Thru 8 games, usage up, assists up, turnovers up, efg% & ts% up, ft% up, 3pt fg% down, Moreyball up, overall boxscore metrics up.
On the court plus / minus up. On / off down. And negative for 3rd straight year, including his last Euroleague season and thus his full run in a very large lead role against top competition.
Doncic
Re: Doncic
After yesterday's game, Doncic's on / off plus minus is now -11.4 pts per 100 possessions for season.
Not strong with any of those in the biggest minute lineup. Best with bench players.
Not strong with any of those in the biggest minute lineup. Best with bench players.
Re: Doncic
Other recent Doncic comments at @bballstrategy on twitter.
Re: Doncic
Doncic, very excellent boxscore, worst on team on court plus minus and on /off.
Now negative for raw on court plus minus for season.
Now negative for raw on court plus minus for season.
Re: Doncic
It kinda looks like "Mavs starters" have bad on-off, and their MOV comes from an above-avg bench.Doncic could be listed as PG, but he's started at SG in 3 games that Brunson (100% PG) started.
Courtney Lee has played in 3 games, all starts, for a total 26 minutes.
mpg are per team game
Brunson+Doncic are -1.2 pts/100 in 104 min.
Code: Select all
. Mavs GS pos mpg On Off
Doncic 10 G 35 -1.7 12.8
Porzingis 9 C 28 -8.1 17.0
F-Smith 7 F 26 1.5 3.7
Curry 6 SG 20 -6.3 8.6
Powell 5 PF 19 7.7 -0.7
Kleber 4 PF-C 25 -1.7 6.8
Brunson 3 PG 21 14.4 -6.6
Lee 3 SF 3 -11.6 3.4
Wright 2 SG 24 8.2 -3.2
Boban 1 C 4 39.3 -1.2
Hardaway 0 SF 23 11.7 -5.9
Jackson 0 PF 13 -3.4 4.8
Barea 0 PG 2 0.0 2.6
Courtney Lee has played in 3 games, all starts, for a total 26 minutes.
mpg are per team game
Brunson+Doncic are -1.2 pts/100 in 104 min.
-
jgoldstein34
- Posts: 249
- Joined: Fri Sep 30, 2016 6:38 pm
Re: Doncic
9 of his 12 most used lineups negative. 7 from about -20 to -80 pts per 100 possessions.
Current record benefiting from some pop from dink lineups.
Current record benefiting from some pop from dink lineups.
Re: Doncic
. Mavs mpg On Off
Doncic 35 -1.7 12.8
Porzingis 28 -8.1 17.0
F-Smith 26 1.5 3.7
Kleber 25 -1.7 6.8
Wright 24 8.2 -3.2
Hardaway 23 11.7 -5.9
Brunson 21 14.4 -6.6
Curry 20 -6.3 8.6
Powell 19 7.7 -0.7
Jackson 13 -3.4 4.8
Boban 4 39.3 -1.2
Lee 3 -11.6 3.4
Barea 2 0.0 2.6
I just copied the table that Mike posted, reordered by minutes per game, and deleted the games started and position columns (and probably wrecked the formatting).
If I'm reading correctly, the top four guys are negative contributors as far as on/off is concerned, and they all play more than half the game on average. The next three guys (Wright, Hardaway, and Brunson) are positive and play a little less than half. Then Curry is negative again, Powell is positive, Jackson is negative, and the last three hardly play.
How is a team where there are more negative players than positive, and they get more minutes, playing to a decent record with a positive MOV? Is that even the right way to think about it? If it isn't, does on/off say very much about Doncic or any other given player?
Doncic 35 -1.7 12.8
Porzingis 28 -8.1 17.0
F-Smith 26 1.5 3.7
Kleber 25 -1.7 6.8
Wright 24 8.2 -3.2
Hardaway 23 11.7 -5.9
Brunson 21 14.4 -6.6
Curry 20 -6.3 8.6
Powell 19 7.7 -0.7
Jackson 13 -3.4 4.8
Boban 4 39.3 -1.2
Lee 3 -11.6 3.4
Barea 2 0.0 2.6
I just copied the table that Mike posted, reordered by minutes per game, and deleted the games started and position columns (and probably wrecked the formatting).
If I'm reading correctly, the top four guys are negative contributors as far as on/off is concerned, and they all play more than half the game on average. The next three guys (Wright, Hardaway, and Brunson) are positive and play a little less than half. Then Curry is negative again, Powell is positive, Jackson is negative, and the last three hardly play.
How is a team where there are more negative players than positive, and they get more minutes, playing to a decent record with a positive MOV? Is that even the right way to think about it? If it isn't, does on/off say very much about Doncic or any other given player?
Re: Doncic
There are different ways to look at this challenge. The one I am going to talk about now is based initially on team effectiveness experienced with pairs.
Of the 20 most used pairs, 8 are more than marginally positive for the season. 4 players have 13 appearances in those pairs. Brunson, Wright, Hardaway and Doncic. Carlisle has used 255 quads but not this one at all (unless I missed it scanning the list). It is possible that all 4 doesn't work but I'd try it before deciding that, before trying 254 other things. He has used trios of these 4. 2 biggest minutes trios without Doncic were fabulous. 2 of the 3 without Hardaway were negative with the lineup with Porzingis being about neutral. Play the quad, the one with Zingis (with and without Hardaway) and the two without Doncic more than just a few minutes per game on average and see what happens. Surely there are many dozens or even 100-200 less rational, pretty meaningless to bull quads that you can sacrifice for this more important testing cause.
Porzingis, Kleber and Powell all have one appearance in the top 8 pairs. Try out one or two of them with 3-4 of these guys. Porzingis without these other two and with Kleber. Kleber and Powell. Focus of solving the management of these 7 guys over any and everything else. For at least a month. I'd devote at least half the game to 5-7 lineups and up to 3/4th of game to these quads. Carlisle is giving over 80% of total time to 8 quads right now, it just that 7 of the 8 are awful to horrific. Time given to top 5 man lineups is light (round 25%), too light as many, probably too many variations are being tried. Around 120 lineups tried so far.
While trying these quads and 5 man lineups, I'd be looking at trade options, primarily for bigs. Bigs who are better shot defenders and / or rebounders. And better fits with these 4 core guys.
Prozingis - Doncic is not working and has faded to -9.3 per 100 possessions. Is it fixable? I dunno for sure but it doesn't look easy and there is no sign yet that Carlisle knows how to run it. I'd take my approach identified very quickly over what is being done. With more time it could be checked harder and refined.
Of the 20 most used pairs, 8 are more than marginally positive for the season. 4 players have 13 appearances in those pairs. Brunson, Wright, Hardaway and Doncic. Carlisle has used 255 quads but not this one at all (unless I missed it scanning the list). It is possible that all 4 doesn't work but I'd try it before deciding that, before trying 254 other things. He has used trios of these 4. 2 biggest minutes trios without Doncic were fabulous. 2 of the 3 without Hardaway were negative with the lineup with Porzingis being about neutral. Play the quad, the one with Zingis (with and without Hardaway) and the two without Doncic more than just a few minutes per game on average and see what happens. Surely there are many dozens or even 100-200 less rational, pretty meaningless to bull quads that you can sacrifice for this more important testing cause.
Porzingis, Kleber and Powell all have one appearance in the top 8 pairs. Try out one or two of them with 3-4 of these guys. Porzingis without these other two and with Kleber. Kleber and Powell. Focus of solving the management of these 7 guys over any and everything else. For at least a month. I'd devote at least half the game to 5-7 lineups and up to 3/4th of game to these quads. Carlisle is giving over 80% of total time to 8 quads right now, it just that 7 of the 8 are awful to horrific. Time given to top 5 man lineups is light (round 25%), too light as many, probably too many variations are being tried. Around 120 lineups tried so far.
While trying these quads and 5 man lineups, I'd be looking at trade options, primarily for bigs. Bigs who are better shot defenders and / or rebounders. And better fits with these 4 core guys.
Prozingis - Doncic is not working and has faded to -9.3 per 100 possessions. Is it fixable? I dunno for sure but it doesn't look easy and there is no sign yet that Carlisle knows how to run it. I'd take my approach identified very quickly over what is being done. With more time it could be checked harder and refined.
Re: Doncic
On/Off means almost nothing. Same 4 +/- can be useful. On/Off...not really.xkonk wrote: ↑Thu Nov 14, 2019 3:14 am . Mavs mpg On Off
Doncic 35 -1.7 12.8
Porzingis 28 -8.1 17.0
F-Smith 26 1.5 3.7
Kleber 25 -1.7 6.8
Wright 24 8.2 -3.2
Hardaway 23 11.7 -5.9
Brunson 21 14.4 -6.6
Curry 20 -6.3 8.6
Powell 19 7.7 -0.7
Jackson 13 -3.4 4.8
Boban 4 39.3 -1.2
Lee 3 -11.6 3.4
Barea 2 0.0 2.6
I just copied the table that Mike posted, reordered by minutes per game, and deleted the games started and position columns (and probably wrecked the formatting).
If I'm reading correctly, the top four guys are negative contributors as far as on/off is concerned, and they all play more than half the game on average. The next three guys (Wright, Hardaway, and Brunson) are positive and play a little less than half. Then Curry is negative again, Powell is positive, Jackson is negative, and the last three hardly play.
How is a team where there are more negative players than positive, and they get more minutes, playing to a decent record with a positive MOV? Is that even the right way to think about it? If it isn't, does on/off say very much about Doncic or any other given player?
Re: Doncic
In each case, the positive (whether on or off) is of greater magnitude than the negative.
jg wrote:
Assume this means he subs out 3 to 5 guys at once? So fresh legs go in, guys who know their role with this set of teammates and generally outperform either weary opponents or their subs?Carlisle is pretty notorious for hockey style subs
Crow wrote:
This seems to be key to the starting order being outscored. Porzingis played C only 1/5 of his time in NY, now entirely C. Also now he isn't the 1st option; and he's not making his usual shots, other than the 3. Obviously a major investment, so give him time.Prozingis - Doncic is not working and has faded to -9.3 per 100 possessions
Re: Doncic
Same 4 would be more useful with more minutes than now, especially with scattershot coaching. What same 4 I could / did do with Doncic looks real bad for Doncic and consistent with the on and on/off data analysis. I'll do more of that later but Carlisle's chaotic lineup management makes that difficult and produces sample sizes of limited meaning.
I'll give plus minus on the court overall and in pairs more weight than on / off but I won't completely dismiss on/ off, especially when it has lasted for 2 seasons before and has gotten so big & intense this season.
Right now the NBA data suggests to me that Doncic's team depends on the quality of its other players, especially in non-Doncic lineups, heavily for team results and is not being led anywhere strongly by Doncic.
We'll see what RAPM and RPM estimates, though that won't completely settle it.
Ultimately we may see playoffs. Playoffs could be even more Doncic centric. That could hurt Mavs or possibly excuse or vindicate the Doncic dominance.
If On/Off means "almost nothing" then conceivably a "great" and / or "successful" player could exist and should be shown to exist with bad or very bad raw on /off. Who is it?
When I checked about a dozen contemporary players @bballstrategy recently (mostly playmakers) I found negative on / off and extremely negative on / off to be rare and brief for these players (most of whom were merely good) and not consistent with team "success". On/Off may mean "almost nothing" to some who presumably don't expect it to last long term... but so far it has with Doncic and this year is a massive acceleration of it. If the massive on/off ends it will cease to matter for Doncic but not because it "didn't matter" but because it ceased.
I'll give plus minus on the court overall and in pairs more weight than on / off but I won't completely dismiss on/ off, especially when it has lasted for 2 seasons before and has gotten so big & intense this season.
Right now the NBA data suggests to me that Doncic's team depends on the quality of its other players, especially in non-Doncic lineups, heavily for team results and is not being led anywhere strongly by Doncic.
We'll see what RAPM and RPM estimates, though that won't completely settle it.
Ultimately we may see playoffs. Playoffs could be even more Doncic centric. That could hurt Mavs or possibly excuse or vindicate the Doncic dominance.
If On/Off means "almost nothing" then conceivably a "great" and / or "successful" player could exist and should be shown to exist with bad or very bad raw on /off. Who is it?
When I checked about a dozen contemporary players @bballstrategy recently (mostly playmakers) I found negative on / off and extremely negative on / off to be rare and brief for these players (most of whom were merely good) and not consistent with team "success". On/Off may mean "almost nothing" to some who presumably don't expect it to last long term... but so far it has with Doncic and this year is a massive acceleration of it. If the massive on/off ends it will cease to matter for Doncic but not because it "didn't matter" but because it ceased.
Re: Doncic
Durant had terrible on / off his first 2 seasons. Then it ceased. And got very good and the team success with the strong Durant on /off validated the value of good on /off over the team failure with the bad Durant bad on /off.
Westbrook had lousy on /off in his third season. A season that ended with a playoff loss to the Mavericks where his higher than Durant usage, lousy shooting, high turnovers, -7.1 on plus minus and even worse -15 on /off hurt the team.
Westbrook's on /off has gone back to terrible this season. Means "almost nothing" or something? The -1 on the court means more, but I'll consider both as not encouraging for playoff success or playoff success because of Westbrook.
Westbrook had lousy on /off in his third season. A season that ended with a playoff loss to the Mavericks where his higher than Durant usage, lousy shooting, high turnovers, -7.1 on plus minus and even worse -15 on /off hurt the team.
Westbrook's on /off has gone back to terrible this season. Means "almost nothing" or something? The -1 on the court means more, but I'll consider both as not encouraging for playoff success or playoff success because of Westbrook.
Re: Doncic
After a couple of big games, Doncic's stats are even better in many places.
3 pt fg% is still below average though. A-TO is barely above 2-1.
Usage went from 30.5 last season to 35.2.
On the court raw team plus minus is now good. On /off still modestly negative but last night's +45 cut the negative roughly in half.
3 pt fg% is still below average though. A-TO is barely above 2-1.
Usage went from 30.5 last season to 35.2.
On the court raw team plus minus is now good. On /off still modestly negative but last night's +45 cut the negative roughly in half.
Re: Doncic
On court plus minus is now strong after some very positive recent games. On /off has gotten much better too, though still negative for the season.