Most Important Players for the Season so Far

Image Courtesy of xkcd.com

Game Theory is the study of games and strategy. It’s focus is all about the choices the players and how that affect the probability of success. A game has players, a set of moves (or strategies) available to those players, each with a specific value or payoffs. As a player you want to undertake moves that optimize your expected value. For the NBA this very much applies. Certain players have a significantly higher impact on the expected outcome of a game. I was inspired by the above xkcd comic to write a post on just those players.

Some quick background. Start at the Basics. For this post I’m judging player importance by their net impact in terms of Wins Produced and Point Margin per Game.I’m also looking for net contribution over the course of the season so far. Since I know each players Wins per 48 minutes from the Automated Wins Produced Site. I can use that number to work out Point margin generated for the team as Follows:

Points  Margin Generated for Team per 48 = 31*(WP48-.099)

Points  Margin Generated for Team=Points  Margin Generated for Team per 48*(PlayerMinutes)*5/Team Minutes

Based on these equations, there are fifty players in the NBA currently that have impacted the bottom line for their by more than 1.5 net point  per game. They are :

Paul and Love are in a class by themselves by delivering 7 pts a game for their teams.  Garnett, Gasol, Horford and James all deliver 5 points (and Garnett.,Gasol and James all do it on superlative teams which as per diminishing returns is harder).

What’s really interesting is that this list has a dark parallel. A mirror if you would.

Mirror Universe Smiley is not amused

There are 52 NBA players that have reduced their teams bottom line by 1.5 points a game or more for the season year to date.


Remember kids, if you’re worse than a .100 WP48 player you’re on a verage costing your team points. The “leader” of this group is Andrea Bargniani who costs his team on average 4 point a game. Landry,Hickson Lopez and Blatche all hover around -3 points a game.

One thing that really pops out from this exercise? If I break this down by team:

The top tier in the league so far comes into very clear relief. There are five teams with at least three players delivering 1.5 points a game or more to their team(let’s call them earners)  (Spurs, Heat, Celts, Lakers and Mavs).  The Celtics are in a class by themselves with five such player and only one -1.5 point producers (call these wasters) (a rookie center Erden who should not see time in the playoffs). The Lakers have 4 +1.5 earners but their waster will get playing time in the playoffs (three guesses as to who and the first two don’t count). Miami and the Spurs feature unsurprising trios as their three earners and both have additional player who could make it a quartet (Blair for the Spurs and Miller for the Heat). As for the Mavs, we have Kidd,Chandler and Dirk as earners and Barea as a waster but once Rodrigue is back, J.J. role should diminish.

 

As for the T-Wolves and Kevin Love? Fear not, Amnesty International has been contacted.

And that dear readers is that.

 

19 Comments

  1. some dude
    12/16/2010
    Reply

    no mention of the lakers 5th earner coming back? I’m still hoping Fisher’s minutes are reduced even in the playoffs.

    I didn’t realize how bad Carl Landry has been. I wonder why he was so good on Houston and so terrible on Sacramento. Hickson got a lot worse, too.

    Is it just me or are a lot of players nowhere near last year’s WP48?

    Now, just to play Devil’s advocate here. This means that is Minnesota swapped Love for Gerald Wallace, Minnesota would have a margin of -15? I have a hard time believing this…

  2. 12/16/2010
    Reply

    some dude, I imagine if Gerald Wallace had to play center (or even lots of minutes at PF) the Timberwolves would be in even worse trouble than they are now. For what it’s worth, Hollinger thinks it’s a 14 game swing: http://games.espn.go.com/nba/tradeMachine?tradeId=29jfmbd (granted I have no idea how he does the rotation adjustments and other stuff it claims at the bottom).

    • 12/16/2010
      Reply

      Not quite that bad. You have to figure that Wallace (or pretty much anyone’s numbers) will go up slightly from going to the Twolves (diminishing returns exist in the negative sense as well) and some of KLove’s minutes would go to Tolliver who’s not horrible. But they would also go to Darko and Beasley. Ick.

      That might actually be dividing by zero.

      If Khan trades Klove straight up for Bargniani that might cause the universe to collapse.

      • some dude
        12/16/2010
        Reply

        I agree that would make the the worst team in the league, easily. But I’d also be willing to bet the house that the Wolves wouldn’t finish -20 point margin average for the season, which is what the chart says. heck, I’d even be willing to concede potential worst point margin all-time possibility, but -20 is just too much.

        • 12/16/2010
          Reply

          SD,
          And I agree with you. The math is a little more worked. Take Love out and replace him with Jefferson and everyone else on the T-Wolves would improve slightly (although not even close to enough to compensate). It is incrementally harder at the edges to move away from the mean and that’s what we would see on this kind of team. So the saying “stats on bad teams” has some real meaning. I would think they’d hover around the worst ever mark and not lower.

          • some dude
            12/17/2010
            Reply

            okay cool.

            I guess i was wondering if there is a way to account for those edges. I’m thinking the equations are linear except at the edges, which is why this happens.

            • 12/17/2010
              Reply

              There is and I’m working on something around that. Think “Chemistry” and roles.

    • some dude
      12/16/2010
      Reply

      Must be because Hollinger sees Wallace as just a SF and Love as a PF in the automated stuff. That said, I’ve never liked his trade machine analysis, though love the ability to concoct trades properly.

      I didn’t realize Love played some C (according to 82games, about about 20% of his minutes are there). That’s due to me avoiding Wolves games and for good reason :P

      How about this one? Al Jefferson. He’s at .110 WP48 right now and .130 last year, so a +1. (they were -9 at a team last year). So we’re looking at a -14 team. Still seems far-fetched.

      I think they record is somewhere like -11 for a team. -14 is an astounding number. Could this be the curious case of players at the extremes (playing on a horrible team)?

  3. HU
    12/16/2010
    Reply

    Arturo,
    Does WOW analysis take into account opp. shooting %? As in, how would WOW analysis conclude on a player who shoots 60% with 15 board per 36m but gives up a 100% shooting percentage and 50 PP36m from his opp. counterpart (and just to prove that he is not helping teammates’ defense, he only hangs out in the attacking half of the court)?
    You probably have already covered this and I missed it but just wanted to ask. I noticed opponents shoot a really low AdjFG% versus some players (Rondo, Lebron, Camby, Shaq, etc) and terrific versus others (all of the Wizards – I assume bad coaching, Bargnani, Beasley, etc)
    Thanks,

    • greyline
      12/17/2010
      Reply

      Check out Courtside Analyst for an attempt to account for opponent production. Accounting for defensive contributions beyond rebounding, blocks and steals is an issue with every metric I’ve seen.

      • 12/17/2010
        Reply

        HU,
        Yeah, WP approximates individual defense thru a team defense and the individual stats in the boxscore. This is a good but not great approximation. Great defenders (Artest,Perk,Battier) do tend to be undervalued. We’re currently working on game by game data automation (and this might let us take it to position based defense) but we have a data source for play by play (which would let us do individual defense).

        As greyline says, Ty over at Courtside is doing this by hand (i.e. he figures out the opponent WP to account for individual D).

        • some dude
          12/17/2010
          Reply

          I read that charges drawn and taken are soon to be an official stat? This would help if included into the WP48 formula as a drawn charge is not recorded as a steal.

          There’s obviously much more missing, but it could be something!

  4. Man of Steele
    12/17/2010
    Reply

    HU, I’ll butt in and answer your question with a no. This is the single greatest weakness of the WoW metric – its inability to account for defensive efficiency on the individual level.

    • HU
      12/17/2010
      Reply

      If this is the case then how does opp. counterpart shooting %, if added to the WOW metric, alter results?
      This is not a perfect statistic but it is certainly much better than nothing as people work on a better defense measure, if others don’t exist.
      Can somebody check this? It would certainly add to basketball analysis.
      Thanks,

Leave a Reply

Your email address will not be published. Required fields are marked *