Andre Miller and his peers: a comparison with fun graphs

I’m a big fan of charts, especially fun ones. Peer studies are even better, it gets so tiring to see guys compared to league averages, or players in which they have nothing in common.

It doesn’t seem that Roy really respects Andre and his game. I remember a few games when Rex played well and Roy was drawing and dishing to Rex for the open three or letting Rex push the tempo for himself. Roy’s body language screamed “Yeah, do it, GREAT”. But with Andre Roy seems a little pouty like Andre took his ball away. Not saying Roy is right or wrong, just his body language does not look good trying to fit in with Andre.

Dre has the 5th best PER
And the 5th best APG

out of this (interestingly selected, nice job) peer group.

is Dre’s closest peer…

Do you think if we dealt them straight up for each other Roy would still be pouting about not getting his shots?

 feed him and trust him. Roy knows how to finish a game.
 Let Dre do some of his magic in between.

Let LA, a growing Batum, or a healthy Oden be the second option.

is not about Andre Miller.

It’s about how teams can make this work well, and how our team seems to…not want to.

(It’s okay, I remember last Dec-Jan, when Miller started and Roy was burninating. I’m not worried.)

great post.

rare to see use of statistical tools so intelligently. Pro job!

Andre’s not a pretty player like Parker or Westbrook, but he seems to get it done, and for now that’s A-OK with me…just hope Roy will see that.

Rec for putting this together.

But I just don’t see how the statement “Andre Miller can fit next to Brandon Roy” is really justified by essentially a two-factor analysis.

We are to believe that because 3 other successful teams have high usage-poor shooting second wing options that it can work in our case?

Well, first, one of those examples is pretty clearly not comparable. I’ll give you Pierce-Rondo and Westbrook-Durant as being similar to Miller-Roy, but Parker-Duncan? The better comparison is Parker-Ginobili. So basically we are left with two successful examples of the poor-shooting/high-usage #2 wing. In a league with 30 teams and 16 playoff teams though, doesn’t that strike anyone else as a sparse number? Two examples doesn’t tell us it can work in every case, it just tells us it isn’t impossible, but that’s a much different notion.

Also, look at the offensive quality of those two teams. OKC and BOS are good, but they are good mostly because of defense. They were 9th and 5th in the league in DE last year, but a less impressive 12th and 15th in OE, which is what we are interested in here (although Rondo and Westbrook are both fine defenders, so I don’t think these relative efficiency rankings are unrelated to PG play). Just as a reference, the Blazers were 15th in DE and 7th in OE, and the top 5 teams in OE were PHX, ATL, DEN, ORL and TOR (who all have shooters, shooters, and more shooters….)

ALSO … we know that Pop decided to bring Manu off the bench rather than start him and TP together. That makes that an example that runs counter to the OP conclusion. And we could also mention Kobe-Odom as another counterexample as well, who has a higher usage rate and PER than Artest. That makes two championship caliber endorsements for the idea that having two ball-dominating guys play together isn’t good when one of them can’t hit from outside.

In the end though, while other teams can hint at templates of how things can be done well, what matters is what we can get our unique roster to do. Basically, can we get Andre Miller and/or Brandon Roy to stay productively involved in the offense while the other has the ball? And how does that happen? A lot of people seem to want a motion offense, but look at those two guys and tell me which one can list “motion” as one of his strengths? (unlike Rondo, Parker or Westbrook, for example). They are both high usage players not by choice, but because they aren’t very good without it.

MAYBE there is a way to fix that … but I’m doubtful. Certainly the idea that we have to live with some limitations imposed by player availability is true, but I think this is the year we find out whether the combo of Andre Miller and Brandon Roy will work. I wouldn’t say it’s any one person’s responsibility to make it happen, but it’s basically on all of them (players and coach). If it doesn’t, the team won’t panic and make a desperation grab for a weak PG, but it will actively look for opportunities to improve the position and make it the highest roster priority.

The point I was trying to make was that the examples of teams successfully using a high-usage/weak shooting second option wing are pretty sparse, and that having so few successful examples means it’s a tough strategy to make work. Is any part of that not true?

But going through your comments point by point:

1. I’m not really seeing how bringing Manu, who was easily one of their top 3 players, off the bench for half the games in his career is anything but an endorsement of the idea that those high usage guards are more effective apart than together. Pop could have played them together and let offensively capable (who said putrid?) Brent Barry and Beno Udrih run a few minutes in the second unit, but the player pairs data show he decided to maximize the amount of time Manu and TP were separate.

Also, our team is one star guard with a second banana who can’t shoot. Their team is not comparable in a couple ways. First is that their first banana was a PF. Second is that their premier guard was the one that couldn’t shoot, which makes a big difference. It’s a lot easier for a shooter like Many to play off the ball than for a slow non-shooter like Andre Miller. Maybe if we could get Roy to run off of picks like an Argentinian with his hair on fire, we could just let Andre run the show, but Roy doesn’t seem to be that kind of player.

2. BOS and OKC not only aren’t PHX … they aren’t even above average in offensive efficiency. Those examples of “success” aren’t really that successful. Actually, the Blazers with Miller and Roy and a million injuries were actually better on offense last year, which doesn’t disprove the notion that it can work of course, but it does make it odd to claim those other teams show what we could do if we were more successful at integrating Miller. You can argue that OKCs overall roster quality is to blame for their unspecial OE rate, even though they have arguably the league’s 2nd best offensive player on the roster. If you want to do that though, then the same standard should be applied to prior Celtic teams. Sure, they had a better OE the prior year, but that’s with three HOF players still near their primes (which is another reason that team is not a very comparable example to Portland) and a very good cast of role players. Point is: if you put a great roster out there it can make up for deficiencies in fit, just like you said with OKC that a weaker roster will drag down efficiency.

But even with their great roster, Boston’s 6th best OE numbers were beaten by the Blazers that year (#1!), when the Blazers were running with steady Steve Blake at the point, along with youngish Roy and Aldridge and baby Batum to go along with a tough but not offensively exemplary center tandem of Oden and Przybilla. Since it’s pretty tough to argue the 09 Blazers had more talent than the 09 Celtics, it looks like fit really does make a big difference.

3. You’re probably right about the Lakers sending Odom to the bench because he is more of a PF.

But that still leaves the point I think. We have few examples of how the supposed tandem of a star guard and a second banana poor-shooting/high usage guard can work, and even the ones we have all have problems. OKC isn’t that great on offense, Boston isn’t that great on offense without the overwhelming talent advantage of multiple HOFers just barely past their prime, and from San Antonio the Parker-Ginobili combo has the pecking order reversed, their coach has voted with his rotation by opting to separate them as much as possible, AND the Spurs also haven’t been very efficient on offense for quite some time.

I’d really like for the Miller-Roy combo to work well for us … but I’m not seeing much in the way of evidence that it will.