18th May 2011
In gathering links for StatHead yesterday, I came across this post at the Wages of Wins, wherein Prof. Berri mentions that the losses of Allen Iverson, Carmelo Anthony, & Rudy Gay did not hurt their respective teams. He then writes:
"In each of these examples, the loss of a scorer led people to forecast doom. In each case, the team losing the scorer managed to survive and even improve.
Readers of The Wages of Wins and Stumbling on Wins understand this basic story. Scoring is overvalued by many NBA observers. Top scorers do not always have the impact on wins that people imagine. But no matter how often this story repeats, each time a scorer is lost we still see the same arguments offered by adherents to the conventional wisdom (for example, this week the Grizzlies insisted they would never dream of letting Gay depart)."
That's anecdotal evidence, though. What if we looked at every instance of a team losing its leading scorer? Would the typical team in that situation be impervious, or are those just a few cherry-picked exceptions to a larger rule?
Well, luckily, at BBR we have boxscores for every regular-season game since 1985-86. So I gathered our data, considering a team's "leading scorer" to be the player who led the team in PPG among players who played more than half of the team's games. I then looked at each team's offensive rating in every game, noting whether the designated "leading scorer" played in that game or not.
I also accounted for the strength of the opposing defense in each game by measuring how many pts/possession the opponent allowed in every game of the season except the one at hand. The end result will measure how well each offense performed relative to what we would expect from a league-average team facing the same opponent -- split by whether the team's "leading scorer" played or not.