Position Adjustments

When considering a basketball player's quality, you obviously need to consider position. But still, certain kinds of crude position adjustments seem to me to produce perverse results. For example, see Dave Berri's comment on Rashard Lewis:

And all this returns to a point I made last summer when Lewis was signed. At small forward, Lewis can post numbers that might justify his contract. At small forward he is an adequate rebounder and an outstanding scorer. At power forward, though, his inability to rebound is problem.

Lewis is listed at 6'10" and Hedo Turkoglu, the Magic's starting small forward, is listed at the exact same height. What's more, Turkoglu is a somewhat better rebounder than is Lewis. But it can't be right to say that if the Magic were to start calling Turkoglu the power forward and Lewis the small forward that Lewis would suddenly become worth paying more than he is now. The ability to guard multiple positions is an asset. If Lewis couldn't hack it at the four and had to be played strictly as a small forward, he'd be a less valuable player, not a more valuable one.

Comments

Larry Bird has a very arcane formula for placing a value on potential acquisitions. The crack-whore/marijauna/stolen gun quotient figures prominently, as does how much armor plating their SUV brings to the table. Extra points are awarded for quality of bar fighting and the ability to participate in rape while retaining plausible deniability.

Posted by steve duncan | March 17, 2008 12:23 PM

I think the point is Rashard Lewis' contract would be justified from the Magic's point of view if they didn't already have Turkoglu doing a lot of things Lewis does well, pushing him out of the SF spot. Rashard Lewis playing PF is a lot worse than Rashard Lewis playing SF.

However, Rashard Lewis playing PF is better than Brian Cook playing PF and Rashard Lewis warming the bench, so yes he is "more valuable" than that alternative.

Posted by right | March 17, 2008 12:25 PM

I think which interpretation makes sense depends on how you answer "compared to what?" If you comparing a player in isolation to the "replacement level player" then the position adjustment exercise would be - what's better:

1) Lewis at SF and a replacement level PF (who will have good rebounding to offset Lewis' weakness) or
2) Lewis at PF and a replacement level SF (who won't rebound so well, paired with Lewis' lower rebounding)

In that case it makes sense to say that any player at position X needs to be compared to the kind of production (scoring vs. rebounds vs. assists, etc) we'd expect from the average player at that position. This exercise is probably more useful for talking about players in the abstract and/or talking about moving players from one team to another.

You're asking a different question: given the particular players that are already on team X (and their idiosyncrasies), what's the best way to use those players? From this perspective changing the label of Lewis from SF to PF would just mean calling Turkoglu a SF, and wouldn't change anything. Given that, it'd always be better to have a player who can play well at the 3 and ok at the 4, rather than a player who is just a 3. However, comparing across teams, it's probably fair to say that Player A who has played all his minutes at SF (and played well) has been more productive this year than Player B who split time at PF and played less well there.

Posted by Scipio | March 17, 2008 12:26 PM

Nice spinning on MSNBC. "She has had less support in general" so one would expect she'd have "less support online." That's true to the extent that Obama has a couple hundred thousand more votes than she does, but this is out of about 10 million a piece, so in the large scheme of things, she's running about even with him.

Posted by Joe Strummer | March 17, 2008 12:28 PM

I think the real question is whether either Turkoglu or Lewis can _defend_ a true power forward. In practice, there's nothing wrong with playing two SFs as long as one can defend post players.

Posted by AWC | March 17, 2008 12:33 PM

Berri sounds like he's talking about baseball or football, where you have set positions and somebody's got to fill each one. Basketball positions are fluid -- you can start 2 point guards or 3 small forwards if you want. This is yet another way in which Dave Berri is a moron. But my favorite example is that he thinks the Pacers got the better of the DunMurphy for Jackson & Harrington deal. Because Stephen Jackson is terrible. And Dunleavy should have been an all-star. Right.

Posted by too many steves | March 17, 2008 1:02 PM

exactly right, awc. Was magic a PG? Sure, on offense, but he was a 3 or 4 on defense and on the boards. LeBron is anywhere between PG and PF depending on your offensive philosophy, but offense is more flexible than defense because you're controlling who holds the ball where. You can make your 7 footer the PG as long as he doesn't have to bring the ball up the court, it might work (ask Don Nelson).

The problem with lewis at the PF is that he's a beanpole who can't rebound or defend anyone over 225 lbs. Lewis isn't worth the money the magic gave him under the best of circumstances, but he's really not worth it to be a poor PF, because, as the Rockets have shown, good PFs are easy to find. Or is being a PF on the Rockets like being a RB for the Broncos?

Posted by Crusty Dem | March 17, 2008 1:07 PM

The hilarious thing is Berri's absurd gymnastics to get around the fact that he refuses to acknowledge the question of defense.

A PF needs to get defensive rebounds. A SF doesn't have to get as many, but needs to be able to guard faster players. Lewis is a good offensive player. This is well known. But can he play defense? Can he get enough rebounds as a PF or guard other true PFs? And if not, is he fast enough to guard SFs? Of course, for the Magic, this 2nd question is irrelevant, b/c it's beetter to make Lewis the PF and Turkoglu the SF.

Berri has this weird thing where he requires PF to get more rebounds than SFs. This doesn't make any sense when evaluating players b/c the name their coach gives them now determines their ability. The problem is the missing piece: guarding smaller, faster players. Without it, this theory is incoherent.

Posted by mpowell | March 17, 2008 1:22 PM

I think Scipio is basically right. Theoretically, having a guy who can play either position is more valuable than having a guy who can only play one position. But if the coach actually plays Lewis at the position where he is less effective, it is likely that the time he is actually on the floor will be less productive.

Now, as for Orlando, the coach may be doing the right thing, since Lewis-Turkoglu-Evans is likely better than Cook-Lewis-Turkoglu. Evans is closer to replacement level than Cook is.

But the failure here is on the GM's part in not getting a better PF, to put Lewis in a position where he is more useful. It's not Lewis's fault that the team has to play him at his less effective position.

Posted by Al | March 17, 2008 1:29 PM

Terrible, ignorant post, Matt. Makes you look like you view the NBA strictly from a fantasy sports position.

The whole point is defensive matchups, and it does matter, a lot.

Offensively, what position a player plays does not matter as much, as your 2 & 3 typically play the wing, and your 4 & 5 the low or high post, and depending on the offense, can be interchangeable to an extent.

But on defense you need to matchup, and playing Lewis at the 4 presents serious problems for the Magic against teams with legitimate power forwards who will kill Lewis on the boards and on the blocks.

It's sort of like saying that Player X can play Gold Glove caliber defense at 1st Base, but would suck at short.

Lewis is a very mediocre power forward but is getting paid like a great one.

Posted by Eric Blair | March 17, 2008 1:52 PM

Worse pundit, Berri or Kristol?

How many times does it need to be demonstrated that Dave Berri knows FUCK ALL about basketball before ostensibly smart people stop linking to him?

Posted by Pooh | March 17, 2008 1:54 PM

Terrible, ignorant post, Matt. Makes you look like you view the NBA strictly from a fantasy sports position.

Most of Matt's basketball posts accomplish that particular objective.

Posted by Swopa | March 17, 2008 2:03 PM

So I decied to torture myself by reading more Dave Berri, and I found something dumber than his opinion of Stephen Jackson: Dave Berri thinks Brent Barry is a better player than his father was. I'm not kidding:
http://dberri.wordpress.com/2008/03/09/the-better-barry/#more-746

Posted by too many steves | March 17, 2008 3:23 PM

The reason the Lewis signing was terrible is because they offered him about 25-30 million more than the next highest bidder. Lewis is not a max contract player, for many reasons.

Posted by Craigory Gregory | March 17, 2008 3:24 PM

Too many,

of course he does. Because, and repeat after me, he knows FUCK ALL about basketball...

Posted by Pooh | March 17, 2008 3:28 PM

The ability to guard multiple positions is an asset. If Lewis couldn't hack it at the four and had to be played strictly as a small forward, he'd be a less valuable player, not a more valuable one.

Actually Rashard Lewis just about hacks it guarding the SF as well. He's a below average defender that tries to make up for it with his height where he will draw some shot blocks, but most forcing shot adjustments from the player he's guarding.

The problem for Lewis is that he's soft. He doesn't like contact. You bruise him up early, simmer him for the 2nd and 3rd, and by the 4th he's got nothing left he's practically falling apart.

The Lewis contract completely destroyed contract extensions for the 2004 crop of rookies. After Lewis signed that monster deal, Okafor, Gordon, Deng, and Igoudala all turned down their 5 year contract extension offers.

Posted by Gabriel Sutherland | March 17, 2008 4:06 PM

Re: Steves

You have to love the way Berri writes, though. He talks about how Brent posted a better shooting% than Rick, but then acknowledges how Rick partially makes up for it through a high ft%. It really does give the impression that he doesn't know anything about bb, b/c he never even thinks that Rick's high ft% might suggest that his low fg% was due to some difficult to capture detail, like the fact Rick was the focus of the opponent's defense. A good statistian would acknowledge the potential inaccuracies in his evaluations. With Berri, there is only Wages of Wins. It is God's Truth on Hoops revealed to use through Berri.

Posted by mpowell | March 17, 2008 5:24 PM

If Lewis couldn't hack it at the four and had to be played strictly as a small forward

BTW - I'm curious about this phrase. How do you measure "can't hack it"?

Isn't it possible that Lewis can't "hack it at the four", but yet SVG is playing him there anyway because their potential replacement - Brian Cook - can't "hack it at the four" either? In fact, isn't it possible that nobody on the roster can "hack it at the four"? So I really don't get where Matthew's coming from here.

Posted by Al | March 17, 2008 5:29 PM

The other faulty assumption that both Berri and Matt make is that positions are static, where in fact the role of the "small forward" or "power forward" varies widely between teams. Both Lewis and Turkoglu aren't great rebounders, but that matters a lot less when you're playing alongside Dwight Howard.

Posted by vanillablue | March 17, 2008 5:40 PM

"A good statistian would acknowledge the potential inaccuracies in his evaluations."

An intellectually honest person would acknowledge the limits of their model.

Posted by Petey | March 17, 2008 5:42 PM

Wait, is petey going to counsel us on intellectual honesty now?

Chutzpah, sir.

Posted by Pooh | March 17, 2008 6:57 PM

"Wait, is petey going to counsel us on intellectual honesty now? Chutzpah, sir."

Passionate advocacy while making sure to color within the lines of intellectual honesty is the entire Petey schtick.

Posted by Petey | March 17, 2008 7:07 PM

Passionate advocacy while making sure to color within the lines of intellectual honesty is the entire Petey schtick.

Well, there certainly coloring going on.

Posted by SomeCallMeTim | March 17, 2008 7:10 PM