DFDepressed FanDepressed Fan

All  

Sixers

, all the time

Who Gets Coach Collins Love?

The second chart will be fixed in a few minutes to reflect the most recent data. It really shows how valuable Jrue, Thad and Brand have been to this team- and the positive impact Vuce and Lavoy have had so far.

No sure why Iguodala does not have a greater net impact when he is on the floor. Definitely a case where the numbers and subjective observations do not match up. But basically it says that the team does about the same when Iguodala is on the bench as it does when he is on the floor.

One possibility is that Doug somewhat mirrors Andre's minutes with the minutes of the best wing on the opposing team since he is guarding them. I am not positive that actually the case but it could at least partially explain the gap.

That probably has something to do with it.

here are more detailed numbers on Iguodala:

http://basketballvalue.com/player.php?id=52&year=2011-2012

rtg on: +11.33
rtg off: +9.77
So "unadjusted +/-
" is +1.47

But his adjusted +/- is -1.11.

So certainly playing alongside stronger teammates (Jrue and Thad) pulls his adjusted number down. While Lavoy and Vuce get a big adjusted bump since they tend to be on the floor with the other negative players (ET, Meeks, Lou.)

FYI, I corrected and added some of the data above.

Basically, Jrue, Thad and Brand positively impact the team's performance. Iguodala nets about zero based on this metric. While Lou, Meeks and ET all have a negative impact on overall measured team performance. This all correlates pretty well with minutes played.

Nice post TK, confirms some thoughts I've had and debunks others. Like the logic applied here - not too simple or too complicated.

yeah man!
Most math challenges me, but this was easy to get thru and super informative.

Props for bringing something like this to the table. Well done

Question for you tk76 - how does this take into account all of our blowout victories (and two blowout losses to Miami)? The entire 4th quarter in a handful of our games was mostly garbage time, so did the players in at those times hurt or help their +/-?

user-pic
Matt reply to MCT on Feb 14 at 21:19
+/-

That minutes amount is a much smaller amount that you probably think. Nocioni has played 43 of a possible 7010 minutes. That's .6% of the total.

Assume you are referring to Turner playing a lot of minutes in garbage time? He's played 31 minutes with Nocioni all year. That's less than 5% of his total 681.

user-pic
Matt reply to Matt on Feb 14 at 21:28
+/-

Excuse me that total number was all 5 positions, obviously it was too high. Nocioni has played 4% of the minutes this year. The 5% of ET's minutes was right though.

user-pic
MCT reply to Matt on Feb 14 at 21:37
+/-

All three - Meeks, Turner, and Lou. I feel like they have all played more garbage minutes than the others, probably since we don't have any guards deep on the bench.

Turner definitively gets more Graveyard Shift minutes than the other wings. But I think he gets enough non-garbage minutes for his numbers to be valid. But I can't say that is a sure fact. Fair to say it makes his statistical "confident interval" wider than some of the other players. In fact that is the case for all of the players who either play more garbage minutes or have less minutes overall.

Also, the more minutes you play with Thad (or Jrue) the better you +/- will be. Because those to guys appear to be the difference-makers thus far based on these numbers.

Again, the point is that the numbers for Jrue and Thad are so strongly positive while ET, Meeks and Lou are so strongly negative that you can be pretty confident the numbers are telling you something significant. But the numbers lack the precision to compare individuals within those groupings or to make a clear conclusion about someone like Iggy, who is close to zero. Also, the lower minute results for Lavoy and Vuce have to be taken with a grain of salt, because they probably lack the statistical power to be considered valid.

Part of what makes this an interesting exercise for someone with a statistics background is that you can see both how data matches up with you subjective observations, and you can analyze the data to determine what of it is statistically relevant and what might just me random noise.

Also, using charts and graphs helps to get a real feel for what the numbers are saying in broad terms without getting caught up in the details. For example, looking at the minute distribution shows how ET's minutes are really quite close to a large number of players (although his recent numbers are way down.)

user-pic
MCT reply to tk76 on Feb 14 at 22:11
+/-

Definitely some interesting stuff though, thanks for taking the time to post it tk.

user-pic
Matt reply to MCT on Feb 14 at 22:21
+/-

Yeah like I said, only 5% of Turner's minutes have involved Nocioni (Who I feel like is the main sign that it's garbage time). Assume that Meeks and Lou are an even lower percentage. Don't really think that is a large enough percentage to change anything, do you?

I don't know the answer. My sense is that once they get to garbage time the game tend to be played somewhat more even (in fact the graveyard shift often blows leads.) So it is the players who have actually built the leads during relevant parts of the game that get rewarded by a better adjusted +/-.

Again, I think the big take away points are the overall trends of the numbers and their relation to minutes played. The more you look at the details or focus on low minute players the less valid the data will be.

If you are a stats person... the trend that Jrue/Young/Brand are net positives while ET/Meeks/Lou are not is statistically significant. There also is a clear statistical correlation between minutes played and adjusted +/-. A correlation that breaks down once you look at lower minute deep reserve bigs.

And lie I said below, the team's large positive point differential (+9.1) means that although ET/Lou and Meeks each have negative adjusted +/-, the raw +/- is positive for all three players. So the team is winning with them on the floor, but they simply are winning by even more when they are on the bench... if that makes sense. If the team was closer to .500 like in years past, then the adjusted +/- and raw +/- would be closer.

Great job tk76.

The most interesting observation to me is the unbelievably positive impact the Sixers bigs bring to the team. The bigs are largely "bashed" by somewhat educated fans (such as us DF posters) for a variety for issues (mostly issues related to lack of athleticism), while everybody likes the Sixers guards.

The most surprising thing is these numbers show that the team has only one guard that makes a positive impact on the team (Holiday). On the other hand, all 5 rotation bigs have performed fairly well.

user-pic
Just Saying reply to Xsago on Feb 14 at 19:03
+/-

Or the data is flawed or incomplete (as any singular data point will tend to be) because the sixers defensive rebounding woes can be directly attributed to the poor rebounding issues of the bigs you think should be praised.

Interesting...and very nice to not see Nocioni anywhere in those charts!

user-pic
tk76 reply to das411 on Feb 14 at 20:59
+/-

Noce has incredibly negative numbers using this metric, but his minutes are too low to make it valid.

Speaking of people whose initials are DC, Demarcus Cousins has 16 and 10 at halftime, and Derrick Coleman just sprouted a new roll of fat on his neck.

Overall these numbers do need to be taken in the context of the team's overall performance. The team has such a big positive point differential that in most cases a negative adjusted +/- simply means that the team is still successful with a player, just not as incredibly successful as they are on average.

So I don't take the numbers to mean that ET or Lou are bad players or hurt the team. Or that Iguodala is any less important than we think he is. More that the SG/SF's on this team each have their own clear strengths and weaknesses and so the team does not necessarily excel more with one on the floor as compared to the others.

But I do think the numbers highlight how big of an impact Thad and Jrue have (and to some extent Brand.) Both bring a certain skill-set that no one else on the team has. Young with his mismatch capabilities and Jrue with his combination of defense, scoring and shooting. While other backcourt players seem to have strong aspects of their games but lack the total package that Jrue brings.

As for the huge numbers from Vuce and Lavoy- I think that is a bit of a small sample size that is skewed by a few huge runs and probably will regress a bit to average over time.

user-pic
Heard on WIP on Feb 14 at 22:00
+/-

[Ten minutes ago]

Caller named Jay: "If the Sixers are gonna make a trade, I don't want one of these old retread guys like Pierce and them. If I were the Sixers, what I'd try to do is trade Jrue Holiday for a guy like Tyreke Evans."

Rob Ellis: "I'm trying to figure out why the Kings would do that."

[One minute later]

Ellis: "I love it from the Sixers' standpoint, but I can't see the Kings doing it.

[One minute later]

(Next caller disagrees with the trade idea because the Sixers have waited for a PG for so long and Holiday has so much potential)

Ellis: "Yeah, that's true, we often forget he's only 21. But he hasn't really developed the way I'd hoped he would this year ... He doesn't defend. He doesn't play good enough defense on the ball."

[One minute later]

Same Caller: "Where the Sixers really went wrong was the year when they drafted Iguodala, and they passed on Jameer Nelson. That was the true point guard they could have had all those years."

I wonder if people really start following the Sixers if the level of discourse would rise a bit. WIP is always a joke, but Sixers talk on WIP is even more bizarre.

Iguodala and Nelson were in the same draft? I guess they were.

that's some hilarious shit. thx for posting.

well linsanity struck again as he hit the game winning 3
i feel like i would like the stroy more if it wasnt the knicks

I like Lin as a person. As a basketball player, I think he can start for the Knicks and be successful. he lacks the athleticism, defense or explosiveness to go anywhere beyond that IMO.
I just hope when he comes down to earth (and once the scouting report if half filled) and the faithful realize he is not the messiah, it wont turn media ugly. Tough thing playing with all that pressure. Hope the best for the young man.

I agree, it is a nice story and not much beyond that. he does not fundamentally change the NBA landscape like some have tried to suggest in the media.

Certainly, Lin has had a remarkable stretch of initial starts. But had that happened to another sub .500 team (say the Bucks, who had a similar record 1 week ago) would this have even been a story? For example, Ramon Sessions put up some huger numbers in his initial starts- including a 20pt/24 assist game followed by a 25pt/15 assist game and nobody talked about it. I'd say Sessions and Lin were equally low profile.

So I think this is 50% a New York hype machine story (they tried to make Shumpert into a story earlier), 25% a racial and human interest thing and maybe 25% an actual basketball story. Although many fans talk about it like it is 100% a basketball story- as in Lin is as good as... and the Knicks are now contenders.

user-pic
Tray reply to tk76 on Feb 15 at 1:56
+/-

I think Lin set some record for most points in his first 4 starts.

He's got to be pretty close to the record for most turnovers in the first five starts of a career as well.

Since I got accused of loving Alec Burks the other day after posting a mildly complimentary comment about him a month ago, I'm just going to start trollling for him and point out that in his last two games, he has 12 FGA to 16 FTA. On the season, he takes just over 2 shots per free-throw. And, 28 and 17 for Cousins on 20 shots, with 11 combined fouls and turnovers, 3 steals, 2 blocks.

Excellent post tk76; even an analog guy like me was able to digest this easily. Stats in basketball is like a chinese puzzle, you were about to solve it until Iguodala's numbers threw a monkey wrench in there.


Brian: Maybe you can link this post to the Meeks-Turner thread.

Yeah, Brian and I were discussing about how the Iguodala number just does not make all that much sense. Either the numbers are telling us something that are eyeballs are missing, or the stats are somehow flawed. Don't know what is the case with Iguodala.

Another interesting thing to look at is 2-man combos: http://www.nba.com/statistics/plusminus/plusminus_sort.jsp?pcomb=2&season=22011&split=9&team=76ers

In terms of raw +/- the best 2-man combos for the Sixers are:

Jrue-Thad: + 160 in 524 min
Jrue-Iggy: +160 in 800 min
Jrue-Meeks: +133 in 526 min
Jrue-Brand: +130 in 641 min
Iggy-Meeks: +127 in 577 min


See a trend... Jrue is a difference maker.

The funny thing is that most of those 2-men groups are Jrue and another starter. And yet the starting 5 is barely a net positive and is not among the team's best 5-man units. Yet when Jrue plays with another stater in a mixed line-up (typically with Thad) then they have more success.

If you look at their most successful 3 man units the trend continues: http://www.nba.com/statistics/plusminus/plusminus_sort.jsp?pcomb=3&season=22011&split=9&team=76ers

Holiday is part of 7 of the 8 best 3-man units. Iguodala is part of 5/8. No other player part of more than 3 of the top 8.

As much as Jrue's brilliance on the defensive end is a subject of discussion, I fail to see why Thad's humongous leap in that department is not as great a factor.

Thad's help defense turns breakdowns into stops and turnovers for the other team. His mobility allows Jrue to defend the pick and roll as far as the 3pt. line. Improved defensive rebounding and long arms have poked out many a steal. Even though he cannot guard big power forwards, he still holds his own in spurts.
An all-NBA defensive team mention is not laughable.

"An all-NBA defensive team mention is not laughable."

Which is unbelievable considering where he is coming from defensively.

His jumper is also fixed.

Wow. OK, so I wanted to take a closer look at Iguodala's off-court numbers and there's something in there that's really shocking.

With Iguodala off the floor, the team has played 380.16 minutes. In those minutes, their OFR is 111.84, DFR is 101.97. The differential is 9.86.

Now if you break that number down further. For 221.83 of the minutes when Iguodala has been off the floor, Jrue has been on the floor. In those 221.83 minutes, the team's OFR is 117.8 and the DFR is 99.74. A differential of 18.1.

For the 159.13 minutes when AI9 is off the floor (and Jrue is off the floor) the team's OFR is 103.6, DFR is 104.8. A differential of -1.19.

So for 58% of the minutes when Iguodala is off the floor, Jrue is on it, and the team is playing great basketball.

Now if you do the same for Jrue's minutes, the dif when AI9 is on the floor (but Jrue is out) is 6.2. The dif when both are off the floor is the same, -1.19. And Iguodala plays about 56% of the minutes when Jrue is off the floor.

So basically, Iguodala's adjusted number is being pulled down by how great the team has been when he's been on the bench while Jrue was still on the floor.

Are those numbers up to date or based on 82games.com?

Anyhow, that seems to be why Jrue's adjusted +/- is so much better than Iggy's. Also it is a bit of a credit to Turner. Because I think ET has done well replacing Iggy, while no-one really replaces Jrue- particularly in terms of how well he defends PG's. When Lou replaces Jrue the team plays well offensively but the defense really suffers.

BasketballValue.com's stats, just pulled them 15 minutes ago.

Also, don't let Collins see those numbers you came up with. It makes a strong argument that the team could be successful even if they traded away Iggy. A strong argument that is unlikely to hold up in the real world :)

I think Iggy is a very good player, but I do honestly believe that if he were traded for someone competent, and Turner got more minutes to replace Iggy, the team wouldn't fall much.

I wonder if Iggy still wants to be traded (going on the presumption that skipping his exit interview followed by Thorn calling around to make offers meant that he wanted a trade last summer).

It's fine for you to believe that, but there's absolutely no evidence to support it. Turner hasn't performed to a level higher than his current role on the team. That's just a fact.

Overall the team is more effective when Turner is on the bench. But that does not diminish how improved his game has been this year.

ET's efficiency numbers have gone from league worse to league average. While his raw +/- has gone from -50 to +95. So he very well might be on track to earn a bigger role, but I agree he has a ways to go. When you have a scorer/play-maker like ET still only putting up average efficiency it does not exactly scream "unleash him."

So he will have to stay in the Doug-House a bit longer. I'm hoping that by next year he will have a role similar to Harden and then by the time Iggy's contract is up they can seriously ponder where to go next.

Harden was efficient from the second he stepped on an NBA floor, though.

user-pic
bebopdeluxe reply to Brian on Feb 15 at 0:43
+/-

I call BS. As spiller said in another forum, saying that Turner doesn't deserve more minutes than he is getting now conviently avoids the fact that his minutes were cut BEFORE the dropoff in his game...not the other way around.

This all started with the Denver game, IMO. In that game, Turner played the entire 4th quarter and OT...while Lou only played 19 minutes - his lowest total of the season. Up until then, we were seeing a LOT of JTI ar rhe end of games, and it was working pretty f'n well I thought. After that game, it seemed like Collins went to LouWillVille and put the Boss back in charge in the 4th quarter, and one towel-kicking incident (after an out-of-conrol rant by Collins into Turner's grill for EVERYBODY to see) and one loose ball that Reggie Evans got to, and Turner is now the back-up SF and ball- handler for Lou and Meeks.

Yay.

Can you tell me what Turner was doing wrong during the first month of the season that earned him the material cut in minutes hat we have seen over the last 2-3 weeks?

I think if you wait another 2 weeks Et will be back to 25 minutes a game. His minutes are dramatically down the last 5 games- but I'm expecting that to be a blip.

Overall ET, Meeks, Lou, Thad, brand and Hawes all average within 4 minutes of each other. Only Iggy and Jrue get dramatically more minutes. You can make a much stronger argument for Brand or Thad deserving more minutes than you can for the other guys in that group.

So IMO the only issue with ET and minutes has been the last 5 games. And I guess for me 5 games is too small a period of time to get worked up. If ET is still only playing 15 minutes a game over the next 20+ games then I'll join you in being concerned.

Brand and Thad I agree with. Don't forget that ET has gotten a lot of garbage time minutes that skews those averages.

user-pic
bebopdeluxe reply to spiller27 on Feb 15 at 1:06
+/-

QFT. Relative to guys like Thad and Brand, how often were they out there in the last 10 miuntes of the game while Turner was out there with the deep reserves? i also wonder to what respect Turner's being out there with the scrubs has affected both his offensive and defensive efficiency ratings relative to some of the other guys...and even if Meeks plays some of that scrub time with Turner, he also plays a LOT of 3rd quarter minutes that were truly ri-donk-u-lous in therms of overall efficiency.

But in my original post the point was that neither Meeks, Lou or ET had net positive impact as compared to when they are on the bench. So basically none of them are showing themselves worthy of increased minutes compared to the others... which jibes almost exactly with their actual minute distribution.

I'm not saying any of them are hurting the team or are not valuable. Just that none have separated themselves the way that Jrue, Thad and Brand have (along with Hawes when he played this year.) That is what the secod chart is saying. Those stats might not be correctly portraying the reality on the court- but that is why I posted it for discussion.

That's a great point. Garbage time has unquestionably hurt his defensive efficiency because we've had so much garbage time this year (due to many blowouts) and they simply don't play defense during garbage time.

4 of our 6 best lineups on a +/- basis include Turner (including the #1 lineup). He's more important to the team that most here believe. It also helps why the team has not looked as good since his minutes started dropping.

ET is absolutely important and he gets roughly the same minutes as 6 of their other key guys in their 8 man rotation.

But the team does just as well when ET is on the bench (actually it does a bit better.) Same issue for Meeks and Lou. So it is hard to say any of those guys have proven they deserve a big jump in minutes.

They do have very redundant skill sets, and in many ways Turner's scoring touch/instincts counters out Iggy's aestheticism and defense.

But like I've said before, at this point I think ET does so much better when he is guarded by smaller players. Plugging him in as a full time SF probably does him a disservice- although I really hope that he grows to where long defenders give him less problems.

Although I have no data to back up this impression.

user-pic
bebopdeluxe reply to tk76 on Feb 15 at 8:02
+/-

My response to this is MORE JTI.

What was that group doing wrong the first 15-20 games of the season to have it subsequently roll off of the face of the earth?

I put in a bit more data suggesting how impactful Jrue has been to this team as compared to his other teammates.

Best 5-man unit for the Sixers:

http://www.nba.com/statistics/plusminus/plusminus_sort.jsp?pcomb=5&season=22011&split=9&team=76ers

Holiday and the night shift (Lou, Turner, Young and Vucevic). They are +36 in 56 minutes they have played together.

OFR: 125.47
DFR: 92.38
Adj +/-: +33.18 (there is actually another great lineup but the sample size is too small)

http://basketballvalue.com/teamunits.php?year=2011-2012&sortnumber=17&sortorder=DESC&team=PHI

And this right there pretty much explains the Sixers success this season and some of our observations over the last 2 months:

- Jrue has the biggest impact on the improvement
- The bench builds up the leads
- Lou is a positive when he plays off of Jrue instead of ball hogging
- Vucevic plays better with the bench (Collins did hint a month ago he plays well with Young)

user-pic
tk76 reply to Xsago on Feb 15 at 5:25
+/-

A rare post in that it will probably make everyone happy...

This is a link that may be relevant,regarding RAPM (regularized adjusted plus-minus)
http://stats-for-the-nba.appspot.com/ranking_rec
Holiday, Young, and Iguodala all rank fairly close by this metric, however the big difference is that AI is very effective on defense, while hurting the offense, while the others are opposite. In fact, AI is one of the few wing players who ranks up there with Garnett, Howard, M. Gasol, Bogut et. al. in defensive RAPM.

If you click on "info" next to the player names, a useful feature comes up: it shows how the player compares with others who have played with the same 4-man units. For AI, the only players who have played significant minutes with the same units are Holiday and Turner, and in the cases of both he is, again, much better on defense, while hurting the offense (green shading is good, red is bad). The larger sample size of last season tells a similar story.

Interesting site that i have never seen before. Thanks for the link. There are some nice stuff in there.

user-pic
tk76 reply to Lance on Feb 15 at 6:47
+/-

Not sure I totally follow what the numbers mean but definitely interesting. A few observations:

Jrue made a huge leap from negative to positive defense.

Jrue one of the most balanced offensive and defensive players on the list. Most others are strongly positive on only one category.

Iguodala the only top rated player with a negative offensive value. I guess his assists were not enough by this metric.

Linsaniy

user-pic
tk76 reply to tk76 on Feb 15 at 6:49
+/-

Also some really odd players showing up as highly ranked by this measure.

user-pic
Sharone Wright on Feb 15 at 8:34
+/-

Knicks have won 6 in a row...and they're still 6 games behind us. Eat that, Knickerbockers.

Has anyone else noticed that Williams Never, under any circumstances, passes the ball to ET?

I wonder if ET's numbers are better if he's in the game w/o L.Will being in the game, although that might be a really small sample.

ET is +75 in the 500 minutes he has played alongside Lou.

ET is +20 in his remaining 181 minutes when not paired with Lou.

More importantly:

ET is +92 in in 370 next to Jrue.
ET is only +3 in his other 310 minutes on the floor.

Jrue has a higher +/- with Lou, Meeks and Iguodala than he does with Turner.

Likewise, Iggy has a better +/- with Jrue, Meeks and Williams than he is with Turner.

Lou has a higher +/- with Jrue than Turner, but is better with Turner than with Iguodala or Meeks.

And Meeks is better with Iggy and Jrue than he is with ET. But that is still better than Meeks is to Lou.

So basically:

1. Everybody does better playing with Jrue.

2. Jrue, Meeks and Iggy all play better with each other than they do with ET.

3. Lou does not pass, so Meeks and ET both struggle paired with Lou as compared to when the are paired with others.

Iguoda

user-pic
mchezo reply to tk76 on Feb 16 at 15:54
+/-

Just curious, isn't it arguable that the 5 man rotation numbers are the only ones that really matter? you have to play five guys.

user-pic
tk76 reply to mchezo on Feb 16 at 16:45
+/-

There are probably about 60+ 5 man units involving those players. Sorry, but it would take too much time to parse out the same numbers. While on nba.com you can pull up the 3 man units easily.

user-pic
jsmoove reply to ojr107 on Feb 15 at 13:09
+/-

Minus the hyperbole, not too wise to pass the ball back out to the perimeter to a guy who doesn't exactly make perimeter shots at a good clip.

@tk: to say that Lou does not pass is not accurate. he leads the team in assists per 36 at 5.2 and assist rate at 24.8%. His previous reputation is overshadowing his on-court performance this year and last year too when it comes to distributing.

user-pic
tk76 reply to jsmoove on Feb 15 at 13:13
+/-

Point taken. Probably more accurate to say the Lou is their highest usage rate player.

Out of curiosity, does anyone check out other team blogs w/ any regularity? Are they as split on their team's players as Sixers fans seem to be?

Like after every game on OKC blogs is there a huge argument about Harden not starting over Sefolosha, or Westbrook taking too many shots instead of passing to Durant, and certain people are constantly riding Westbrook and bashing Brooks for "punishing" Harden?

I'm from Philly. That makes me way to provincial to care about other teams. And I say that with a bit of pride :)

user-pic
Lance reply to Brian on Feb 15 at 15:17
+/-

I follow a number of different team blogs, and yes, there are lots of these disagreements. OKC fans argue about Westbrook helping and hurting the team; complain about the limitations of Ibaka and Perkins; berate their coach for allowing so much isolation play in their "offense". Boston fans argue about how good Rondo is and whether their vets are bringing him down or vise versa. Many Spurs fans think Popovich is ruining the team by optimizing his rotations for offense instead of defense (Bonner and Blair are widely despised). Anyway, I think Sixer fans are rather normal, and at least never the manic depressive mess that Knicks fans tend to be.

I think most other cities are happy with any kind of success when it comes to their teams (with a few major markets as the exceptions).

I think Philly fans are just extremely opinionated and have trouble letting it go if someone makes a comment they disagree with (myself included - I am a repeat offender). No one wants to let the other person get the last word in an argument.

user-pic
MCT reply to MCT on Feb 15 at 13:55
+/-

meant to reply to Brian's last comment


Expand/Contract all comments

Leave a comment


back-to-story.gif