Had this conversation in another thread about how awkward the two big lineups were: So I decided to test my hypothesis using NBA.com's stat tracking. Here are the on/off differentials for these two man lineups: Vonleh/Leonard - +1.1 in 15 games played and 3.5MPG Vonleh/Plums: -0.6 in 69 games played and 12MPG Vonleh/Davis: -0.3 in 31 games played and 6.2MPG Davis/Leonard: +0.6 in 57 games played and 14.3MPG Davis/Plums: +1 in 24 games played and 4MPG Leonard/Plums: -0.5 in 51 games played and 6.1MPG So here is what I am getting out of this: -All of Vonleh's minutes with a significant enough sample size indicate that he is not effective at all. -Stotts realized early on that the Leonard/Plums lineup wasn't working and moved Leonard to the bench. The Leonard/Davis lineup is the only one that has a positive on/off differential with a big enough sample size. -The Vonleh/Plums lineup isn't really doing much better, but the Davis/Vonleh lineup was just bad as well, so overall a very good move. -Davis and Plumlee play well together, as I suspected, but the sample size is very small. I'd like to see Stotts experiment more with this one. Next I wanted evidence that the small lineups were better, so I decided to see how the bigs did when paired up with Aminu or Harkless. Harkless/Plum: +0.7 in 65 games played and 7.4MPG Harkless/Leonard: +0.3 in 46 games played and 9.9MPG Harkless/Davis: -0.6 in 68 games played and 10.8MPG Harkless/Vonleh: -0.9 in 28 games played and 5.4MPG Aminu/Plum: +0.5 in 79 games played and 22.3MPG Aminu/Leonard: +0.3 in 59 games played and 7.7MPG Aminu/Davis: +0.1 in 73 games played and 6.1MPG Aminu/Vonleh: -0.3 in 66 games played and 12.5MPG It seems pretty clear that in general: - The bigs pair up better with a wing. - Okay, Vonleh is just bad. Of course these don't really say anything about who else is playing with the team so I also looked at some of the most used 5 man lineups with a big enough sample size to see if there is a trend. Aminu,Al-Farouq - Harkless,Maurice - Lillard,Damian - McCollum,CJ - Plumlee,Mason: +1.8 in 33 games 7.2 MPG Crabbe,Allen - Davis,Ed - Henderson,Gerald - Leonard,Meyers - Lillard,Damian: +1.3 in 30 games 4.3 MPG Aminu,Al-Farouq - Crabbe,Allen - Lillard,Damian - McCollum,CJ - Plumlee,Mason: +0.7 in 36 games and 3.5MPG Crabbe,Allen - Davis,Ed - Henderson,Gerald - Leonard,Meyers - McCollum,CJ: +0.1 in 34 games and 5.7MPG Our most used line up is Aminu/Vonleh/Plum/CJ/Dame which was -0.2. Based on these stats, the Aminu/Hark/Lillard/CJ/Plum lineup is by far the best, and I'm glad Stotts realized that. Some other interesting notes: The Aminu/Hark/Plum is one of the better 3 man lineups at +1.1 in a pretty good sample size. Aminu is CONSISTENTLY in the top 2, 3, 4 and 5 man lineups. No matter who he's playing with, Aminu is clearly impacting the game. Meyers is consistently a positive as well. In fact, he's in 3 of our top 2-man lineups with a significant enough sample size. Meyers/CJ is a +2.2 (by far the best 2 man lineup we have), Meyers/Hendo is a +1.9, and Meyers/Crabbe is a +1.1. Meyers/Dame is also a +0.6 and Meyers/Aminu is a +0.3. In fact, it's pretty hard to find a negative lineup with Meyers in it, regardless of sample size. If this comes as a surprise to you, it comes as a surprise to me too. But for all you people saying Meyers is a ineffective.....the stats point the opposite direction. I think he's clearly been effective in his bench role. If you have made it this far, congratulations! TL;DR: Small lineups rock, big linups suck.
WOW! Great work, and the bolded section is certainly shocking. Interesting to see how well our bench mob played with Lillard. I'm not ready to give up on Vonleh yet--I'm hoping for development over the summer and better productivity next year, but he should certainly be relegated to 9th-man status for the rest of this season.
A lineup I don't see (but it's pre-coffee) is CJ/Dame/Chief/Hark/Legend; I bet it would be interesting; Leonard hangs back while Hark drives, potentially giving the lineup two different looks.
Great analysis. Yeah I agree the two big lineups don't work. I'd certainly resign Harkless this summer; and bring in another SF/PF hybrid. Marvin Williams has been excellent as such this year. I'd target a player in his mold. So who are the top two options out of Plumlee, Davis, Meyers? If we only play one big; which I agree is far more effective, one of these guys should be the 3rd string option.
It really seems like you have three different "builds": Passing/Pick&Roll (Plums), Offensive Rebounding/Put-backs (Davis), and Shooting/Defensive Rebounding (Meyers). Each have different weaknesses (shooting, FTs, help defense). It's totally situational which two are best.
Yeah....would be nice to have a front court player who is more multi-dimensional. As it stands, each guy is only good at one or two things (outside of Vonleh, who is good at none).
I'd like to see us turn Plums and Leonard into one center who could do more (and also isn't an insane asshole -- have to clarify that for some folks!). If I had to keep only one, I'd keep Davis.
Yeah, I like Davis the best, although all the stats say Plumlee is the most effective. Probably a mixture of both, but Plums will get paid and might fetch a greater return anyways.
I'd keep Davis because if we took two centers and turned them into one center, that center would start and Davis is best off the bench. If our best center is a marginal starter, you don't keep him and also look to improve on him.
