Starting Lineup

[Q12543 [enjin:6621299]]

It seems like Mitchell has settled on a fairly steady starting lineup consisting of KAT-Gorgs-Wiggins-LaVine-Rubio the rest of the way. I realize that Prince may be subbed in for LaVine every once in a while, but LaVine will undoubtedly still get plenty of run with the other four even when he doesn't start.

Here is our top three 5-man combinations in terms of minutes played and how they have fared in terms of Ortg and Drtg:

KAT-KG-Prince-Wiggins- Rubio:
Ortg - 103.3
Drtg - 92.6

KAT-Gorgs-Prince-Wiggins-Rubio:
Ortg - 104.7
Drtg - 108.4

KAT-Gorgs-Wiggins-LaVine-Rubio:
Ortg - 119.4
Drtg - 116.6

Wow. Look at what happens to our offense as we first replace KG with Gorgui and then replace Prince with LaVine. It goes up 16 points per 100 possessions!

Double Wow. Now look at the corresponding uptick (which is not a good thing) in our opponent scoring (Drtg). It goes up a ridiculous 24 points per 100 possessions simply by replacing KG and Prince with Gorgs and LaVine.

So the question posed in a prior thread by Kurrdog is whether the starting lineup that includes both Gorgs and LaVine is effectively a "championship" level starting 5. It's obviously way too early to tell, but they score with some of the most elite 5-man units in the NBA (the only one they can't really touch is the Golden State "lineup of death": Green-Iggy-Barnes-Curry-Thompson).

The question is whether this 5-man team can ever defend? Is there a swap-out or two with a hypothetical free agent or draft pick that would balance the best of both worlds? Or can this group grow together defensively and become a killer starting 5?

[thedoper]

The KG Prince lineup is freakishly long, not that there isn't more to that on D. But the length is a good place to start for team D. We become a very average sized lineup with the Lavine / Gorgs, particularly because Gorgs can play small in many ways. This is why the temptation will always be so great to give Lavine minutes at the 1. Any coach would have experimented with it. I just don't know that Mitchell was the guy to bring out Lavine's talent in that area. I wonder what we will see from other coaches if Taylor will pull his head out of his ass and hire a real talent.

[Carlos Danger]

We also have to realize that the number being quoted are relatively small samples. Dieng has 1590 minutes this season and the rating provided in the OP is based on only 244 minutes (or roughly 15% of Dieng's total minutes played).

Most games I've watched on TV, the announcers have singled out Dieng for his good defense. I still believe there is a system/scheme problem more than a talent problem. KG is a HOF player and of course he helps when he's in there. Unfortunately, it doesn't appear as though his mentoring is working a lot on the defensive side so perhaps it's time for a new voice. Towns, Wiggins, Dieng, LaVine, Rubio all certainly have the physical ability to defend well. The effort seems like it's there. But the results are not. Maybe they need time. Maybe they need a new voice/system. Maybe they are just not as capable as I think. IDK. I will say that I'd want to see the same group at least get a chance with a new coach before changing any pieces.

[Q12543 [enjin:6621299]]

We also have to realize that the number being quoted are relatively small samples. Dieng has 1590 minutes this season and the rating provided in the OP is based on only 244 minutes (or roughly 15% of Dieng's total minutes played).

Most games I've watched on TV, the announcers have singled out Dieng for his good defense. I still believe there is a system/scheme problem more than a talent problem. KG is a HOF player and of course he helps when he's in there. Unfortunately, it doesn't appear as though his mentoring is working a lot on the defensive side so perhaps it's time for a new voice. Towns, Wiggins, Dieng, LaVine, Rubio all certainly have the physical ability to defend well. The effort seems like it's there. But the results are not. Maybe they need time. Maybe they need a new voice/system. Maybe they are just not as capable as I think. IDK. I will say that I'd want to see the same group at least get a chance with a new coach before changing any pieces.

The sample size is both small and large at the same time. All three of these 5-man lineups have played together for 200 or more minutes. That's not a lot on the face of it. On the other hand, that counts any time they have been on the floor together the entire season. And since a 5-man unit rarely spends more than 10-15 minutes per game all together at the same time, those 200+ minutes represents a lot of different opponents over many, many games.

So I feel like the Ortg and Drtg figures are directionally correct, meaning it's not just based on pure randomness.

As for Dieng, he's a pretty good pick and roll defender. But if someone is taking him on mano-a-mano in the paint, whether it's a guard coming right at him or a big post player in the paint, that's where he really struggles.

[Carlos Danger]

The sample size is both small and large at the same time. All three of these 5-man lineups have played together for 200 or more minutes. That's not a lot on the face of it. On the other hand, that counts any time they have been on the floor together the entire season. And since a 5-man unit rarely spends more than 10-15 minutes per game all together at the same time, those 200+ minutes represents a lot of different opponents over many, many games.

So I feel like the Ortg and Drtg figures are directionally correct, meaning it's not just based on pure randomness.

As for Dieng, he's a pretty good pick and roll defender. But if someone is taking him on mano-a-mano in the paint, whether it's a guard coming right at him or a big post player in the paint, that's where he really struggles.

I would agree that your argument/data is directionally correct. KG is a HOF defender. He's off the charts on the team based on the metric you are using (Drtg) and nobody else on the team even comes close. So - yes, if you replace KG with anyone else on the team (including Dieng) it's going to have a negative effect on Drtg. But I've never been able to bring myself to completely trust Ortg/Drtg/Net Rating. There seems to be a lot of anomalies that do not match my eyes and/or other metrics. As an example, currently Kevin Martin has the 4th best Drtg on the team (behind KG, Prince and Rubio) and ahead of Wiggins. Furthermore, using the same Drtg metric, KAT and Bazz rate out fairly similar as middle of the pack on the team. Nothing of that makes sense to me which is why I tend not to put too much stock in that specific metric.

[Q12543 [enjin:6621299]]

Carlos, Keep in mind that Martin and Bazz are playing mostly against opposing bench players and KAT is starting and invariably in the game at crunch time, always playing against the opponents' best lineups. Same with Wiggins. If KAT and Wiggins came off the bench and Bazz and Martin were starters, my guess is you would quickly see their Drtgs diverge, to the benefit of KAT and Wiggins.

+/- or ON/Off metrics can be noisy and tricky to navigate. But I feel like we just don't get the total picture without them since the box-score metrics (like PER, WS, VORP, etc.) misses the tons of little things players can do to impact the game, especially defensively.

[khans2k5 [enjin:6608728]]

The best lineup hasn't played together yet, but it's Towns/KG/Wiggins/Lavine/Rubio. The offensive rating doesn't go up much when you swap Dieng and KG, but the defense gets way worse. That lineup has 4 good defenders and 3 good scorers. That's about as balanced as we can get. That lineup would produce the biggest differential if they played together in my opinion.

[Carlos Danger]

Carlos, Keep in mind that Martin and Bazz are playing mostly against opposing bench players and KAT is starting and invariably in the game at crunch time, always playing against the opponents' best lineups. Same with Wiggins. If KAT and Wiggins came off the bench and Bazz and Martin were starters, my guess is you would quickly see their Drtgs diverge, to the benefit of KAT and Wiggins.

+/- or ON/Off metrics can be noisy and tricky to navigate. But I feel like we just don't get the total picture without them since the box-score metrics (like PER, WS, VORP, etc.) misses the tons of little things players can do to impact the game, especially defensively.

Q, I think you are using sound logic in your interpretations. However, I just can't totally trust it myself due to the many variables (like you pointed out above). In fact, if we compare the ratings in both NBA.com and BasketBall Reference, we get completely different results on the same players. So....IDK. I'm sure I'm missing something obvious. But, I just don't use it much for those reasons.

[AbeVigodaLive]

I dunno.

But I do know the most recent concoction makes for some pretty entertaining, high-scoring games.

If the team is going to lose more than it wins... at least have it be more entertaining, right?

[MikkeMan]

The sample size is both small and large at the same time. All three of these 5-man lineups have played together for 200 or more minutes. That's not a lot on the face of it. On the other hand, that counts any time they have been on the floor together the entire season. And since a 5-man unit rarely spends more than 10-15 minutes per game all together at the same time, those 200+ minutes represents a lot of different opponents over many, many games.

So I feel like the Ortg and Drtg figures are directionally correct, meaning it's not just based on pure randomness.

As for Dieng, he's a pretty good pick and roll defender. But if someone is taking him on mano-a-mano in the paint, whether it's a guard coming right at him or a big post player in the paint, that's where he really struggles.

Sample size is still too small if total amount of minutes is around 200. If you check the 4th most frequent Wolves lineup. It is Bjelica, Dieng, Lavine, Martin and Muhammed. It has played more than 140 minutes together and that lineup has Ortg 96.7 and Drtg 96.9. I think there is no way that lineup would be better in defense than Dieng, Towns, Prince, Wiggins and Rubio lineup but still based on drtg numbers is would be. And the difference between 216 and 144 minutes is not that big that we could assume there wouldn't be similar noise in those three most frequent lineup numbers.