Starbury and the Celts: Can He Help?

David Berri makes a few good points here. First note that, in terms of wins produced, Rajon Rondo is the most productive player on the Celtics. His WP48 (wins produced per 48 minutes) is more than 3 times the league average, making him an elite player, and one of the most productive point guards in the NBA. Eddie House, Rondo's backup, is also, because he is an efficient scorer who does not often turn the ball over, an above average NBA player. Both Rondo and House are good because they reduce the possibility of empty offensive possessions by scoring efficiently and not giving the ball to the other team.

Stephon Marbury, by contrast, is by this measure a below average NBA player. An inefficient scorer who takes a lot of shots and turns the ball over is a player who creates a great many empty possessions, and thus hurts his team's chances of winning.

I have one issue, with Berri's post, however. I don't think the Celtics are bringing Marbury in to take minutes away from Rondo and/or House. Given how much more effective Rondo and House are than Marbury, that would be a tremendous mistake. I think that they are brining him in - he's expected to sign with the team sometime today, after he clears wavers - to replace Tony Allen.

While he is an inefficient scorer, he is able to create shots and handle the ball against pressure, two skills desperately needed on the Celtics' second unit since Allen went down. He is also big enough to guard the shooting guard position, which means that he, like Tony Allen, can play alongside House on the second unit rather than take minutes from him.

Finally, since Marbury - while a below average player at this point in his career - has some NBA-level skills and is signing at a pro-rated portion of the league minimum salary, if his performance on or off the court hurts the team, they can easily cut him.

How the Rockets Stopped LeBron James

Cleveland has arguably the best team in the NBA. The Cavs rank third in offensive efficiency (a fact often overlooked because their slow pace reduces the number of possessions, and thus points, in a game) second in defensive efficiency, and first in efficiency differential, arguably the best single measure of team performance.

They also have LeBron James, by almost any measure the best player in basketball today, and one of the best to ever live.

Last night, however, the Houston Rockets - a well constructed team with a great defense that ranks fifth in the league in defensive efficiency - throttled the Cavs 93-74, and frustrated LeBron James all night. In his 34 minutes of playing time he shot only 7-21 from the field, was held to 6 free throw attempts, and, for the first time in his career, had no assists.

How did they do it? Kevin Arnovitz of TrueHoop explores this in painstaking detail. Remember, hoop heads, defense, while often overlooked, is half the game.

LeBron James' Pace-Adjusted Triple Double

If LeBron James played like he has this year at the pace NBA games were played at in 1962 (when Oscar Robertson became the only player ever to average a triple double for a season), his per game averages would be a stunning 40.1 points, 10.3 rebounds, and 10.0 assists. No wonder he is on track to post the highest PER in history, topping even Michael Jordan's best seasons!

But, because he's playing fewer minutes than he has in the past (because the Cavs are winning a lot of blowouts, while also trying to keep him fresh for the playoffs, James is over 3 minutes per game off his career average, and over 2 minutes a game off his previous season low, which was in his rookie season) and because the Cavs play at such a slow pace, no one outside the stat-geek community seems to be noticing.

Anyway, this piece by Neil Paine of Basketball-Reference.com is a must read for someone who wants to understand how to translate player performance across time.

Best Missed Free Throws EVER!

I'm a little leery of "sports as a metaphor for life" arguments. And don't get me started on the "builds character" stuff. I've seen a lot of bad characters formed in youth sports leagues.

But this is a story that blows all of my skepticism of the value of sports away. Of course, it didn't have to happen like this. The coach could have put winning the game above all else. The player could have been concerned about his personal stats. But they didn't.

That's exceptional. And exceptions move me.

Appreciating Vince Carter

I've never been a huge Vince Carter fan. I tried three different drafts of a sentence to follow that, only to realize that I don't really know why I don't like him. Unlike most Kentucky fans, it isn't because of his time as a Tar Heel. I'm the rare Kentuckian who also pulls for Carolina, as I've got family in Chapel Hill, where my grandfather has taught for, well, ever.

Now I'm confronted with TrueHoop exposing my anti-Carter bias with some pretty good evidence that, far from being over-rated, Vince Carter is actually considerably under-rated. Not only is he an elite-level offensive talent, but advanced statistics demonstrate what scouts have known for years: despite his reputation for indifference on the defensive end of the court, he's one of the better perimeter defenders in the NBA.

Add that to his otherworldly dunking ability and his reputation for - despite bailing on Toronto - being a really nice and relatively selfless guy, and perhaps I ought to pull for Carter.

Appreciating Shane Battier As a New Way to Think About Basketball

While I detest Duke, I have always had an appreciation for Shane Battier. In a sense, he is the anti-Dukie. What I mean by that is that while players from Duke tend to be vastly over-rated, Battier is consistently under-rated. He's the kind of player - like Chuck Hayes or Tayshaun Prince - who should have been coached by Tubby Smith.

Anyway, while this article by Michael Lewis of the New York Times pretends to be about Shane Battier, it is really about a new way of thinking about basketball. And, as you might guess from my posts here on the sport, I strongly endorse that new way of thinking about basketball.

Or, perhaps it isn't a new way of thinking about basketball as it is a new way of thinking about the statistics that purport to measure player performance in basketball. The traditional box-score statistics tend to reward selfishness over selflessness. This article does a great job of digging into why that is, and what is now being done about it. A great read for anyone who really cares about basketball.

Kobe v. LeBron

David Berri demonstrates a few things remarkably well:

1. Kobe's 61, 0, and 3 was, surprisingly, greater than LeBron's 52, 9, and 11, at least in terms of Wins Produced per 48 minutes. This is because of Kobe's otherworldly scoring efficiency in that game.

2. Overall, LeBron James is a much more productive player than Kobe Bryant. This is true any way you slice the numbers. Both are good players, but LeBron beats Kobe (and just about anyone else, ever) hands down.

3. Bill Simmons is a much, much better writer than David Berri. Despite Berri's excellent insight, the highlight of the article, hands down, is the Simmons quote. Berri's prose, next to it, looks somewhere between pedestrian and perfunctory.

Mo Williams, All Star?!?

As David Berri notes, statistically speaking, Mo Williams is essentially the same player he was last year in Milwaukee. Last year, nice player, above average, but by no means an All Star. This year, same guy, roughly the same performance, but he's an All Star. What gives?

His team is winning more games not because he's a much better player, but because he's on a different team, surrounded by better players (including especially LeBron James!). I'm all for winning being the ultimate goal in basketball, but team winning is not the best measure of individual player performance.

To wit, the team I ran with won every game we played last week in the Y. Does that mean I was a better player than everybody on every team we beat? Hell no! Anyone who saw me play could confirm that!

The "Best Player on the Best Team" MVP Criterion is Stupid

Watching SportsCenter this morning, several NBA "experts" argued that the true definition of MVP is "the best player on the best team." They did this to demonstrate that Kobe Bryant should be considered, at this point in the season, the NBA MVP, because he's the best player on the team with the best record.

That's simply ridiculous.

By that line of reasoning, you could argue that if Boston has the best record at the end of the year, Rajon Rondo could be the NBA MVP. Don't get me wrong, Rondo is an incredibly useful (and damned fun to watch) player, but he's no NBA MVP (though he really should be an All Star).

In the same way, Kobe Bryant is not the MVP, was not the MVP last year, and should not be the MVP this year. Every decent statistical measure shows LeBron James and Chris Paul as clearly the two best players in the NBA, and two of the best players ever. Kobe Bryant is well above average, but simply not in the same class as those two.

I agree that to be "most valuable" a player has to create some actual value, which means wins. But it simply isn't the case that the "best player on the best team" creates the most wins. The case of the Celtics demonstrates this pretty well. Statistically, the best player on that team is Kevin Garnett, followed closely by Rajon Rondo. After that come Ray Allen and Paul Pierce. All four of those players are very, very good. But the team wins a great many games not because any of those four are "most valuable," but because those four are very good, and are surrounded by above average players.

The same is true on a smaller scale with the Lakers. Kobe Bryant obviously creates a few wins for the Lakers. He is an excellent player. That has some value. But he is an excellent player surrounded by a few excellent players (Pau Gasol, Lamar Odom) and more than a few good ones (especially the vastly underrated Trevor Ariza).

Being surrounded by those players neither makes Kobe Bryant the most valuable player in the NBA, nor disqualifies him from being the most valuable player. Rather, it helps explain team success. The Lakers have the best record in the NBA right now not because Kobe Bryant is the best or most valuable player, but because they, like the Celtics, have a great many good players on their team.

Remove Kobe Bryant from the Lakers and they still have an excellent front court (Gasol, Odom, and, when healthy, Andrew Bynum) and a decent back court. In fact, given how well Ariza has been playing, they may not lose a whole lot, if Bryant's lost minutes go to him.

The stories are very different for LeBron James and Chris Paul. Tom already noted Paul's value to the Hornets by pointing out how they've played without him. Similarly, picture the Cavs without James. While they'd still have a few good players, would they even be a playoff team?

In this case, advanced statistics do an excellent job of measuring the relative "value" of MVP candidates. Those statistics tell us not only that James and Paul and far and away the two best players in the NBA, but also that their teams would lose many more games and win many fewer games without them.

They both create more value, and more wins, for their team, than Bryant does for his.

Because I'm so pedantic, I wanted to push this a little bit farther: First, identify the best teams, and then identify the best players on those teams, too see if picking the "best player from the best team" would produce a viable MVP choice.

First for the best teams. I'm using John Hollinger's efficiency ratings to identify the best team. The thinking here is that the difference between offensive efficiency and defensive efficiency - efficiency differential - identifies the relative strength of a team. This is like point differential adjusted for pace.

By that (or any other measure) four teams stand out: The Boston Celtics, the Orlando Magic, the Cleveland Cavaliers, and the LA Lakers. In terms of efficiency differential, here's how they rank:

1. Cavs: 11.3
2. Celtics: 10.3
3. Magic: 10.0
4. Lakers: 8.8

From this measure it seems clear that the Cavs are still the class of the league, though the Lakers and Celtics each have slightly better won-loss records.

Now for the best player on each of those teams, according to Hollinger's PER (Player Efficiency Rating), one of the two best statistical measures for performance.

1. Cavs: LeBron James 31.86. This mark not only leads the league, but threatens to break Michael Jordan's record for best single season PER. James is having a historic season, and is clearly a worthy choice for MVP. So, did the criterion of "best player from best team" work? Maybe, but the extent to which is worked may have been a coincidence.

2. Celtics: Kevin Garnett 21.23. Here is where the argument begins to break down. If the Celtics catch the Cavs (and, in terms of won-loss record they already have), is Garnett the MVP? His PER ranks 18th in the league. While PER does not measure defense, given James' emerging strength as a defender it is hard to argue that Garnett's defensive prowess makes up a 10.63 point and 16 player gap.

It is worth noting that Rajon Rondo's PER of 18.89 ranks second on the Celtics.

3. Magic: Dwight Howard 25.49. While the distance between Howard's 25.49 and James' 31.86 may seem great, Howard ranks 4th in the NBA in PER. That shows just how stellar James' season has been. Most years Howard would be a pretty good choice for MVP, especially considering his defensive impact.

4. Lakers: Kobe Bryant 25.22. Bryant ranks 6th in PER, behind James, Chris Paul (29.75), Dwayne Wade (28.24), Howard, and Tim Duncan (25.38). In fact, given the stability Duncan has provided an injury plagued Spurs team (2nd in the West despite losing many of their top players for chunks of time) it could be argued that Duncan would be at least as deserving of the MVP as Bryant.

All of this is to say that when two separate commentators argued that Kobe Bryant was clearly the NBA MVP thus far this season, because he's the best player on the best team, they were full of crap. It is neither clear that Bryant is the best player on the best team, nor that choosing the MVP based on that criterion is the best idea.