February 13, 2010

A Steal for the Mavs.

Filed under: Uncategorized — wwinston @ 9:26 pm

It looks like the Mavs are going to trade Josh Howard, Drew Gooden, James Singleton and Quinton Ross for Caron Butler, Brendon, Haywood, and DeShaun Stevenson . This is a great trade for the Mavs. The major players are Howard, Gooden, Butler and Haywood, so let’s focus on them.

 Let’s first look at the Mavs big weakness: poor defense in the backcourt. Here are the Adjusted +/- Defensive Ratings for the Mavs Guards:

Terry +8, Kidd +8, Barea +7, Beaubois +14.

This means, for example, if Beaubois plays 48 minutes in lieu of an average NBA defensive player we believe the Mavs would yield 14 more points per 48 minutes.

These are the worst defensive numbers for any backcourt in the league except for the Wizards:

Arenas +13, Foye +10, Young +4, Stevenson +8, Boykins +9.

So what do you do to mitigate the weakness of a bad defensive backcourt? Bring in a great defensive center who can cover up the guards’  mistakes. Enter Brendon Haywood whose -16 Adjusted +/- Defense Rating is the league’s best. Haywood is weak offensively ( -9 Adjusted +/- Offense Rating), but he should be able to cover for many defensive lapses by the Mavs guards. If he does not work out, the Mavs do not have to resign him.

   Caron Butler has a +3 Adjusted +/- Points Rating this year.  Butler has a +2 Points Offense Rating and -1 Points defense Rating, so he is not a bad defender. Butler’s contract does run through 2011, so the Mavs do need him to pan out.

 Butler’s current numbers are similar to Josh Howard’s past ratings. If Josh can regain his pre-injury form, the Butler for Howard portion of the trade is probably a wash.   So unless Howard recaptures his old form, this part of the swap is great for the Mavs.

    We also have an Impact Rating, that is based on how player’s change the chance of their team winning,. For example, if you are down by 20 points with 2 minutes to go and your team loses the game by only 5 your Adjusted +/- rating shoots up, but your Impact ratings is  hardly affected.  Here are how the four players rank in Impact Rating among the 200 NBA players who have averaged at least 25 minutes per game.

Butler 43rd, Haywood 59th, Howard 83rd, Gooden 190th.

Again, by this metric the trade is a steal for the Mavs

Drew Gooden’s Adjusted +/- rating this season  is -4 and the past 3 years he has averaged an Adjusted +/- of -6, so he will probably not help the Wizards much. Certainly the Haywood for Gooden part of the deal is a huge plus for the Mavs. 

   Of course, when players are traded, they often perform better or worse than expected. But as we said, the Mavs need to cover for their poor defensive guard play, and Haywood has done that with the Wizard’s poor defensive backcourt, so there is every reason to expect that his great defense will be lead to a revitalized Mavs defense.

  We leave you with one final statistic:

When Butler and Haywood are both in this year for the Wizards, the Wizards play 2 points better than average. The rest of the season the Wizards play 13 points worse than average. So together Butler and Haywood can turn a poor team into an above average team. 


  1. Looks like a very good trade.

    On the technical front though, there is something that I have begun to pick up on. Haywood, Jamison and Blatche are all positive on basketballvalue’s Adjusted +/-. They combine for 88% of the total PF & C minutes. Meanwhile the most guard-like players are all negative. There are other examples where this pattern occurs- really positive across the front court, very weak in the backcourt, maybe more than the eye would expect- (e.g. the last championship Pistons and this season’s Nuggets) and I’d tend to think the answers that Adjusted provides- with the limited amount of time played with alternatives at PF & C- or at least their size might need to be treated with some caution. Any further thoughts about why this arises and how to deal with it?

    Comment by Crow — February 14, 2010 @ 1:43 am

  2. Very insightful.

    Could you explain the “Impact Rating” calculation a bit more? I would to add this metric to my website.

    Comment by Serhat — February 14, 2010 @ 6:52 am

  3. Each team starts with 50% chance of winning game. Suppose my time outscores you by 10 points in 1st quarter and I am in. Then our chance of of winning may be say 80%. I now get credit for 30 Impact Points while adjusted +/- gives credit for 10 basketball points. In short, the metrc is change in chance of winning game instead of change in game margin.

    Comment by wwinston — February 14, 2010 @ 8:51 am

  4. Good point, as usual. I think this is because when a PF is out one more of the bad defensive guards is in,. I do not think this should mess up things. The team does better when these guys play because they are better players than the other guys, and that is what the numbers say.

    Comment by wwinston — February 14, 2010 @ 8:52 am

  5. Thanks. The 04-05 Pistons and this season’s Nuggets and Wizards all have very bad big men playing those limited minutes not played by the top 3 big guys. There is the argument that the data is over-fitted based on the comparison of the results with the main guys with the bad team results in these quite limited minutes. At quick glance Denver and Washington have used some 3 guard lineups and that might a factor too.

    Comment by Crow — February 14, 2010 @ 2:11 pm

  6. fyi, gooden is younger the haywood

    Comment by observer — February 14, 2010 @ 3:01 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment

Powered by WordPress