Below is the new, simplified Win Chart format I created for the 2010-11 Orlando Magic. In order to distinguish Marginal Win Score, and to make the sources of Marginal Win Score clearer, I have omitted the former “Win Contribution” column (which I don’t think many people understood anyway), and have replaced it with columns listing the (1) Player’s Win Score average and (2) his Opponent’s Win Score average.
The reason I did so was to make it clear that the “MWS” number for each player is simply the difference between the WS that player produced and the WS that player’s collective “Opponents” have produced, divided by two. The thinking behind the metric is that half of a basketball player’s impact comes from what he himself produces in the box score for his team(WS), and half of a basketball player’s impact comes from what he prevents his opponent from producing in the box score for his team (OppWS). Half from the offensive end and half from the defensive end, to put it in loose terms.
From those MWS averages I can then calculate each player’s expected winning percentage. From each player’s expected winning percentage I can then calculate the number of wins and losses each player has produced based on his winning percentage and the number of minutes the player has played (or his “games played”). Remember, the number of “games played” by each player is simply the player’s minutes played divided by the total number of team player minutes per game.
BTW, I also did a calculation column (which did not fit so I had to eliminate it) that recalculated each player’s wins and losses according to the MWS the player posted last season. If you go by that calculation, the Orlando Magic would have been expected to win 22.3 games at this point… virtually identical to the number they have won.
I bring that up to make this point. The Orlando Magic’s “disappointing” start was not due to the poor play of Vince Carter or Rashard Lewis. Carter was producing at the same level as last season, and Lewis has been well below average for several seasons. The difference was the loss of SF Matt Barnes to the Los Angeles Lakers, something everyone seems to omit from their memories.
Similarly, the Magic’s improved play is not due to Gilbert Arenas or Jason Richardson (though it is due in small part to Turkoglu). The main reason the Magic have improved so much is the substitution effect created by losing the below average Rashard Lewis and replacing him with the more productive power forwards Brandon Bass and Ryan Anderson.
Also, while I believe that Win Score (as well as Adjusted +/-) has shown a decline in the play of Orlando center Dwight Howard, Marginal Win Score has not. According to MWS, Howard is producing at exactly the same level as last season. (the Win Score analysis I am citing could be stale).