Print and Go Back TrueHoop [Print without images]

Wednesday, September 23, 2009
The Great NBA Swap Meet

Posted by Kevin Arnovitz

Yesterday, we made mention of Brett Hainline's swap machine, which uses a player's offensive and defensive efficiency ratings to determine how swapping one player out for another would improve your team's overall performance. 

Once Hainline went live with it, I immediately did what any Los Angeles Clippers fan would do -- nixed the uniquely inefficient Al Thornton from the starting lineup. To fill Thornton's place at small forward, I opted for efficiency poster boy Shane Battier. 

It's important to keep in mind that with salary cap restraints, such a trade would be impossible in the real world, but I was more interested in approximating how much better would the Clippers be with a player of Battier's mold on the wing. 

The results were fascinating. Queen City Hoops estimates that the Clippers would be 10 games better with Battier in Thornton's place. Here's QCH's breakdown:

Troy Murphy

For a larger image of this chart, click here.

To better understand how the Clippers pick up those additional 10 wins, I asked Hainline to walk me through what all this stuff means: 

We are looking at both ends of the small forward spectrum: Al Thornton is a high-volume yet inefficient scorer who plays little defense. Shane Battier is regarded as one of the league's best defenders while being an ancillary player offensively, taking few shots but converting them at a high rate. My fascination with Allen Iverson aside, it frustrates me to see players recognized as being great when all they are really doing is shooting a lot (remember Adam Morrison making the All-Rookie team?). That pet peeve of mine makes this opportunity all the sweeter -- this is a chance to show what kind of impact those players really make.

The first table shows actual statistics from last season. The efficiencies shown are for their respective teams: When Al is on the court for the Clippers, they had a net efficiency of -10.5, but with him off the court, they actually improved to -7.4. The reverse was true in Houston, as Shane helped the Rockets to a +4.9 mark, but that number dipped to +2.8 when Shane was on the pine. The last four columns are individual statistics.

From those numbers, we can estimate how another player would impact a team by replacing someone. By taking the on court efficiencies of the Clippers, and the respective numbers for Al and Shane, we get the numbers you see in the first row of the second table.  Notice a significant boost on both sides of the ball, as their offensive efficiency is predicted to rise by 2.4 points and their defensive efficiency is expected to decline by 2.6.  Here is how we got there:

That gives us an estimate of what to expect with Shane on the court for the Clippers -- a 27 win team. It's not great, but it's 10 more games than when Thornton was lacing them up for them.

But what about when Shane is not on the court? With injuries and age being a concern, we should account for the fact that Shane played over 600 minutes less than Thornton did last season, and that is what the final three rows look at. They're estimates of the team's overall efficiencies, including time with Shane on and off the court -- their whole season in other words. 

The initial row projects Shane to just use up all of Thornton's minutes, meaning the now less efficient off-court numbers are used the same amount as they were last season for the Clippers. Given the estimated improvement the Clippers could see with Shane on the court replacing Al, and the same amount of minutes going to the "bench," a weighted average of the on court and off court numbers puts the Clippers with an overall net efficiency of -6.1, good for 25 wins, which is still significantly better than their actual numbers from last year.

However, what if Shane really does need to play fewer minutes?  Due to age and injuries, he may be good for 2000 and no more. Well, the bench picks up those minutes, so instead of 1300 minutes going to a -7.4 efficiency group, they get 1900 minutes. 1900 minutes to a -7.4, 2000 to a -5.5, and the Clippers project to an overall efficiency of -6.4, dropping another win from total.

The final row describes the case where the Clippers need more minutes from Shane than he could provide in Houston, and he obliges, but his knees still won't let him get all the way to Al's minutes. So, we say 2300 minutes with Shane on, 1600 with him off, and we get a -6.3 efficiency for the Clippers on the season, and they get back to 25 wins.

The notion of a replacement player will always be far dicier in basketball than it is in a sport like baseball, where a variable such as "plate appearance" is relatively easy to isolate. As Hainline explains, comparing two players is far more complicated than handing one guy's minutes to another. No two players' minutes are alike. The instant you place Battier on the floor for Thornton, you immediately increase the offensive roles of Eric Gordon, Baron Davis and Chris Kaman, to say nothing about the team's increased reliance on its bench because Thornton, for all his failings, is a more durable player than Battier.

For an infinite supply of amusement, go to Queen City Hoops and assume the role of basketball Zeus.