This past summer (and parts of the fall), reviews of almost every NBA team – using Wins Produced – were posted in this forum (these are all listed at the NBA Analysis Page). The lone exception was the New Jersey Nets. So with this quick post, every team will have been reviewed.
Let me start with what the Nets did last year. Yes, they were the worst team in the NBA. Last February we explored how this happened. Here were the basic points made eight months ago.
- Obviously the roster assembled in New Jersey was quite bad. But it shouldn’t been as bad as it appeared.
- The problem in New Jersey was primarily at the point guard spot. The three players who declined the most were Devin Harris, Rafer Alston, and Keyon Dooling.
- The departure of Alston – and the addition of Kris Humphries – probably helped the Nets avoid the worst record in NBA history.
When we look at the final results for this team (32 games after the story in February was posted), we find support for these stories. The following table indicates the team should not have been as bad as it was in 2009-10. The point guards were the players who decline the most. And the addition of Humphries and loss of Alston probably helped.
Although all of that may be interesting, much of this doesn’t matter in 2010-11. This team now has new ownership. And more importantly, many new players. In fact, of the ten players listed on the depth chart, six were not with the team last year. Because clothes do not win or lose games, we can expect a change in the players in the clothes will cause a change in the outcomes observed.
How much of a change, though, depends on the quality of players added and lost. Let’s focus on the latter first. The above table reveals that the Nets employed five players who a) saw 500 or more minutes of actions and b) posted WP48 [Wins Produced per 48 minutes] marks in the negative range. All five of these players are now gone (and if you now have one of those players on your team, that might not be a good sign). Having these players depart clearly helps.
The team has also added a few players who produce in the positive range. This list begins with Troy Murphy, who posted a 0.294 WP48 last season with the Indiana Pacers. Murphy, though, begins the season injured. Because Murphy’s back-up is Derrick Favors – and Favors played poorly in the preseason – this particular injury could really hurt.
Whether or not Murphy plays, though, the Nets should improve on their 12 wins from last season. This is because the team is primarily employing players who produce in the positive range. The lone exceptions are Quinton Ross (who might be the back-up shooting guard) and the aforementioned Favors (although projecting rookies is difficult).
How much Murphy ultimately plays is not something we can project easily. But we can offer a projection if he does play. At least Arturo Galletti can do this. Arturo has offered projections of every team, and according to his extensive analysis (and it really was quite extensive) the Nets will win 39 games in 2010-11. In fact, Arturo has the Nets making the playoffs. Wow!!
Again, I think Arturo expects Murphy to play and be productive. In fact, Arturo also expects Favors to contribute (despite his preseason numbers). So this might be an optimistic projection.
By the way, if you don’t like this projection – or even if you just want to see more and more and more projections – one can look at the Wages of Wins Network Stat Smackdown. There you will see that even if people agree that Wins Produced is a fairly good idea, projections of the future will still differ.
One last note… Andres Alvarez says the automated Wins Produced numbers should be available throughout the season. He thinks updates each week are possible (with the first numbers coming out in a week or so). So it looks like this season we will be able to evaluate all players via Wins Produced throughout the year. Once again, thanks to Andres for making this possible.
– DJ
Wages of Wins Journal 
ironlynxtk
October 27, 2010
A small suggestion, there should be an update to automate the Wins Produced numbers website by Mr. Alvarez to pull statistics from a public database that is also up to date. (I figure websites like espn wouldn’t like that you are taking statistics from their website for your use, unless you can get their permission.)
That way you won’t have to update manually, you can set it up update every night at 3 am (for example), pull all the statistics for each player who played that day, run the calculations, and you can have updated Wins Produced every morning!
nerdnumbers
October 27, 2010
Iron,
Thanks for the suggestion. Without revealing too much from behind the curtain I’ll reveal that updating is not the hard part. Some of the issue is that this is the start of the season which means things like position averages/team rebounds aren’t stable yet. I have to update some code to handle that. I built this in the offseason/playoffs when the data was all stable. I want to wait until I’ve run this for a few weeks before committing to anything more than that.
A.S.
October 27, 2010
One thing that Prof Berri touches on from time to time is the variability of players’ production from year to year. Prof. Berri has often commented that NBA player’s production is fairly consistent over time. In general, I am very intersted in the degree and causes of variation over time (as I recall, there was some discussion in Wages of Wins?) – e.g., young players get better, but (in general) how much and over what period? In particular for this post, I am wondering whether the wholesale change of player personnel on the team affects player production? We saw last night with the Heat, for example, that players’ production appeared to be affected a lot by the fact that the team had not “gelled” yet. There were lots of turnovers and bad/missed shots. This was caused, I suspect, by the fact that the team consists of a lot of players who are playing together for the first time. Same thing will apply in NJ – a lot of new players playing together for the first time. What is the likely effect of that on their Wins Produced?
Matt
October 27, 2010
So now that each team has been projected based on a seemingly strong calculation, could one of you post a summary article? That is, given the basic (and obviously not perfect) forecast for each team, what would the standings be? Do the total wins forecasts make sense? Do they need to be scaled to a limit of 1,230 league wins?
I think that could go a long way and shouldn’t be overly difficult. It would also be incredibly fun to reference throughout the season and could be used as a basis for gloating, or the inspiration for an article on why certain teams over- or under-performed your expectations.
nerdnumbers
October 27, 2010
Matt,
http://www.nerdnumbers.com/statsmackdown (general summary)
http://nerdnumbers.wordpress.com/2010/10/27/wages-of-wins-network-2011-stat-smackdown-eastern-conference/ (general summary of Eastern Conference)
You’ve read my mind and I intend to update periodically throughout the season. I want to give a huge shout out to all the Wages of Wins Network analysts who put in all the work in the span of a week to do this!
Matt
October 27, 2010
Perfect, thanks.
entityabyss
October 27, 2010
Hey A.S., consider this. The boston celtics, when the big 3 came together for the first time, just started off dominating. The whole thing with jelling, while possible and plausible, is not necessarily true. Dwyane wade just came off an injury, mike miller is out for 25 games, and the game was at TD garden. It was boston’s game to win.
Don’t let the whole jelling thing confuse you. Miami will dominate. Players just had bad games against a defensively strong team. Also, any team can lose to any team. The end result is the concern.
arturogalletti
October 28, 2010
AS,
http://arturogalletti.wordpress.com/2010/07/19/measuring-the-quality-of-basketball-in-the-nba-part2-adjusting-for-pace/
You’re welcome :-)