heat were lowest ranked on recorded at the rim shots too so it is not just a this season with the big 3 issue. And the Heat were 28th on at the rim shots during Riley's last season as coach too. But there has been some change. Whereas they were bottom 10 on FTM/FGA for several years running they got to average last season and now are 3rd best. On combined at the rim shots and FT opportunities they are better but still below average on the combined. If they wanted to get more high percentage at the rim and FT opportunities they haven't made that much progress and I'd want to review the offensive design and defensive responses to it.
Last season Toronto was 22nd, Cavs 16th. This season it is Toronto 2nd, Cavs 16th. What did Toronto do right here? And even though it is not as dramatic how did the Cavs avoid fall-off?
Toronto without Bosh more than offset the loss of Bosh. Both Johnson and Bargnani increased their per minute at the rim shots some. Davis helped back-fill a lot for Bosh. Wright gets at the rim shots at almost twice Turkoglu's rate. Bayless 50% better than Jack. Barbosa 2.5 times Belinelli's rate. Kleiza helps a little too. You can get shots at the rim with the right players and I guess the ball in your hands, with inside space to exploit and focus on doing it and apparently do better than your past performance with a big man star at the core of it all.
For the Cavs, Eyenga and Sessions helped a lot. Gibson doubled his at the rim rate, Williams up 50%. Some combo of having the ball, room and effort offset the loss of James.
Last edited by Crow on Mon Feb 21, 2011 11:10 pm; edited 10 times in total
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 70
PostPosted: Mon Feb 21, 2011 4:54 pm Post subject: Reply with quote
Don't have individual lineup dynamics as the year progressed. In terms of overall team efficiency as gathered from the team page at hoopdata....
106.95: first 17 games until players only meeting after loss at Dallas
110.14: next 23 games (22-1) until LeBron got hurt in Clippers game
104.10: next 6 games with assorted injuries/migraine goggles
111.33: last 10 games since getting healthy (9-1)
So, that's 110.14 and 111.33 at the team level during a 31-2 composite since the team meeting at Dallas when everyone was healthy.
There may be an issue here where teams do enough to win rather than doing enough to maximize their efficiency, since winning is what matters. And, some of that involves the lineups featuring the big three because they might back off in terms of intensity when they get comfortable leads. Save the horsepower for when you need it. Or, don't even use your maximum horsepower unless you need it. If the big three spead the ball around more with a lead, or run a bit more clock before forcing up a less than ideal shot...then maximum efficiency isn't likely.
Maybe this Miami group is similar to 2008-09 Cleveland because both were reaching the point that it took to attain comfortable victory margins...and there's no practical reason to move much beyond that.
Or, some of the assumptions some had before the season about offenses may have been false, or about opposing defenses abilities to disrupt offenses.
Oh, what's the latest on aging patterns? Are NBA players peaking at 26 and starting their decline phases then. Is there a sustained plateau over a few numbers? Or, is it up a hill then down a hill...with the "down" beginning at 26, and 29 being decidedly over the hill?"
Back to top
View user's profile Send private message
horsecow
Joined: 01 Dec 2009
Posts: 12
Location: Iowa City, IA
PostPosted: Mon Feb 21, 2011 4:59 pm Post subject: Reply with quote
I was curious so I looked at Miami's offensive efficiency and offensive efficiency adjusted for quality of defense over time, and I don't see any upward trend from October to now: http://jamerchant.wordpress.com/2011/02 ... -learning/
I'd be interested to see some numbers on pick and roll efficiency with both Wade and James in the game (as opposed to Wade with someone else/James with someone else), since that has been such a core concept for both players in the past. The configuration of the P&R can get pretty messy with certain line-ups for the Heat: it's James/Wade with the ball at the top, which means Wade/James is one of the spot-up shooters spacing the court and Arroyo, Chalmers or Miller is the other. With Ilgauskas/Bosh setting the pick and Bosh/Ilgauskas on the weak side, there's not a lot of solid perimeter (i.e. 3pt) shooting out there.
ANother thing: I know Bosh is a better overall offensive player than Varejao, but he does lack some offensive skills Varejao had in spades: He is a little too willowy to really set effective screens a la Varejao or Ilgauskas, and his willowiness also doesn't make him a great offensive rebounder (his OR% numbers are down this year (6.1% down from ~8.5%).
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 70
PostPosted: Mon Feb 21, 2011 5:23 pm Post subject: Reply with quote
Does the trend line inch upward if you take out the games from 1/13 to 1/28 when they weren't all in the lineup at the same time? Or, if you start at Dec 1st (giving them a month to adjust to each other) and then exclude the shorthanded games? May not move it much. Just wondering....
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Mon Feb 21, 2011 6:17 pm Post subject: Reply with quote
The best raw team +/- per minute for a PG - regular C pairing is Arroyo with Z at about +10 per 48 minutes. They use that for about half of Z's time. Chalmers-Dampier is equally as good by the raw data. Chalmers with Z is only half as good by the raw data. Anthony with House is great, with Chalmers half as strong. Anthony or Dampier with Arroyo I don't know exactly but they must not be strong. Can't immediately see the less good or bad PG-C combos but the team should be on top of that and probably somewhat responsive to pair performance by the end of the season.
Bosh at center for a bit over 200 minutes is +16 per 48. Maybe you don't do it in the regular season, but I'd up use of it a lot in the playoffs as long as match-ups aren't worse than average for it. Haslem's return could make this easier though his Adjusted +/- recently varies considerably by method (traditional and RAPM) and timeframe and what he will be like upon return is unknown.
Wade at PG probably isn't a big part of the answer, at just +3.6 per 48, less than half as good as team average, though this is raw like everything else. Adjusted +/- for pairs would probably help some in the review of the situation.
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Wed Feb 23, 2011 4:51 pm Post subject: Reply with quote
Based on raw +/- when the big 3 on the court together they are on average +15 per 48 minutes. They have played together about 22 minutes per game. In the playoffs they should try to increase that.
For pairs without the 3rd guy being on the court Wade-Bosh is a nice enough +6 per 48 but the other pairs are actually negative on raw +/-. James-Bosh without Wade is -2, James-Wade without Bosh is -7.5 per 48. Wade on the court without James or Bosh is also -7.5. Organize the rotation to minimize the time of these negative sets and try to pick the right times with more match-up analysis.
Adjusted +/- pair data would be better though.
I'd hope Miami knows their pair without the 3rd guy data. But how many of their opponents do? And if they do, would they use it to strategize getting one of the big 3 in foul trouble more than otherwise? I would and perhaps especially Bosh.
Have any public articles notably touched on Heat pairs without the 3rd guy topic even in general terms? It is fairly small minutes and the flipside of the big obvious story of the 3 together on the court but it is a part of the full story.
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Sun Feb 27, 2011 3:58 pm Post subject: Reply with quote
Wayne Winston generously worked up the Heat Adjusted +/- data for pairs.
http://waynewinston.com/wordpress/?p=1072
Back to top
View user's profile Send private message
Mogilny
Joined: 05 Aug 2010
Posts: 25
PostPosted: Mon Feb 28, 2011 7:32 pm Post subject: Reply with quote
Andrew Gelman posts a Heat efficiency plot with confidence intervals included.
http://www.stat.columbia.edu/~cook/mova ... t_map.html
Back to top
View user's profile Send private message
huevonkiller
Joined: 25 May 2010
Posts: 15
Location: Miami, Fl
PostPosted: Sat Mar 19, 2011 4:15 pm Post subject: Re: more Heat, less Heat Reply with quote
schtevie wrote:
Here is the benchmark comparison exemplifying the (relative) failure. In Cleveland in 2008-09, the top two most used lineups featured James and Ilgauskas. Rounding these out were Mo Williams, Delonte West, and either Ben Wallace or Anderon Varejao. These had an Off Rtg of 6.6 and 7.1 over the league average, respectively.
Compare this to this year's most used lineup for the Heat. Added to the same James and Ilgauskas are Wade, Bosh, and Arroyo. They have an Off Rtg of 7.1 over the league average. (And the swapping of Chalmers for Arroyo shows much worse.)
Not to denigrate the offensive prowess of Wallace and Varejao, but this result simply should not be. Diminishing returns, indeed.
Is part of the explanation that the similarity really isn't. That the Heat have been significantly improving over the course of the season? Does anyone know of a source which shows a time plot of Off Rtg? The Heat Index has a recent suggestive plot of a secular decrease in long 2s, and perhaps that is part of the story.
What is the story?
Sorry but again, I think someone is overstating the Heat's problems.
There's nothing wrong with the 08-09 Cav's offense during their prime. They were a completely different team in the post-season, notably their defense is overrated and Mo Williams is a joke.
The 2011 regular season (when healthy)exposed the Cavaliers supporting cast for the fraud they are on defense.
No one knows what the diminishing returns are, because the post-season hasn't started
Author Message
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Tue Feb 15, 2011 6:17 pm Post subject: more Heat, less Heat Reply with quote
So, the season's well past half over, and I thought it might be worth focusing a bit of attention on the Heat. Why I write is that I am looking for the best explanation as to why the Heat have done such a bad job integrating the offensive talents of James, Wade, and Bosh. To me this is more than a bit of a mystery.
Here is the benchmark comparison exemplifying the (relative) failure. In Cleveland in 2008-09, the top two most used lineups featured James and Ilgauskas. Rounding these out were Mo Williams, Delonte West, and either Ben Wallace or Anderon Varejao. These had an Off Rtg of 6.6 and 7.1 over the league average, respectively.
Compare this to this year's most used lineup for the Heat. Added to the same James and Ilgauskas are Wade, Bosh, and Arroyo. They have an Off Rtg of 7.1 over the league average. (And the swapping of Chalmers for Arroyo shows much worse.)
Not to denigrate the offensive prowess of Wallace and Varejao, but this result simply should not be. Diminishing returns, indeed.
Is part of the explanation that the similarity really isn't. That the Heat have been significantly improving over the course of the season? Does anyone know of a source which shows a time plot of Off Rtg? The Heat Index has a recent suggestive plot of a secular decrease in long 2s, and perhaps that is part of the story.
What is the story?
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Tue Feb 15, 2011 8:56 pm Post subject: Reply with quote
An offensive rating of 114 is pretty good. The current Heat lineup with that rating is about 10th best in the league of the 40 used over 200 minutes. The two from the Cavs in 2008-9 were about 10th best for the top 50 most used lineups.
The current Heat lineup is notably worse than those Cavs lineups on net turnovers according to the 5 man data at 82 games. Not sure if or how much of that difference is coming on offense though.
The game level team efficiency data is available at this link and you could compile how you want or it is probably also in the motion chart section there too (or the motion chart at DSMok1's site).
http://hoopdata.com/teamgl.aspx?team=MIA
Back to top
View user's profile Send private message
greyberger
Joined: 27 Sep 2010
Posts: 50
PostPosted: Wed Feb 16, 2011 12:52 am Post subject: Reply with quote
The 2010 Cavs with Lebron on the court shot 54.8% eFG (+6.4% compared to off court) and made 20 FTs per 48 (+5 versus off).
The 2010 Heat with Wade shot 51.5% eFG (+6.8% versus off) and made 19 FTs per 48 (+6 versus off). The Toronto units with Bosh in also got about 5 more FTs made per 48, FWIW.
The Heat 'starter' lineups this year shoot ~52% and make ~23 FTs per 48. Is that better than the 2010 Heat, or worse than the 2010 Cavs? The team as a whole is about 6th best in eFG, 3rd best in offensive FTR. What's the right comparison here?
What would it take to exceed expectations? 60% eFG and the same free throw rate?
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Wed Feb 16, 2011 7:01 am Post subject: Reply with quote
An Off Rtg of 114 is pretty good. It's, of course, simply zero improvement on what was.
Is this not more than a little bit surprising? You swap out Williams, West (decent offensive players) and a millstone offensive PF/C and replace them with Arroyo, the best (?) offensive shooting guard, and one of the best offensive power forwards in the league, and the result is....nothing?
The numbers are a bit dated but compare the sums from Steve and Aaron's 2007-08 Offensive +/-s. A swap of -5.7 for +11.1, a net change of +16.8. One might have expected a slight improvement.
That there is but one ball to share suggests that there would be some diminishing returns, but what explains this?
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3597
Location: Hendersonville, NC
PostPosted: Wed Feb 16, 2011 8:05 am Post subject: Reply with quote
Superstars each have a team built around them. Put them on one team, and it's a different environment. The best teams have been a superstar and some strong supporting players. On occasion, 2 such players (Robinson and Duncan, Kareem and Magic) .
The nearest similar case might be the assembling of the 2008 Celtics. They did OK right off the bat, but those guys played distinct positions, unlike this bunch. LeBron's kind of bouncing around between PF and PG.
So now we have a sample of 1.5 such seasons of gathering +/- superstars.
And with the Celts, a few others emerged in significant roles: Rondo, Perkins, Posey, House .
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Wed Feb 16, 2011 9:09 am Post subject: Reply with quote
Mike, I like seeing the conventional wisdom overturned as much as next guy (probably more, actually) but I think you assign a wee bit much to the importance, make that the centrality, of role players. PR aside, why bother to go to all the trouble assembling a Dream Team (apologies for using this horrible term) for the Olympics, if a couple of superstars and some solid role players would have equal effect?
The predicate is that these three guys had histories of being very efficient at being able to score, distribute, and not turn the ball over. It wasn't an illusion that they were offensive superstars. (The Adj +/- numbers are but one reference.) Accordingly, in combination, the expectation wasn't and isn't (?) that there should be no net offensive improvement.
Back to top
View user's profile Send private message
Ilardi
Joined: 15 May 2008
Posts: 265
Location: Lawrence, KS
PostPosted: Wed Feb 16, 2011 11:34 am Post subject: Reply with quote
Schtevie,
I think there are two plausible explanations (not mutually exclusive) for the conundrum you've identified:
1) As you know, APM values simply reflect a player's impact in a given role; thus, when you add two or more superstars to the same lineup and thereby cause their respective roles to change substantially (as in the case of Wade and LeBron, both of whom were formerly in the role of single-handedly dominating the offense), you likely wind up with substantially diminished contributions from each;
2) Adjusting/developing an offense to capitalize efficiently on the presence of multiple superstars with overlapping roles may take more time than generally appreciated . . . thus, it would be highly informative to look at time-trend analyses of the Heat's offensive efficiency this year (especially on a lineup by lineup basis) to see if indeed there is a trend toward greater efficiency as the season has progressed.
Also, of course, role players really do enter into the equation: It may well be that advancing beyond an already-lofty efficiency of 114 with James/Wade/Bosh requires the presence of two highly specialized role players with skill sets that may or may not be present on the Heat's current roster.
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Wed Feb 16, 2011 12:50 pm Post subject: Reply with quote
What do the top 50 minute lineups over 114 on offensive efficiency from the last 2 seasons look like in general?
About 2/3rds have a bigtime assist PG.
About 2/3rds have a notable post scorer or threat.
About 2/3rds have 3+ decent 3 pt. shooters on the court at the same time.
2/3rds of the lineups were from 8 teams.
About 87% were playoff level teams.
It doesn't appear that many of these teams are fastbreak leaders.
The teams with 2 or more such lineups include Boston, Dallas, Denver, Houston, Clippers, Miami, Lakers and OKC. 6 of the 8 Coaches played PG in the NBA, Phil Jackson and Eric Spoelstra did not.
Arroyo is far from a bigtime assist PG this season. They do get assists from others but they are 27th on team assists per game.
Miami gets the least team shots at the rim per game in the league and Bosh is only getting 3 per game.
Arroyo, Wade, James are three 3 point shooters, but 2 are below average on 3pt FG%. As a team they are 8th on 3pt FG% but only 14th on attempts. They are #2 on long 2s (16-23 ft.) per game.
Looking at the top Heat lineup it appears that most of stints where it had really poor offensive efficiency the opponents had a shotblocker or at least a big / tough dude inside.
With the 08-09 Cavs, James had a bit more help on assists from Williams, and better 3 point shooting but again very low shots at the rim (28th). With Shaq in 09-10 only moved them up to 16th on shots at the rim (and team assists went up too). He probably needs a "bigger" post scorer or more frequent drivers, more good passing and more 3 point shooting. Getting to 114 on offensive efficiency without being stronger on these criteria, without a former NBA PG as coach, without team familiarity from last season may be a semi-accomplishment. (With the Cavs James also lacked a former NBA PG as Coach as did Bosh and Wade with their former teams.)
Last edited by Crow on Wed Feb 16, 2011 1:27 pm; edited 1 time in total
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 68
PostPosted: Wed Feb 16, 2011 1:26 pm Post subject: Reply with quote
Crow, what do you think about trey volume here? I know that's something you've followed closely (you follow everything closely!).
Noticed:
08-09 Cleveland: 8.8 treys per 100 possessions
10-11 Miami: 7.4 treys per 100 possessions
Maybe it's hard to crack a historic efficiency at just 6.9 treys per game, particularly if you're not getting an inside/outside mix (you mentioned Miami not getting many shots at the rim). Just a thought.
Might connect back to Schtevie's earlier thread about coaches not maximizing the trey influence. Might be a dead end. Figured I'd throw it out there because I had remembered that 08-09 team having some perimeter threats LeBron could kick out to. That 1.4 treys per 100 possessions is at least a noticeable difference between the two teams in the discussion...
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Wed Feb 16, 2011 1:32 pm Post subject: Reply with quote
Thanks Jeff. I agree with you about at least looking hard at the role of 3 point shooting on top efficiency teams and top performing teams in the playoffs.
'08-09 Cleveland was 3rd on made 3s per game. One leg of the inside / 3 pt / assist combo but just one leg. (Positing that this combo matters to see what is found.)
'08-09 champs Lakers were a bit below average on 3 pt makes but compensated with near top on team assists and top 10 on shots at the rim.
Last season the Lakers weren't top 10 at any of these regular season and don't appear special on at least 2 of these in the playoffs either but made it with experience, a top 4 overall playoff offense (aided by #1 offensive rebounding rate) and 3rd best playoff shot defense. They did it another way outside being strong on these 3 criteria. (They were high on 3 pt attempts in the playoffs but shot below average on them.)
In '07-08 the Celtics were top 10 on 2 legs (assists and 3s).
In '06-07 the Spurs essentially had all 3 (one 11th place).
This season the Lakers are balanced and nearly have all 3. Denver also almost have 3. Orlando, Dallas and Atlanta 2. Boston just 1. Miami almost 1. OKC and Chicago not close to even one.
3 of the last 4 champs were top 10 on 2. Whether it will be the case again this season will help build or weaken the case for the 3 criteria being pretty important.
Last edited by Crow on Wed Feb 16, 2011 2:31 pm; edited 4 times in total
Back to top
View user's profile Send private message
horsecow
Joined: 01 Dec 2009
Posts: 12
Location: Iowa City, IA
PostPosted: Wed Feb 16, 2011 2:08 pm Post subject: Reply with quote
I wonder if free throw shooting has something to do with this. This year, all of the big three have gotten to the line less:
FTA
LeBron down from ~10.0 to 8.8
Wade down from ~9.5 to 9.0
Bosh down from ~8.4 to 5.7
= down from ~27.9 FTA to 23.5 FTA
To equal the same number of points, their collective FT shooting would have to improve considerably, but it's actually gotten worse (especially Wade's). Free throw shooting is the one area where having a superstar teammate doesn't make completing the shot any easier, so if these three are getting to the line less (and the team's improved offensive efficiency in other areas isn't compensating), the team is scoring less.
Why they are getting to the line less is another question, though. Lack of floor spacing for Bron/Wade drives? A "superstar foul" quota by league refs?
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Wed Feb 16, 2011 2:20 pm Post subject: Reply with quote
Free throws could be a 4th leg to a strong offense... and turnovers and offensive rebounding could 5th and 6th if you want to stretch back out beyond field shooting and the related assists. Maybe some teams compensate for not having 2 of the 3 legs by having the 4th or 5th or 6th. '09-10 Lakers were 0 for 3 on the 3 initially name criteria but go 2 for 6. Enough I guess. How much the mix of ingredients matters in strong offensive efficiency and what combo has the best track record would take more research / review and writing. There are some old threads where I and a few others looked at champ and top contenders by the 4 Factors and overall efficiencies. And I now see there was a 2010 JQAS article on this very topic.
If you look at the 6 offensive criteria, Lakers nearly have 4, Denver almost 5. Orlando 3, Dallas and Atlanta still just 2 Boston almost 2. Miami almost 2. OKC 1 and Chicago almost 1.
Of course defense matters too.
I think it was useful to quickly compare the Heat's offense with these others but back to focus on the Heat themselves if you wish...
Back to top
View user's profile Send private message
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Mon Feb 21, 2011 1:47 pm Post subject: Reply with quote
I am a bit bewildered, perhaps even bothered, by the general sense - if this is true - that there is nothing really surprising about the Heat's offensive performance this year: An Offensive Rating of 114 is a very good start; Shallow learning curve and all that; Nothing in the historical record suggesting that they should have been special. Eh.
I put the possibility aside, for now, that the average performance over the year is not representative, that the Heat offense today is actually far more potent now than at the beginning of the season. I do so because I don't think it's true. Having intermittently followed the relevant statistic over the season, I have no sense that there has been a trend showing improved offensive efficiency over time. I could be wrong. (Perhaps Tom H will choose to anchor the discussion at the eponymous blog with this paramount Heat Index.)
If this impression is correct, however, there is then the issue of explaining and contextualizing the Heat's offensive performance. And my point is that what has transpired in this regard is completely at odds with what was anticipated. The degree of diminishing returns was (correctly) the focus of preseason discussion; the possibility of no returns (correctly) not being part.
In addition to their reputations of having high (at least, not low) basketball IQs, Wade, James, and Bosh going into this season each had a clear record of being a superior ball handler, scorer, and passer. The Heat also acquired 3 point shooting specialists. The direct implication of these facts was that opposing defenses would be stretched and in turn each component of the Heat offense would be expected to see less pressure. That there is one ball to be shared only implied that the gross statistical totals (Pts, PPG, Assists, etc.) of one or more players would be expected to decrease. The overall efficiency of the offense however was necessarily expected to increase. But there is no compelling evidence that it has (compared to a reasoned baseline of expectations) and much that it hasn't.
There is the highly suggestive evidence provided by the benchmark Cleveland Ilgauskas/James starting lineup of 2008-09. Expanding on this, the minute weighted average Offensive Rating of all Ilgauskas/James lineups, in 2008-09, was 114.3, 6.0 points above the NBA average. By contrast, this year's I/J average lineups have yielded 113.1 points per 100 possessions, 6.1 points above this years NBA average. Again, no change. Additionally, the average of all I/J/W/B lineups this year (there have been five) have been 114.2, 7.3 points above the NBA average. Compare that to the top ten I/J lineups (in terms of time on court) from 2008-09. These clocked in at 115.6, 7.3 points above that year's NBA average. Again, no change. So, it isn't just a starting lineup phenomenon.
What then accounts for the dysfunction in Miami, holding Ilgauskas constant? Interestingly and surprisingly (I suppose) it has been the addition of Wade and extraction of James from the Ilgauskas lineups that has most improved offensive performance (though this should be taken with a few grains of salt as it is an observation based upon a small sample of infrequently used lineups).
For all lineups (693.47 minutes worth) that minimally include I/J/ and B (that is when Wade is not necessarily on the floor), these yield an Off Rtg of 113.1. Then of these, in the 53.38 minutes when Wade isn't on the floor, Off Rtg plummets to 99.0. Similarly, the I/J/W then "not Bosh" numbers are 113.7 and 109.0 (in 707.26 and 67.17 minutes respectively). So, Bosh is apparently positive. Then finally, there is the benefit of getting rid of James. The I/W/B and "not James" numbers are 114.8 and 125.1 (in 677.69 and 37.6 minutes respectively).
The conclusion that the Heat have produced, um, sub-par offensive performance this year does not hinge on Ilgauskas however. The basic and general evidence can be found in the year on change in the effectiveness of the three stars. A list of some unanticipated results:
(1) The Assist %ages of James, Wade, and Bosh have all decreased (by 7.1 %age points on average). Well above average passers, with anticipated greater opportunities to make assists, suddenly unable to pass? How does that happen?
(2) The Turnover %ages of James and Wade have increased (a bit) despite decreased individual usage. With what story of offensive organization is that compatible?
(3) But more comprehensively, there have been significant decreases in the offensive efficiency of James and Bosh, offset only in part by a smaller improvement by Wade.
Consider a play efficiency measure, inclusive of the effect of turnovers, call it TS*%, equaling (Pts/2)/(FGA + 0.44*FTA + TO). Compared to last year, Wade's TS*% has gone from 0.494 to 0.504, a 1.05 %age point improvement. By contrast, James' has gone from 0.530 to 0.496, a drop of 3.43% points. And Bosh's has dropped 1.81 %age points, from 0.529 to 0.511.
These results, it should be repeated, are incompatible with inferences from basic offensive theory (assuming increased scoring potential, of course): individual scoring opportunities would necessarily collectively drop, but the realized efficiency should have increased.
James and Bosh haven't been injured, have they? Looking at the year on shot clock comparisons at 82games shows no difference to suggest that this might be a mitigating factor. (And by the way, let me take this opportunity to, yet again, thank Aaron and Roland for the wonderful resources that they provide to the community. It would be a much poorer world....) The coach seems as aware as any, regarding statistical issues, yet he somehow hasn't been able to combine these extraordinary resources in a way that yields a (non-trivial) net improvement.
What is left to explain the result (again, assuming no underlying time trend in the data) an appeal to randomness?
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 816
PostPosted: Mon Feb 21, 2011 2:21 pm Post subject: Reply with quote
Restating and expanding points made above:
1. The Heat are below average in the league on offense from the PG position and not very directive with passing from that traditional hub. And very low on team assists. Maybe 2 heavy ball-handling wings with weak PGs is not that overpowering in pressure on the defense.
2. Just average on team attempts from the 3 point zone. Miller being back helps some but hasn't make them consistently elite on 3 pt frequency yet. Hard to keep Jones or Miller on the court a lot with Wade and James playing the same positions most of the time. Wade and James are averaging 4.6 3 pt attempts in 77 combined minutes per game over the last 10 down from about 6.5 for the season. I'd think that is low for starting wing combos. In fact, using hoopdata, it appears that starting wing combos at the same minute level would average about eight 3 point attempts, so what they get from House, Miller and Jones is needed to help get them back to average.
3. 30th at shot attempts in the paint (and presumably very low even in lineups with the big 3 and Z since they are dead last). If I was the defending them I'd guard the paint pretty hard and live with them trying to win with mid-range or 3 point shots.
All things that make them easier to defend I'd think. And having a Coach who didn't play in the NBA, especially at PG may limit their greatness on offense too. I expected some issues from points 1 and 2 but I didn't expect the 3rd point would be as bad a problem with Bosh and 2 drivers.
Last edited by Crow on Mon Feb 21, 2011 4:09 pm; edited 3 times in total
Back to top
View user's profile Send private message
schtevie
Joined: 18 Apr 2005
Posts: 411
PostPosted: Mon Feb 21, 2011 2:49 pm Post subject: Reply with quote
Crow wrote:
If being below average in the league on offense from the PG position and not very directive with passing from that traditional hub, being average on team attempts from the 3 point zone and being 30th at shot attempts in the paint (and presumably very low even in lineups with the big 3 and Z since they are dead last), all things that make them easier to defend I'd think, and having a Coach who didn't play in the NBA, especially at PG, doesn't help your search enough I don't think I have anything extra to add at the moment.
Not sure what to make of the "Coach who didn't play (PG) in the NBA" correlation. Why should we believe that that should have a functional relationship with outcomes, offensive or defensive?
And it was always expected that offensive distribution through the PG wasn't going to be a feature of the offense. If you are importing three really good passer/scorers, you neither need nor want that function to be performed by another. Surely (?) if the ball was going more through Arroyo or Chalmers, things would be worse, not better.
Finally, that the offense is not getting enough high quality shots (3s and shots in the paint) is an alternative expression of the offense sucks (compared to potential) phenomenon not an independent explanation. They surely (?) have the raw material that gives them the potential to score from far as well as near.
I do note that Tom H has a nice bit today about how Bosh has regressed and could improve on the latter account.
recovered page 2
Author Message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Mon Feb 21, 2011 2:53 pm Post subject: Reply with quote
It is more a starting point than conclusive answer, but I noted above that of the 8 teams with 2 or more big minute lineups over 114 on offensive efficiency this season 6 of the 8 Coaches played PG primarily in the NBA. Eric Spoelstra did not. About 40% of coaches played PG in the NBA so this group has almost double the % representation on teams with 2 or more big minute lineups over 114 on offensive efficiency than expected. It could be looked at deeper.
Knowing from firsthand experience how things really are on the NBA court and what works best might be worth something in designing and managing an offense. It is not a strange suggestion, in my mind, to think that not having that experience could either hamper very high offensive performance or at least the coaching not be helping as much as it might with such experience.
If a Coach without NBA PG experience doesn't get the Heat a title in a year or two I'd expect we will see a new Coach with at least NBA experience and perhaps with NBA PG experience.
If a Coach with either version of NBA experience were to take over the Heat in the future you could get a team specific comparison of the height of the top lineup and overall offensive efficiency achieved by the Coach without such experience to with.
Last edited by Crow on Mon Feb 21, 2011 10:51 pm; edited 7 times in total
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3627
Location: Hendersonville, NC
PostPosted: Mon Feb 21, 2011 3:00 pm Post subject: Reply with quote
I'll suggest that at age 26, both LeBron and Bosh can not be expected to get any better and may well be in decline. Wade at 29 is decidedly over the hill.
Chalmers is 24, Joel Anthony is 28, and everybody else on the team is 30 or older. Z, Juwan, and Damp all have actual gray in their beards.
While playoff experience will prove to be important in the playoffs, the major players here have had a lot of (undue?) playoff pressure in their young lives. Probably a trade-off there.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Mon Feb 21, 2011 3:34 pm Post subject: Reply with quote
Last season the Heat were lowest ranked on recorded at the rim shots too so it is not just a this season with the big 3 issue. And the Heat were 28th on at the rim shots during Riley's last season as coach too. But there has been some change. Whereas they were bottom 10 on FTM/FGA for several years running they got to average last season and now are 3rd best. On combined at the rim shots and FT opportunities they are better but still below average on the combined. If they wanted to get more high percentage at the rim and FT opportunities they haven't made that much progress and I'd want to review the offensive design and defensive responses to it.
Last season Toronto was 22nd, Cavs 16th. This season it is Toronto 2nd, Cavs 16th. What did Toronto do right here? And even though it is not as dramatic how did the Cavs avoid fall-off?
Toronto without Bosh more than offset the loss of Bosh. Both Johnson and Bargnani increased their per minute at the rim shots some. Davis helped back-fill a lot for Bosh. Wright gets at the rim shots at almost twice Turkoglu's rate. Bayless 50% better than Jack. Barbosa 2.5 times Belinelli's rate. Kleiza helps a little too. You can get shots at the rim with the right players and I guess the ball in your hands, with inside space to exploit and focus on doing it and apparently do better than your past performance with a big man star at the core of it all.
For the Cavs, Eyenga and Sessions helped a lot. Gibson doubled his at the rim rate, Williams up 50%. Some combo of having the ball, room and effort offset the loss of James.
Last edited by Crow on Mon Feb 21, 2011 11:10 pm; edited 10 times in total
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 70
PostPosted: Mon Feb 21, 2011 4:54 pm Post subject: Reply with quote
Don't have individual lineup dynamics as the year progressed. In terms of overall team efficiency as gathered from the team page at hoopdata....
106.95: first 17 games until players only meeting after loss at Dallas
110.14: next 23 games (22-1) until LeBron got hurt in Clippers game
104.10: next 6 games with assorted injuries/migraine goggles
111.33: last 10 games since getting healthy (9-1)
So, that's 110.14 and 111.33 at the team level during a 31-2 composite since the team meeting at Dallas when everyone was healthy.
There may be an issue here where teams do enough to win rather than doing enough to maximize their efficiency, since winning is what matters. And, some of that involves the lineups featuring the big three because they might back off in terms of intensity when they get comfortable leads. Save the horsepower for when you need it. Or, don't even use your maximum horsepower unless you need it. If the big three spead the ball around more with a lead, or run a bit more clock before forcing up a less than ideal shot...then maximum efficiency isn't likely.
Maybe this Miami group is similar to 2008-09 Cleveland because both were reaching the point that it took to attain comfortable victory margins...and there's no practical reason to move much beyond that.
Or, some of the assumptions some had before the season about offenses may have been false, or about opposing defenses abilities to disrupt offenses.
Oh, what's the latest on aging patterns? Are NBA players peaking at 26 and starting their decline phases then. Is there a sustained plateau over a few numbers? Or, is it up a hill then down a hill...with the "down" beginning at 26, and 29 being decidedly over the hill?"
Back to top
View user's profile Send private message
horsecow
Joined: 01 Dec 2009
Posts: 12
Location: Iowa City, IA
PostPosted: Mon Feb 21, 2011 4:59 pm Post subject: Reply with quote
I was curious so I looked at Miami's offensive efficiency and offensive efficiency adjusted for quality of defense over time, and I don't see any upward trend from October to now: http://jamerchant.wordpress.com/2011/02 ... -learning/
I'd be interested to see some numbers on pick and roll efficiency with both Wade and James in the game (as opposed to Wade with someone else/James with someone else), since that has been such a core concept for both players in the past. The configuration of the P&R can get pretty messy with certain line-ups for the Heat: it's James/Wade with the ball at the top, which means Wade/James is one of the spot-up shooters spacing the court and Arroyo, Chalmers or Miller is the other. With Ilgauskas/Bosh setting the pick and Bosh/Ilgauskas on the weak side, there's not a lot of solid perimeter (i.e. 3pt) shooting out there.
ANother thing: I know Bosh is a better overall offensive player than Varejao, but he does lack some offensive skills Varejao had in spades: He is a little too willowy to really set effective screens a la Varejao or Ilgauskas, and his willowiness also doesn't make him a great offensive rebounder (his OR% numbers are down this year (6.1% down from ~8.5%).
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 70
PostPosted: Mon Feb 21, 2011 5:23 pm Post subject: Reply with quote
Does the trend line inch upward if you take out the games from 1/13 to 1/28 when they weren't all in the lineup at the same time? Or, if you start at Dec 1st (giving them a month to adjust to each other) and then exclude the shorthanded games? May not move it much. Just wondering....
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Mon Feb 21, 2011 6:17 pm Post subject: Reply with quote
The best raw team +/- per minute for a PG - regular C pairing is Arroyo with Z at about +10 per 48 minutes. They use that for about half of Z's time. Chalmers-Dampier is equally as good by the raw data. Chalmers with Z is only half as good by the raw data. Anthony with House is great, with Chalmers half as strong. Anthony or Dampier with Arroyo I don't know exactly but they must not be strong. Can't immediately see the less good or bad PG-C combos but the team should be on top of that and probably somewhat responsive to pair performance by the end of the season.
Bosh at center for a bit over 200 minutes is +16 per 48. Maybe you don't do it in the regular season, but I'd up use of it a lot in the playoffs as long as match-ups aren't worse than average for it. Haslem's return could make this easier though his Adjusted +/- recently varies considerably by method (traditional and RAPM) and timeframe and what he will be like upon return is unknown.
Wade at PG probably isn't a big part of the answer, at just +3.6 per 48, less than half as good as team average, though this is raw like everything else. Adjusted +/- for pairs would probably help some in the review of the situation.
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Wed Feb 23, 2011 4:51 pm Post subject: Reply with quote
Based on raw +/- when the big 3 on the court together they are on average +15 per 48 minutes. They have played together about 22 minutes per game. In the playoffs they should try to increase that.
For pairs without the 3rd guy being on the court Wade-Bosh is a nice enough +6 per 48 but the other pairs are actually negative on raw +/-. James-Bosh without Wade is -2, James-Wade without Bosh is -7.5 per 48. Wade on the court without James or Bosh is also -7.5. Organize the rotation to minimize the time of these negative sets and try to pick the right times with more match-up analysis.
Adjusted +/- pair data would be better though.
I'd hope Miami knows their pair without the 3rd guy data. But how many of their opponents do? And if they do, would they use it to strategize getting one of the big 3 in foul trouble more than otherwise? I would and perhaps especially Bosh.
Have any public articles notably touched on Heat pairs without the 3rd guy topic even in general terms? It is fairly small minutes and the flipside of the big obvious story of the 3 together on the court but it is a part of the full story.
Back to top
View user's profile Send private message
Crow
Joined: 20 Jan 2009
Posts: 826
PostPosted: Sun Feb 27, 2011 3:58 pm Post subject: Reply with quote
Wayne Winston generously worked up the Heat Adjusted +/- data for pairs.
http://waynewinston.com/wordpress/?p=1072
Back to top
View user's profile Send private message
Mogilny
Joined: 05 Aug 2010
Posts: 25
PostPosted: Mon Feb 28, 2011 7:32 pm Post subject: Reply with quote
Andrew Gelman posts a Heat efficiency plot with confidence intervals included.
http://www.stat.columbia.edu/~cook/mova ... t_map.html
Back to top
View user's profile Send private message
huevonkiller
Joined: 25 May 2010
Posts: 15
Location: Miami, Fl
PostPosted: Sat Mar 19, 2011 4:15 pm Post subject: Re: more Heat, less Heat Reply with quote
schtevie wrote:
Here is the benchmark comparison exemplifying the (relative) failure. In Cleveland in 2008-09, the top two most used lineups featured James and Ilgauskas. Rounding these out were Mo Williams, Delonte West, and either Ben Wallace or Anderon Varejao. These had an Off Rtg of 6.6 and 7.1 over the league average, respectively.
Compare this to this year's most used lineup for the Heat. Added to the same James and Ilgauskas are Wade, Bosh, and Arroyo. They have an Off Rtg of 7.1 over the league average. (And the swapping of Chalmers for Arroyo shows much worse.)
Not to denigrate the offensive prowess of Wallace and Varejao, but this result simply should not be. Diminishing returns, indeed.
Is part of the explanation that the similarity really isn't. That the Heat have been significantly improving over the course of the season? Does anyone know of a source which shows a time plot of Off Rtg? The Heat Index has a recent suggestive plot of a secular decrease in long 2s, and perhaps that is part of the story.
What is the story?
Sorry but again, I think someone is overstating the Heat's problems.
There's nothing wrong with the 08-09 Cav's offense during their prime. They were a completely different team in the post-season, notably their defense is overrated and Mo Williams is a joke.
The 2011 regular season (when healthy)exposed the Cavaliers supporting cast for the fraud they are on defense.
No one knows what the diminishing returns are, because the post-season hasn't started