APBRmetrics

Ben F. · Joined: 07 Mar 2005 Posts: 391

Iverson going to Denver presents a situation we've been looking for: AI and Melo together. According to Davis Wylie's stats Iverson was using 35.6% of his team's possessions before the deal, Melo used 32.9%. So what happens here? Who loses possessions? How does efficiency change?

Maybe Bob's simulator making a prediction here would be interesting, as would seeing what David Berri says.

94by50 · Joined: 01 Jan 2006 Posts: 499 Location: Phoenix

My - perhaps the - standard assumption would be that both of their possession percentages will drop (who knows how much), but that their efficiencies will rise (again, who knows how much).

Ben F. · Joined: 07 Mar 2005 Posts: 391

So thinking about this whole issue, I got on the track of working backwards - that is, to counter the debate, what would someone say if their prediction was wrong? In other words, how would someone who assumes the inverse relationship between usage and efficiency respond if Iverson's usage does decrease but his efficiency doesn't increase correspondingly?

And I thought of two obvious answers: 1) Iverson has to learn the offense - he's in a new system which takes some time to get used to. 2) He's not used to a different role - he's had a usage rate over 30 since his third year in the league, so to decrease drastically in usage is a huge adjustment for him, and large adjustments usually don't help efficiency.

And how would I respond to these justifications?

1) Karl's system, and especially his approach to Iverson (at least while Melo is still out) is one of just letting his player's play. I wouldn't think he would force too much on AI in terms of offense. Even when Melo returns, AI is such a premier player my feeling is that the offense would more likely be tailored to Iverson's strengths than force him to learn something completely new. This could be some justification, but it doesn't hold too much weight in my mind.

2) This I think is incredibly important to focus on. The fact is, perhaps we can assume an inverse relationship between usage and efficiency as simply the logic behind the game of basketball (and to me this makes sense), but what Bob Chaikin keeps harping on is that there are so many cases that could go either way. The fact is, a decrease in usage doesn't come anywhere close to guaranteeing an increase in efficiency.

Going further it seems to me that we have to break out of the labeling of "Usage vs. Efficiency." Because I don't think that's important. Rather I think that we have to begin thinking of "roles" and what it means for a player to play within their role or outside of it. And more specifically, how a player will respond when forced outside of their traditional role.

Look at Gary Payton as a good example. Payton's last full year in Seattle he had a usage rate of 26.9 and and offensive rating 9 points above the league average. Clearly he was a player who had the ball in his hands a lot and could score quite efficiently. Fast forward to Miami, 2006. Payton has aged and can't do the things he used to be able to do. He takes a back seat to Wade and Shaq. His usage declines drastically to 13.6 and so does his efficiency - to 1 point above the league average. You could attribute this to aging (and certainly Payton saw some erosion of skills due to getting older), but it also has to do with Payton being forced out of his usual role. He's not a back seat kind of player, and never really was. This was by far the lowest usage of his career. He never needed to learn the skill set of how to pick his spots, hit his jumpers from kick-outs with a quick release. And so when suddenly he could pick his spots, he didn't have a sudden spike in efficiency. To put it in more traditional terms: "Some players just need the ball to be effective."

So how does this relate to Iverson? I would say he's one of those players that just needs the ball. And if this is an argument that makes sense, how does it change the "usage vs. efficiency" debate? I think drastically. It's not so simple that you can look for a change in one and a change in the other. There are too many mitigating factors. Rather we should be looking at how players play when forced out of their roles. Boris Diaw's role changed from Atlanta to Phoenix, and with that came a spike in usage. But he was more comfortable with the latter role, and saw a subsequent spike in efficiency. Could this have been predicted from watching him in Atlanta? The other night against Washington, Marcus Camby certainly showed he could break his role without much loss to efficiency. Granted it was against a porous defense but he was hitting jumpers, making crisp passes, even driving the lane off the dribble. Traditionally he doesn't handle that much of the offensive load, but he showed that he could step outside his role (at least for a game) and thrive. Could he keep that up? I don't know, but I think showing that skill set gives a good indicator he could.

So at the beginning of a great test case for our argument, I'm arguing that I think the context of our argument must be shifted. I'd still encourage predictions as to what will happen, but I'm not sure it's as simple as "loss of usage equals gains in efficiency." How comfortable will Iverson be if forced out of his role? How comfortable will Carmelo be outside of his newfound role? (I read an argument that said Melo had a skill set that could easily be applied to a second-banana scorer - this certainly applies here.) That's what I think needs to be the focus.

Neil Paine · Joined: 13 Oct 2005 Posts: 774 Location: Atlanta, GA

Good points, FFSB. That's why I was hoping there might be certain indicators (ballhandling skills, size, athleticism) that could predict which guys can and which guys can't handle a role change... Anyway, if he's reading this, I think Bob's sim would be great for at least predicting the touches/min for Denver's new starting 5 -- which I assume is Iverson-JR.Smith-Anthony-Najera-Camby. Among AI and Melo, who loses touches, and by how much?

kjb · Joined: 03 Jan 2005 Posts: 865 Location: Washington, DC

Neil Paine · Joined: 13 Oct 2005 Posts: 774 Location: Atlanta, GA

That English-Vandeweghe-Issel team played at a monstrous pace, and the league itself played at a much faster pace back then, too. Those usage rates look huge, but they only amount to 20-25% of team possessions while on the floor. Iverson and Melo are at, like, 35% apiece. Something's gotta give (or, more accurately, somebody's gotta give). 'Tis the season for giving, after all...

asimpkins · Joined: 30 Apr 2006 Posts: 245 Location: Pleasanton, CA

Ben F. · Joined: 07 Mar 2005 Posts: 391

We have David Berri's prediction...

asimpkins · Joined: 30 Apr 2006 Posts: 245 Location: Pleasanton, CA

Since Iverson has sat out a lot of those games, it wouldn't really be fair to compare to their record so far. I'd say last seasons win% would be a roughly fair comparison though.

Ben F. · Joined: 07 Mar 2005 Posts: 391

deepak · Joined: 26 Apr 2006 Posts: 665

You could alternatively look at the OnCourt +/- numbers at 82games.com or (if you prefer wins/losses) look at the floor time win/loss record for Iversion.

asimpkins · Joined: 30 Apr 2006 Posts: 245 Location: Pleasanton, CA

Mike G · Posted: Thu Dec 21, 2006 6:07 am Post subject:

asimpkins · Joined: 30 Apr 2006 Posts: 245 Location: Pleasanton, CA

Denver is already leading the league in pace by a good margin. They are +1.5 possessions over the Warriors, who are +0.9 possesions over the Suns, who are +2.9 possessions over the rest of the quickest teams.

I wonder if they can increase it more without really hurting their efficiency.

Mike G · Posted: Tue Dec 26, 2006 7:42 am Post subject: