Defensive Box Scores: Lakers-WolvesGame charting by Patrick McCarthy
Patrick has charted all kinds of details for every play of the Lakers-Wolves NBA 2003-04 Western Conference Finals series and assigned defensive responsibility for field goal attempts, free throws, and turnovers. Admittedly many instances require subjective evaluation, and frequently the credit/blame is split between two players or even allocated to the "Team" as a whole.
Explanations: (all stats reflect assigned responsibility to a player).
Please note the RTG is not currently adjusted by the quality of the opposing player guarded, although this is likely to happen in the future!
Lakers player defensive performance
82games series defensive MVP: Karl Malone!
Timberwolves player defensive performance
For the Timberwolves, Hassell played solid defense, but Johnson's good RTG is deceptive -- it came out because Shaq missed so many free throws awarded on Johnson's ticket! Szczerbiak's numbers look okay (40% FG allowed) until you realize he gave up 11 three-pointers!
-- How does the full charting compare to the "Counterpart Defense" statistics?Since we've been publishing individual defensive stats derived from the matchups within a game (e.g. a player considered to be at PF at a given point of a game will be held accountable for the performance of the opposing PF in the game at that same time -- see for example Tim Duncan's counterpart production) it's an important point to understand: how reliable are the counterpart numbers in contrast to real charting?
Here are the counterpart stats for the same series, with the Charted RTG ('cRTG') for comparison purposes:
Lakers player COUNTERPART statistics
Timberwolves player COUNTERPART performance
Analysis: the majority of players have similar results for the RTG ratings on the "auto-defense" -- particularly those with substantial defensive possessions to count, but there are several notable exceptions, including Szczerbiak (75 counterpart versus 100 charted), Madsen (86 vs 103), Hassell (95 vs 74) and Sprewell (102 vs 88).
The big differences are likely to come mainly from one type of situation: when an opposing player drives past a perimeter defender to encounter the big men underneath. With the counterpart defense we charge the outcome of that possession (barring a block or foul) to the perimeter defender, whereas with real charting Patrick is often giving 1/2 or even full responsibility on the play to the interior defender(s).
Consequently real charting shows less shots against PG, SG, SF types and more shots against PF, C types. Now many people who have played the game of basketball will argue that this is the right way to do it (Dean Oliver in the excellent Basketball on Paper elaborates on charting individual defense at length).
Still, this leads to an imbalance perhaps in defensive ratings if the interior defenders are facing higher degree of difficulty in stopping a player on help defense versus their own one-on-one play. Likewise the perimeter players may be benefitting from not being charged full value when getting beaten off the dribble.
Unfortunately without a much greater supply of charted games to compare, we are hard pressed to get a good sense of the full effects. Of course in terms of our package of opposing defense numbers, we also like to consider rebounds allowed and per minute averages in creating the PER based ratings.
For now we'll rest with the belief that actual charting is preferable, but the counterpart defensive stats appear to hold their own reasonably well, with presumably better strength over the course of a full season than in a short series.
Copyright © 2004 by 82games.com, All Rights Reserved