NBA Team Wins and Losses attributed to the players who created them

The two month project is over.  I have produced, before the end of the postseason, NBA Team Win Charts attributing each team’s wins and losses during the 2009-10 season to the individual players who created them through their marginal statistical production, as measured by the metric known as Marginal Win Score.

click here for

THE 2009-10 NBA TEAM WIN  CHART PAGE

Before I get into the chart glossary, a few notes.  Using Marginal Win Score and its parent metric Win Score, I and Professor David Berri were the only two people on the planet who accurately calculated that the Milwaukee Bucks would win 40+ games last season when most had them pegged as a deep lottery team.  (NOTE:  Please don’t tell me Steve “The Homer” True predicted the Bucks would win “in the middle 40s”.  That’s called an opinion, and it was based on a reach behind himself and into his  anal cavity — the same as “George Webb predicts the Milwaukee Brewers will win 12 in a row“.  Nevertheless, we love the Homer, at least he approaches the sport with intelligent analysis.)

Second, I have now found 12 teams, including the Bucks, whose wins and losses created by players who were active the season before could have been predicted within 3% using each player’s Marginal Win Score from the season previous  (I am still calculating the other teams.  I also make no claims about rookies or European transfers.  All I can provide for those two categories of player are educated guesses.  I simply have found no reliable method of predicting those two types yet, though Europeans are easier than collegians.  I will explain this paragraph further in future posts).

Finally, in every NBA season in which I have calculated wins and losses using MWS48, the Association-Wide calculations have consistently been on a per team basis within 5.3% of pythagorean wins.

With that bit of cheerleading behind, below I provide detailed explanation about the information found in this season’s Win Charts.  The explanations are provided by column number, with the number reflecting the left-to-right appearance on each chart.  The players are listed on each team chart according to the number of Win Credits attributed to them in descending order:

GLOSSARY OF TERMS USED IN

THE 2009-10 NBA WIN  CHARTS

COLUMN 1:  “Position”

This season instead of posting that the player played “x” minutes at small forward and “y” minutes at shooting guard, I gave each player a single numerical position designation based on the traditional numerical postitional references (a “1” is a Point Guard, a “2” is a shooting guard, etc) but with the traditional references altered for players who split time at different postions with the alteration reflecting the percentage of time spent at the different positions.  (EX:  a player spends 70% of his time at point guard and 30% at shooting guard.  That player is a “1.3”.  The math is( .70*1 + .30*2).  The way I’ve been shorthanding it is this way: PG (1.0-1.5); SG (1.6 to 2.5); SF (2.6 to 3.5); PF (3.6 to 4.4); C (4.5 to 5.0)

COLUMN 2: “Games” or “Game Responsibilities“

This season instead of just repeating the player minutes as reported by Basketball-Reference I have instead translated those minutes into the player’s “Game Responsibilities” or the number of game outcomes the player is accountable for based on his floor minutes.  The math is done like this: Player minutes / (Total Team Minutes/82).  The logic behind charging player’s with game outcomes is simple.  In basketball each team has five players on the court at all times.  If those five played the entire game each player’s performance compared to his counterpart on the other team must account for 1/5th of the outcome.  Their can be no other logical method to divide responsibility.  Each player has is held responsible for 1/5th of the outcome for every minute he is on the court, and after about 241 minutes of floor time that 1/5th responsibility matures to a full game responsibility.

COLUMN 3: “Marginal Win Score / 48”

The chief player win production metric used by this blog.  I ask you to go here for a fuller explanation.

COLUMN 4: “Win__Loss Credits”

Quite simply, it is an expression of the number of wins a player created for his team during the season through his marginal statistical production.  And using the logic that for every minute a player is not creating a win he is creating a loss, “Loss Credits” are simply:  (“Loss Credits”= “Win Credits” – “Game Responsibilities”).

Column 5: “Win Contribution Index”

An expression of the overall “win value” a player delivered to his team based upon his marginal statistical production and the percentage of the team’s overall floor minutes that player used.

Column 6: “Wins Above Bench”

An expression of the player’s “absolute value” to the team by comparing the player’s production against the  Win_Loss production the team could have expected by employing easily acquired talent at the position. (See the Pages section for a fuller explanation).

Column 7: “Player Winning Percentage”

This stat used to be called “Player Win Average”.  Quite simply, PWP is the number of  wins a player created for his team for every “Game Responsibility” he accrued.  A player who accrued more than 3 game responsibilities and posted a winning percentage greater than 1.000% is a superstar.

More Win Chart Analysis throughout the summer

Now that I have completed this project so early, look for a series of posts throughout the summer analyzing the numbers, breaking down performance, and attempting  to project future performance.