Last year the Dallas Mavericks won 67 games and appeared poised to win the first NBA championship in franchise history. And then Dallas met the Golden State Warriors in the first round and a different kind of history was made.
In the off-season the Mavericks made very few changes, returning the same starting line-up and most of the same players off the bench. And with basically same the cast of characters, the expectation was the results would be similar.
Well, after 28 games in 2007-08, this team is not quite the same. The record stands at 19-9, which places them on a pace to win 56 games. And although this record is off the pace of last year, it actually exaggerates how good this team has been.
Last year this team’s efficiency differential – offensive efficiency minus defensive efficiency – was 7.8, which is consistent with a team that wins about 60 games. This year their mark stands at 4.2, which translates into about 51 victories. In sum this team is about nine games off last year’s pace.
So what happened? To find an answer, let’s look at the productivity of the individual players. Table One offers two projections. The first looks at how many wins we could expect if each player plays as well as he did last year. The second estimates how many wins each player will produce if he keeps playing as well as he has this season.
Table One: Projecting the Mavericks after 28 games
When we look at Table One we see that most of the main cogs on this roster are doing pretty much what they did last year. Josh Howard and Erick Dampier each posted a WP48 [Wins Produced per 48 minutes] in excess of 0.200 last year. This year they are producing at this level again. Jason Terry, DeSagana Diop, and Devin Harris were all above 0.100 last year. In 2007-08, each is above average again.
The Nowitzki Story
And then there is Dirk Nowitzki. Last year, on the team with the most wins, Nowitzki produced the most victories. When the season was over, Nowitzki’s Wins Produced stood at 18.1. Although seven other players in the Association eclipsed this mark, it was Nowitzki who was named Most Valuable Player by the sports media.
This year, as was revealed a couple of days ago, Nowitzki’s production of wins has declined. In fact, his production is now being eclipsed by seven players within his own division. One of these players – Josh Howard – plays in Dallas, which means that last year’s MVP is not even the most productive player on his own team.
When we look at the magnitude of Nowitzki’s decline we see that most of the team’s drop-off can be laid at his doorstep. And this leads us to ask, what has happened to Nowitzki’s production?
The answer is in Table Two, which compares Nowitzki’s production of each individual stat – both this year and across his career — to what we see from the average power forward.
Table Two: Dirk Nowitzki after 28 games in 2007-08 and across his career
Let’s start with the MVP year. When we compare Nowitzki in 2006-07 to the average power forward, we see a player who is above average with respect to shooting efficiency, scoring, rebounds, and assists. He was slightly below average with respect to blocked shots and steals, and virtually average with respect to turnovers.
When we look at 2007-08, we still see a player who is above average with respect to shooting efficiency, scoring, and assists. But his advantage in shooting efficiency has declined significantly. And this decline is almost entirely due to a drop-off from three-point range. From two-point range, Nowitzki shot 51.5% last year and is shooting 51.1% this season. But once he steps beyond the arc we see a different story. Last year he shot 41.6% from downtown. This year he’s only connecting on 30.1% of these shots. In sum, the MVP is missing quite a bit.
Beyond shooting efficiency, Nowitzki is also below average on the boards and committing turnovers at a rate not seen since his rookie season. It’s important to note that Nowitzki is still a very “good” player. But because he has been a “great” player, his drop-off in performance has a dramatic impact on the team’s fortunes.
History Repeats
What’s odd about this story is that this has actually happened before. In 2002-03, Nowitzki posted a 16.3 Win Score per 48 minutes, a mark quite similar to what he did during his MVP season. The Mavericks also won 60 games, and with an efficiency differential of 8.2, Dallas was actually a bit better than what they were last year.
The following year, though, this team declined. Nowitzki’s Win Score per 48 minutes fell to 13.7. Like this year, his shooting efficiency and rebounds also declined relative to what he did the previous campaign. Overall, the Mavericks dropped eight games in the standings and one could argue -as Table Three illustrates — that this decline was due to the drop off in Nowitzki’s production.
Table Three: The Dallas Mavericks in 2002-03 and 2003-04
After 2003-04 the Mavericks apparently solved their problem by letting Steve Nash depart for Phoenix. Without Nash in 2004-05, Nowitzki’s Win Score per 48 minutes rose to 15.6.
Okay, Nash really wasn’t the problem in 2003-04. He was actually the team’s best player. Actually I don’t know why Nowitzki’s performance slipped in 2003-04 in the same fashion it slipped this year.
As I have often said, the numbers tell us how productive a player has or hasn’t been. The numbers don’t tell us why. For the why question, we turn to the coaches.
Although the numbers don’t give us all the answers, it’s useful to review what answers we did receive.
1. The numbers told us that the Mavericks are not as good as they were last year.
2. Looking at the players, we learned that most of the players who led this team last year are about the same this year.
3. The lone exception is Dirk Nowitzki. His performance has dropped off.
4. Looking at Nowitzki’s numbers, we see a decline in shooting efficiency from beyond the arc, fewer rebounds, and more turnovers.
5. We also learned that such a decline with respect to Nowitzki happened before, and that the decline in his numbers led to a similar drop-off in team production.
All in all, it appears that the numbers do tell us quite a bit. Not everything, but to be able to spot a decline in a team and trace it to specific aspects of a specific player seems pretty neat (at least to me).
– DJ
Our research on the NBA was summarized HERE.
The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.
Wages of Wins Journal 
Darrel
December 23, 2007
So the difference is that… he’s missed eight 3 pointers after 28 games?
Huey
December 23, 2007
Hey Dave, I know this is unrelated to your post, but I was hoping you’d have some thoughts.
In basketball, I often see a coach bench a player in the first half once he gets two fouls. To me, this seems like a poor decision, as the coach is guaranteeing the outcome he wants to avoid — the loss of minutes for a player. It also seems frustrating when the player in question won’t even come close to fouling out later in the game.
Yet, coaches constantly make this decision. What data would I have to look at to conclude that this decision isn’t sound?
dberri
December 23, 2007
Huey,
This is a problem, because few coaches don’t make a different decision. Ideally we would compare what happens when a player with two fouls is left in to what happens when they are removed. If they are left in and they don’t tend to foul out, then pulling them is a bad idea. But to see that, you need some coaches to leave such players in.
Curtis
December 24, 2007
I don’t know that there is a way to measure it, but the counter-argument is that a player in foul trouble plays differently than a player not in foul trouble. A player who is consciously trying to avoid fouls is less likely to block shots, less likely to get offensive rebounds, and so on trying to avoid those fouls. So while Matt Bonner is clearly not as good a player as Tim Duncan, for four minutes at the end of the half, an aggressive Bonner might be expected to be more productive than a passive Duncan. And by sitting Duncan while he is in marginal foul trouble, the coach guarantees that he re-enters with a more aggressive and productive mindset.
Ken
December 25, 2007
Your article ignores some rather important facts. Yes, the record is 19-9, but the Mavs are 5-0 in the last five games and 7-1 in the last eight. Dirk has averaged over 30 points in the last 3 games, and has shot over 30 five times in December. The article misses the fact that Avery Johnson admits that he was tinkering–specifically asking Dirk to take fewer shots and work on the passing, both to get other team mates off to a good start AND to work on handling the double teams… Since the last 8 games have provided a great ride, it is pretty silly to make such dramatic pronouncements about the team, and the MVP based on the first 20 games, when the team was injury plagued and tinkering with a lineup.
Costa
December 26, 2007
Ken, you might not realize it, but you’re actually agreeing with dberri in your comment, not refuting him. dberri himself says that the numbers are telling you “what”, but it’s up to the observer like you or me to fill in the “why”, which is exactly what you’re doing in your comment. Nowitzki’s production is down this year. The numbers bear that out, but they don’t say why. Your comment presents a potential and sensible-sounding reason why. You may or may not be right. I haven’t watched enough Mavs games this year to offer an opinion on that. I just found it neat that while it seems like you think you’re offering an opposing viewpoint to the article, in truth, you’re supporting it. :)
Patrick Minton
December 26, 2007
Darrel,
No, that isn’t the difference. Did you miss the part where Dave referenced rebounds and turnovers?
And, by the way, your comment, which sarcastically implies that 8 missed threes are insignificant, isn’t even funny; 24 additional points would almost certainly mean at least two victories, maybe more.