PDA

View Full Version : Player Pairs Analysis - 2015-16 January Edition



rasuo214
02-02-2016, 04:12 AM
December edition: http://www.spurstalk.com/forums/showthread.php?t=255855
November edition: http://www.spurstalk.com/forums/showthread.php?t=255193

Description for those that don't already know what numbers are all about:

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 10.7 points per 100 possessions when Leonard is on the court with Green.
Players are listed in the same order as last version because it's easier that way.

Green: The player in the column improved the impact of the player in the row.
Red: The player in the column worsened the impact of the player in the row.
Blue: The player in the column had neither a positive or negative impact on the player in the row.

YTD:

http://i.imgur.com/zT59zyQ.jpg

http://i.imgur.com/olXgdu1.jpg

http://i.imgur.com/kzyHYl7.jpg

http://i.imgur.com/9O2z72B.jpg

If you're looking for the difference between Point Differential and Net Pairings, here's the explanation from November (obviously the numbers will be different for the current month):


Difference between Point differential and Net Pairing. In the first Chart we seen Duncan improves everyone, so why does he have some negative net pairings? Basically the difference is in Point differential if the points per 100 possession is higher than his paired teammate then Duncan is having a positive impact (examples: Leonard is +11.3 with Duncan Leonard is +13.9, Parker is +9.9 with Duncan Parker is +12.7 etc.)

In Net Pairing it looks at the impact that both players have on one another, so while Duncan's teammates play better with him on the floor, Duncan doesn't necessarily play better in certain pairings (Leonard, Green, Parker, Aldridge, Butler, McCallum). So Net Pairing looks at whether the pairing is a net positive or negative.

Example: Duncan and Leonard, Duncan has a positive impact on Leonard but Leonard has a negative impact on Duncan. Combined they are still a net positive by +1.2 points per 100 possessions, individually Duncan is +15.3 and Leonard is +11.3, so combined you would expect at least +26.6 (15.3+11.3). Duncan with Leonard is a +13.9 (so while this is lower than Duncan's +15.3) combined Duncan and Leonard are a +27.8 (13.9+13.9) which is 1.2 points more than their combined individual base (27.8-26.6).

Best Pairs:
Kawhi, Green, Duncan with the bench - 3 months in and the Starters aren't doing much better playing with one another but they're great when paired up the bench, particularly Duncan, Green and Kawhi who excelled with the Spurs bench squad in January.


Pop, Please stop playing these guys together:
Mills with another PG - It just doesn't work out very well when Mills is with either Parker or McCallum, thankfully we didn't see the Mills and McCallum pairing in January.
Boban not so great - Last month, Boban was on top of the world dominating other benches, January on the other hand wasn't as favorable. He only had 2 favorable pairings (Leonard and Aldridge)

Offensive Stars: Leonard, Green and West
Defensive Star: Tim Duncan
MVP: Duncan

rasuo214
02-02-2016, 04:14 AM
January:

http://i.imgur.com/qIkCzDy.jpg

http://i.imgur.com/KPg5gdg.jpg

http://i.imgur.com/wxUMorr.jpg

http://i.imgur.com/sX9Eah5.jpg


NOTE: I missed the Parker/Duncan Net Pairing cell (should be green), I'll fix it when I get the chance.

313
02-02-2016, 04:23 AM
:worthy: Duncan, the head of the snake on both ends

:tu Good stuff man, info like this is appreciated more than you know

inb4 player fans

313
02-02-2016, 04:26 AM
Best Pairs:
Kawhi, Green, Duncan with the bench - 3 months in and the Starters aren't doing much better playing with one another but they're great when paired up the bench, particularly Duncan, Green and Kawhi who excelled with the Spurs bench squad in January.

IMO the bench has always been the offensive punch for the Spurs the last 4 years, while the SL were more defensive focused, scoring off TOs, etc but the offense for the SL has always been more stagnant.

SAGirl
02-02-2016, 04:43 AM
Just by a glance it seems like there should be some mix and match. The SL is a negative with each other, and plays better with some bench players than with each other, so it is no wonder we always start slow. It's possible Pop is letting them like this to let them establish chemistry and take off, but at this point of the season, it seems like some tweaking might be necessary.

Not sure what kind of adjustments Pop can make, but there are potential hidden x factor combinations. There are some games you don't want to start off from behind. Like Diaw starting the game in Miami in 2014 bc Pop didn't want the team to have a slow start. The offense went through Diaw as a quarterback.

I suspect we may see some "no PG lineups" strategically at times for a few minutes in the playoffs for defensive reasons, and there is likely to be mixing and matching a whole lot. Also, we have to hope Anderson and Simmons continue to develop.

rasuo214
02-02-2016, 04:51 AM
IMO the bench has always been the offensive punch for the Spurs the last 4 years, while the SL were more defensive focused, scoring off TOs, etc but the offense for the SL has always been more stagnant.

Last season wasn't nearly as bad: http://www.spurstalk.com/forums/showthread.php?t=246848


This year's bench is much better, plus Danny's slow start and Aldridge's implementation are the likely factors.

SAGirl
02-02-2016, 04:58 AM
Last season wasn't nearly as bad: http://www.spurstalk.com/forums/showthread.php?t=246848


This year's bench is much better, plus Danny's slow start and Aldridge's implementation are the likely factors.
And that was counting with a very, very, just awful year for Tony.

I think Anderson would help the starters, but that would require moving Kawhi to SG (although ppl want to see him handle the ball more, so maybe it is a good idea). Right now Kyle is at the 4 because of necessity, so we are not going to see that match. Maybe J.Simms is the answer, but I don't think so. He really has a niche, and that is transition scoring. Lets just say he's a developing player in the half court. But it could be either of these guys. There is just no one else you can spare, unless you do away with Tony entirely and send him to the bench. Mills has shown he's just too inadequate defensively, although he hustles more.

It is hard to say at this point. Pop is still in experiment mode so we will see more combinations.

loveforthegame
02-02-2016, 11:18 AM
Always a good read. Thanks again for putting this together. :toast

Kawhitstorm
02-02-2016, 04:04 PM
-Kawhi/Simmons are a much better pairing than Kawhi/Danny.:toast
-LMA/Manu pairing has also been trending the entire season.:toast

Best line-up,

Tim
LMA
Kawhi
Simmons
Manu

6th man: Patty
Small-ball: Fathead over LMA


Defensively,

-Diaw/LMA >>>> West/LMA (Besides the LMA/Tim pairing, Pop should just roll w/ Tim/West & Diaw/LMA)

look_at_g_shred
02-02-2016, 09:18 PM
It would be nice to see what kind of result we get from Simmons starting instead of green, but I don't think pop does it. His mantra is that you have to "earn" it. I think If any experimenting will be done it will be done while Duncan is out. Pop is stubborn and I have a gut feeling the starting 5 of TD/LMA/KL/DG/TP is what he will roll with to the end of the season and into the playoffs.

GSH
02-02-2016, 09:45 PM
The problem with player pairings is that it doesn't take into account WHY a certain pair may be on the floor at the same time. For instance, why would Kawhi be on the floor with Boban? For the most part, Boban comes in when the game is under control. And usually (maybe even always) when the opponent has mostly bench players on the floor. Comparing that pairing to two other players who are usually on the floor against the best the opponent has to offer is probably not going to reveal a whole lot.

The other thing is, it doesn't tell you how many minutes that pairing has together. You can have some real outliers in small samples.

I love looking at the numbers, and thanks for posting. Just be cautious about drawing too many conclusions, without context.

Kawhitstorm
02-02-2016, 10:55 PM
The other thing is, it doesn't tell you how many minutes that pairing has together. You can have some real outliers in small samples.


There is an app for that: http://www.basketball-reference.com/teams/SAS/2016/lineups/
(http://www.basketball-reference.com/teams/SAS/2016/lineups/)




FYI: Porker/Danny don't appear in any of the top 4 four-man lineups:wakeup


Net (Per 100 Poss)


Rk
Lineup
MP
FG
FGA
FG%
3P
3PA
3P%
eFG%
FT
FTA
FT%
PTS ▾
ORB
ORB%
DRB
DRB%
TRB
TRB%
AST
STL
BLK
TOV
PF


1
L. Aldridge (http://www.basketball-reference.com/players/a/aldrila01.html) | T. Duncan (http://www.basketball-reference.com/players/d/duncati01.html) | M. Ginobili (http://www.basketball-reference.com/players/g/ginobma01.html) | K. Leonard (http://www.basketball-reference.com/players/l/leonaka01.html)
108:05
+13.9
+3.8
+.142
-2.1
-7.3
+.019
+.127
+1.1
-6.5
+.250
+26.7
+2.5
+11.6
+13.8
+11.6
+7.7
+18.4
+8.4
-2.9
+3.4
-0.6
+1.2


2
B. Diaw (http://www.basketball-reference.com/players/d/diawbo01.html) | M. Ginobili (http://www.basketball-reference.com/players/g/ginobma01.html) | K. Leonard (http://www.basketball-reference.com/players/l/leonaka01.html) | P. Mills (http://www.basketball-reference.com/players/m/millspa02.html)
180:43
+8.8
+3.4
+.085
+5.0
+4.6
+.163
+.113
+2.2
+0.7
+.091
+25.0
-1.7
-1.9
+3.3
-1.9
+1.1
+2.7
+8.4
+1.6
+0.8
-4.7
-2.0


3
L. Aldridge (http://www.basketball-reference.com/players/a/aldrila01.html) | M. Ginobili (http://www.basketball-reference.com/players/g/ginobma01.html) | K. Leonard (http://www.basketball-reference.com/players/l/leonaka01.html) | P. Mills (http://www.basketball-reference.com/players/m/millspa02.html)
140:03
+8.5
+0.2
+.095
+7.4
+5.4
+.239
+.137
+0.5
-2.0
+.117
+24.9
-1.4
+0.8
+6.6
+0.8
+2.7
+6.2
+9.4
-1.6
+1.1
-1.7
-1.6


4
L. Aldridge (http://www.basketball-reference.com/players/a/aldrila01.html) | B. Diaw (http://www.basketball-reference.com/players/d/diawbo01.html) | K. Leonard (http://www.basketball-reference.com/players/l/leonaka01.html) | P. Mills (http://www.basketball-reference.com/players/m/millspa02.html)
141:19
+5.2
+1.8
+.049
+5.4
+5.1
+.153
+.078
+6.2
+4.0
+.163
+22.0
-1.0
-1.2
+1.3
-1.2
+0.2
+0.4
+5.5
+3.2
+2.2
-6.4
-2.3

GSH
02-03-2016, 09:33 AM
There is an app for that: [URL="http://www.basketball-reference.com/teams/SAS/2016/lineups/"]http://www.basketball-reference.com/teams/SAS/2016/lineups/

FYI: Porker/Danny don't appear in any of the top 4 four-man lineups:wakeup




Yeah, see, I'm talking about match-ups. Picture Diaw/West on the floor, holding their own. The other team could make one substitution, and Pop would have one of those guys off the floor almost immediately. So the pairing numbers are only valid against certain types of lineups. Duncan/Aldridge would fare a lot better against small lineups than Diaw/West would against really big lineups.

The minutes played are mostly important as a screen against small samples. The "best" 5-man lineups this season are Duncan/Aldridge/Anderson/Green/Parker (+47) and Duncan/Aldridge/Leonard/Parker/Simmons (+51). That either proves that Anderson is better than Kawhi, or that Simmons should be starting for Green... right? Most people here would scream against the first idea (of course), but demand that the second must be true. The fact is, those two 5-man combos have a total of 22 and 18 minutes, respectively.

Most people want to use player pairings in a vacuum. They can do it. But the conclusions will be shit.

YGWHI
02-03-2016, 12:46 PM
Thanks rasuo! Great job as usual. :toast


The minutes played are mostly important as a screen against small samples. The "best" 5-man lineups this season are Duncan/Aldridge/Anderson/Green/Parker (+47) and Duncan/Aldridge/Leonard/Parker/Simmons (+51). That either proves that Anderson is better than Kawhi, or that Simmons should be starting for Green... right? Most people here would scream against the first idea (of course), but demand that the second must be true. The fact is, those two 5-man combos have a total of 22 and 18 minutes, respectively.

One of the best 5-man lineups last season was Patty, Manu, Kawhi, Boris, Tim, the Spurs played that lineup a ton of minutes. Sure, Parker/Tiago injuries helped and while they didn't start Patty over CoJo, they still played good minutes together.

I wouldn't start Simms over Green yet but it won't be crazy to play Kawhi more minutes with the bench. He could help a lot on defense in those minutes, also, the need of Kawhi-Simms pair would be a solved problem.

Kawhitstorm
02-03-2016, 04:25 PM
The minutes played are mostly important as a screen against small samples. The "best" 5-man lineups this season are Duncan/Aldridge/Anderson/Green/Parker (+47) and Duncan/Aldridge/Leonard/Parker/Simmons (+51). That either proves that Anderson is better than Kawhi, or that Simmons should be starting for Green... right? Most people here would scream against the first idea (of course), but demand that the second must be true. The fact is, those two 5-man combos have a total of 22 and 18 minutes, respectively.

Most people want to use player pairings in a vacuum. They can do it. But the conclusions will be shit.

The two 5 man line-ups you listed haven't even played a HALF games worth of minutes.:lmao The four man line-ups I listed have played over two FULL games worth of minutes each & the BEST lineup has Tim in it who DOESN'T play in garbage time nor against most garbage teams.

GSH
02-03-2016, 05:00 PM
The two 5 man line-ups you listed haven't even played a HALF games worth of minutes.:lmao The four man line-ups I listed have played over two FULL games worth of minutes each & the BEST lineup has Tim in it who DOESN'T play in garbage time nor against most garbage teams.


That was my point, numb-nuts. Holy shit, some of you just refuse to read. You said the EXACT same thing I was saying, and then laughed at me for saying it?

Fucking wanker.

GSH
02-03-2016, 05:12 PM
Here....

The Green/Parker Player Pair, as written by Kawhitstorm:
Or... Hop on Pop, Spurs Version

Tony Parker, Danny Green
The worst player pair I've ever seen.

Should they sit against the Dubs?
Would you play them against scrubs?

Yes they should sit against the Dubs.
I would not play them against scrubs.
I do not like that player pair,
I do not like them anywhere.

Do you think that Pop will flinch?
Will he put them on the bench?

I do not thing that Pop will flinch.
He will not put them on the bench.
I don't know why he is so dumb.
As coaches go, he is a bum.

SAGirl
02-03-2016, 05:46 PM
Actually Tony and Danny mesh well in their games. Probably a lot of the - in the SL is due to the Danny's very cold couple of months to start the season + shitty play. LMA also didn't start out too well.

It is a concern to me though that the SL continues to have cold starts often, and maybe there is an x-factor guy out there that you can switch that will improve that unit offensively. But I also mentioned it is too tough to make that call if Pop hasn't seen them together at all through the season, and maybe that guy is Simmons, but I am not 100% on that mostly bc the SG in the SL has a very minor role in spotting up and defense, which is ideal for Danny. They are better with Manu bc Manu is overall just a much better player period. It's not even in the dynamic.

Kawhitstorm
02-03-2016, 05:52 PM
That was my point, numb-nuts. Holy shit, some of you just refuse to read. You said the EXACT same thing I was saying, and then laughed at me for saying it?

Fucking wanker.

Nope, my line-ups are fault proof:
-Tim doesn't play in garbage time
-Tim doesn't play much against garbage teams
-They have played over two games worth on minutes

:wakeup