Reply With Quote
good read.
These Player Pairs statistics take a look at how the Spurs perform when various possible tandems are on the court at the same time. The numbers in the cells correspond to the tandem comprised of the name in the column and the name in the row. For example, as you can see below, the Spurs outscore their opponents by 12.07 points per 100 possessions when Duncan is on the court with Parker.
The cells are color coded based on the following rules:
Light Green - The player in the column improved the impact of the player in the row.
Dark Green - The player in the column improved the impact of the player in the row by a significant* amount.
Light Red - The player in the column worsened the impact of the player in the row.
Dark Red - The player in the column worsened the impact of the player in the row by a significant* amount.
Light Blue - The player's impact when he's on the court under any cir stance.
Dark Blue - The tandem hasn't yet ac ulated enough minutes together.
* - 25% in the first chart, 2.5% in the two subsequent charts
-Box score statistics tell us Manu Ginobili is having his worst season since his rookie year. Our eyes tell us that he’s slowing down. However, despite any perceived flaws, Ginobili continues to be an amazingly positive impact on the team. It’s mindboggling, really. Not only do the Spurs score more points when he’s on the court (111.54) than anyone else, Ginobili also boasts the best defensive number (97.66). Perhaps even more impressive (if that’s even possible) is that he significantly improves everyone on the team outside of Neal. Overall, these are hugely encouraging numbers because there’s no doubt that Ginobili has room to improve individually. This also goes to show how much of a travesty it would be if the Argentine Hall of Famer decides to retire at the end of the season. Ginobili shows no signs of slowing down when it comes to his uncanny ability to make his teammates better … and isn't that what he's all about?
-Tim Duncan’s numbers look very good across the board. His capability of carrying the offense when he’s in the game is apparent. Defensively, he’s solid and the impending return of Leonard and Jackson will help, as will playing more minutes with Splitter.
-Tony Parker and Duncan’s numbers are similar, with Parker being just slightly better. Offensively, the only players he brings down so far are two players (Leonard and De Colo) who he hasn’t played with that much to date. On D, the only worrisome pairing is Parker with Neal.
-Speaking of Gary Neal, his numbers adamantly state that he’s not someone who should become a full-time starter. It’s clear that he drags down Duncan, Parker, Ginobili and Green, which tells me that he definitely needs to return to the bench ASAP. There are problems on both offense and defense. Defensively, we know he’s bad guarding point guard but these numbers also suggest that he’s also bad at defending starting shooting guards since he’s blatantly bad next to Parker and Duncan. Offensively, the biggest worry is that he isn’t meshing well with Ginobili. (When watching games, I’ve subjectively noticed that as well.) If Neal can’t start and also inhibits Ginobili’s offensive mastery, how exactly can he best be utilized going forward? I’m not sure.
-If Neal isn’t the answer at backup point guard due to not being compatible with Ginobili, that opens the door for Patrick Mills and Nando De Colo. Looking at these numbers, Mills looks like the anti-Neal because he does REALLY well next to Ginobili. Defensively, Mills is an unmitigated disaster … unless he’s next to Ginobili, that is. Offensively, Mills and Ginobili have produced an excitingly high number (119.43). De Colo is very interesting because he’s already a healthy difference maker. On offense, the only guard with a better number is Ginobili -- and it’s notable that he shines even better when not paired with another point guard. Defensively, De Colo is subpar but not nearly as bad as Mills.
-Danny Green is a mixed bag. On offense, it’s obvious that he’s best when paired with good passers. He shines with Parker, Ginobili and De Colo. But if you put him with non-passers such as Mills and Bonner, it can get ugly. Defensively, Green has massively underwhelmed thus far according to these stats. There aren’t any huge individual concerns but it’s unacceptable that he drags down pretty much everyone who matters. Green needs to get back to becoming an asset on defense or else Pop could be tempted to juggle his starting lineup.
-There’s probably not much we can take from Leonard or Jackson due to their early injuries. The offense was sputtering early in the season so that’s mostly why they have a lot of red in the offensive chart. However, we’ll have to keep an eye on whether the offense is able to keep humming once the small forwards return. On defense, they help out Duncan and Ginobili, which makes sense because they relieve Duncan from having to crash the boards as hard and keep Ginobili away from the bigger small forward in the league.
-Tiago Splitter’s numbers suggest quite a few conclusions: 1) By looking at what he did next to Leonard and Jackson, we can safely say Splitter wasn’t good early in the season but has been great lately 2) The Splitter and Diaw pairing off the bench was an utter failure 3) It looks like he should fit right into the starting lineup without a problem 4) Duncan and Splitter have learned out to play with each other on both ends of the court.
-Boris Diaw, based on these numbers at least, doesn’t look like a good candidate to come off the bench. The problem is that he also doesn’t look like an especially good candidate to start either. His defense appears to have started off well (based on his pairing with Leonard) but has fallen off a cliff. He’s strikingly bad in some of those tandems. Offensively, he doesn’t really hurt the Big Three … but also doesn’t especially help either. Ugh.
-The point differential number for DeJuan Blair is actually pretty darn good. The problem is that he’s very reliant on Ginobili in order to be halfway decent. It’s clear that he’s been a flop as a starter. Though his defense isn’t bad as a starter, he kills the team’s offense. To carve out a role for Blair, it’s going to have to be as a bench bigman who plays only if Ginobili is also on the court.
-Matt Bonner is a king, per usual. In the regular season, there’s no denying that he’s a fantastic weapon that can help your team win a lot of games. There’s simply no denying that. However, since there’s also no denying that he’s a lock to crumble once the calendar arrives at April, I agree with Pop’s coaching decision to limit Bonner’s minutes so far this season. In my opinion, the best way to use Bonner is as a shot of adrenaline whenever things begin to stagnate. Otherwise, the Spurs need to spend time figuring out how to win without Bonner making life easier for everyone.
-Overall, I'm happy with what I see. The beasting by TD and TP is helping the team. Ginobili is Ginobili when it comes to making people better. There's nothing to suggest Splitter won't work in the starting lineup. Leonard was doing well before he was sidelined. The role players all have some weaknesses but no one appears to be an absolute lost cause at this point. By constantly juggling lineups, Pop is starting to build up some compelling data that should come in handy as the season progresses. Good times
good read.
I've been saying that Mills + Ginobili = Best combination off the bench and now there are stats to prove this!
Btw, timvp, where'd you get these stats from?
Last edited by Boomersgold; 12-16-2012 at 11:39 PM.
Data supports pop starting splitter and ginobili
like he said in the WCF
you have to play your best players
just needs to do some early subs so they can get some run with the bench also and everything is fine
he needs to increase the minutes of his players
guys around the league are getting 40 minutes
stephen jackson was playing 36-40 minutes about 2 years ago
leonard/green/splitter should all be at 30 minimum with leonard probably getting his chance at leading the bench for some stretches
Thoughtful, very well written piece. thanks.
How do these pairs compare when being compared to the team average? i.e Gibobili might make someone better but if they are still below the team average it's still not a good pairing. Also, Green might make someone worse, but again if they are still above average then it is acceptable.
The team averages are:
Point differential per 100 possessions: 7.92
Points scored per 100 possessions: 109.26
Points allowed per 100 possessions: 101.34
I think that's the move that makes the most sense. There's no point in Pop waiting till a crucial playoff game and then try juggling lineups like his horrible rotations in game 5. Diaw also had arguably the best season of his career when he replaced Stoudemire at center in '06. He could provide a similar role of the bench as a backup center.
timvp filling the Hollinger void. well done!
Biggest surprise to me is Neal. Eyeball test says he's been pretty useful including improved ability to score (more in the paint, floaters...) with improved defense.
I also see no reason (outside of Boris not wanting to do well) that Boris could not be a capable big man off the bench. The main argument I can see is that when he starts he fits well next to Tim but also has Tim/TP out there to carry the offense. If you put him on the bench players, it's guys who are more limited offensively and for a guy like Boris that doesn't want to step up his offense, that could lead to some depressed numbers with Boris off the bench.
Good read.
I think your point about Bonner is well-made. He is still useful, but the team needs to wean themselves off him, so they don't need to play without him during his annual playoff collapse.
You mention Blair has to play with Manu. What do you think of the Blair with Jackson results? Is this just a small sample, or is there any meat here?
Blair can only play well when he plays with a player who gets trapped on P&Rs leaving him open and the guy has to throw a pass between someones legs to get the ball to him
so obviously gary neal is the best fit for blair
Nice.
These are great stats.
Saying that, I pretty much disagree with the whole analysis.
A vital aspect that is totally forgiven is the quality of the opponents when the player is on the court. Being able to outscore a crappy bench unit isn't at all like being able to outscore a starting unit.
The whole "Ginobili is awesome and shouldn't retire" is maybe to most way off part of the analysis. Ginobili is playing a lot of his minutes against bench players with a team that is deep. It tremendously inflate his on/off stats (it's the Bonner effect). Players like Parker and Duncan are logically doing better with Manu than without him because Manu is still an upgrade over Green and Neal. Truth is Manu is having by far his worst season since his rookie year. No advance stats can change that.
Bruno you make a really good point, but by the same thinking if Manu is still better than Green or Neal (who aren't bad players), then he can still be a major benefit to this team, even if he's not what he was.Nice.
These are great stats.
Saying that, I pretty much disagree with the whole analysis.
A vital aspect that is totally forgiven is the quality of the opponents when the player is on the court. Being able to outscore a crappy bench unit isn't at all like being able to outscore a starting unit.
The whole "Ginobili is awesome and shouldn't retire" is maybe to most way off part of the analysis. Ginobili is playing a lot of his minutes against bench players with a team that is deep. It tremendously inflate his on/off stats (it's the Bonner effect). Players like Parker and Duncan are logically doing better with Manu than without him because Manu is still an upgrade over Green and Neal. Truth is Manu is having by far his worst season since his rookie year. No advance stats can change that.
Agree, Manu is still a good player, significantly better than Green and Neal. My point was that he is nowhere near as good as these stats show.
It's the same thing than last year, Manu regular season awesome PER was fool's gold because he putted this stats mostly against bench players.
It makes sense that Ginobili's numbers are so high. He plays very limited minutes, almost all of which are spent against the opposing team's bench. That's a recipe for inflation.
Edit: Bruno beat me to it.
Kawhi is going to have to pick up his outside shooting if Tiago is going to be a permanent starter. But that's a pretty damn good defensive unit right there.
Not trying to get into argument about specific analysis in this post. However, I think advanced stats are only useful for players with similar "traditional" stats, like points, rbs, assists, FG%, etc. And for role players, it can only be comparisons among similar types, like defense specialist, 3-pointer specialist, rebounding blue collar, etc. To see something more subtle.
For example, assuming other stats are similar, it's hard to believe a 18 ppg/45% player is better than a 25ppp/44% player no matter what advanced stats might say. On the other hand, a 20ppg/48% player should be strongly presumed better than a 22ppg/42% player assuming others being equal. It has to be something much more tangible than some advanced stats to say otherwise.
In short, traditional stats lay out the foundation for evaluating players, advanced stats give us some deeper analysis on top of that foundation.
Some people are misreading the stats... if Gino is doing great with Tim/Tony/Blair, then that's when both of them are on the floor... at that point, if the opposition are starters or bench, the alleged 'advantage' applies equally to both.
Furthermore, with the exception of Mills, his production with other bench players is actually fairly pedestrian. If he would be 'propping his numbers' by playing against inferior compe ion, then you would see all other bench players gaining equally from such event, but that's not reflected there.
Now you could argue that the sample size is smaller when he plays with the team starters vs the team bench, and that's a valid concern.
Good point. But a vital aspect that is totally forgiven is the quality of the teammates when the player is on the court. Being able to outscore the opposite team playing along with elite players isn't at all like being able to outscore the opposite team playing along with crappy players.Nice.
These are great stats.
Saying that, I pretty much disagree with the whole analysis.
A vital aspect that is totally forgiven is the quality of the opponents when the player is on the court. Being able to outscore a crappy bench unit isn't at all like being able to outscore a starting unit.
Last edited by Brunodf; 12-17-2012 at 11:32 AM.
San Antonio leads the NBA with 25.3 assists per game, its highest total since averaging 24.9 assists in 1995.
http://sports.yahoo.com/news/parker-...4473--nba.html
Fwiw
I agree that the fact that Manu comes off the bench and plays against bench players should be part of the analysis. It's definitely a factor that should be something that is automatically considered. Thus, while I think Ginobili's impact has been a lot more positive than what the naked eye can see, I wouldn't go as far as to say these stats prove that Manu is a better player than Duncan or Parker this season. Making that logical jump is ignoring multiple factors, with the largest of them being Manu's compe ion level.
That said, Manu's numbers are undeniably really, really good. No matter how you frame 'em, they sparkle. Again, it doesn't make him the best player on the team or anything like that, but he's fulfilling his duties very well.
Plus, there are other advanced stats that factor in the compe ion and who Ginobili is playing with -- most notably APM and RAPM -- and they too agree that he is having a very good season.
Let's pretend that Manu's numbers are 100% due to him coming off the bench. Even given that untruth, wouldn't you agree that he's performing his role so well that it would be a shame if he retired? I mean, even with the most condescending eye, you at least have to admit he's doing what his role requires him to do. And last I checked, players who are doing what they are supposed to do at a high level aren't players you hope will hang up the sneakers.
That "Bonner effect" you insinuate doesn't exist. Bonner's plus/minus, on-off and APM stats were all fantastic in 2009 when he started 67 games. The only "Bonner effect" is that he's damn useful in the regular season .... period. Whether he starts, comes off the bench or is only used in spot situations, his advanced stats all point to that conclusion.
Last year was the first time EVER that Manu's advanced stats sucked in the playoffs. It's way too premature to start using the Bonner rules with him. Maybe Manu is only a regular season wonder at this point in his career. Maybe the bench collapsed around Manu so dramatically that even he couldn't stop the destruction. While the former is possible, my money is on the latter. If healthy and as long as the rest of the bench doesn't suck horribly like they did last year in the playoffs (outside of possibly Jack), I highly doubt we see a repeat of the suckage.
Ginobili in as recently as 2011 had the best on-off numbers on the team in both the regular season and the playoffs. To say he's only a regular season specialist now is unfair ... or at the very least speculation.
For reference, 260 out of the 390 minutes Ginobili has played this season have been next to Parker. How often does Pop play both Parker and Ginobili at the same time versus the other team's bench? Not that often, tbh.
I think it's beyond fair to say that at least 50% of Ginobili's minutes comes against quality compe ion. And it's probably closer to 60%, IMO.
I was looking for that, but couldn't find it. Thanks.
Also looking at gameflows, Manu actually closed 18 out of 22 games with Tim and Tony (taking out the first two of the season and Boston, where he wasn't available).
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
