Intro to Defensive Charting

Intro to Defensive Charting

Defensive charting is a concept formalized by basketball analytics pioneer Dean Oliver in his seminal work, Basketball on Paper. More recently, it’s been used by basketball writers/analysts like Luke Winn at Sports Illustrated and David Hess at TeamRankings.com. As Hess explains here, defensive charting consists of rigorous review of a game tape (the 8-second rewind button is a charter’s best friend!) to allocate credit and blame to individual defenders.

Compared to traditional methods like boxscore stats (steals and blocks) and the “eye test,” defensive charting is a vastly superior way to assess individual defenders. That said, it’s not a perfect method. A couple notable shortcomings are:

  1. It’s (generally, and in this case) done independently of the coaching staff. Without perfect knowledge of players’ responsibilities during a given play, it’s possible that blame can be inaccurately assigned. While I’m broadly familiar with Carolina’s defensive schemes (like how the Heels fan penetration to the baseline to set up the help defense/rotations, or how Roy Williams prefers a hedge-and-recover style of defense against ball screens), there are certainly nuances/in-game adjustments that I’m not aware of.
  2. There’s an element of subjectivity involved. Two knowledgeable observers could watch the same play and disagree about which defender (or defenders– credit/blame is often split in this system) to assign the result to.

Despite those limitations, the insights derived from defensive charting data are far more illuminating than those that can be gleaned in a non-charting environment.

Hess’s article (linked above), borrowing from Basketball on Paper, does a great job of explaining the metrics used for defensive charting. The primary ones are Stop% (the fraction of a player’s total individual defensive possessions in which he’s contributing to a stop) and %Possession (a defensive usage rate). These combine to estimate an individual Defensive Rating for a player that is anchored by the team’s DRtg. While my presentation of the boxscore differs a bit from Hess’s, the key concepts are the same. In addition to DRtg, I’ve also included On-Court DRtg in my tables. This is the team’s actual defensive efficiency during a given player’s minutes (rather than the one estimated from the charting metrics)

Below you’ll find three tables:

  1. The defensive charting boxscore for UNC’s latest game (@ Hawaii).
  2. The season-to-date boxscore for UNC with cumulative totals.
  3. The season-to-date boxscore for UNC with summary statistics and per-40 numbers.

As the season progresses, I’ll probably post frequent single-game charting boxscores along with some discussion of interesting defensive insights (like how Carolina switched more ball screens against Hawaii, and with pretty poor results). I’ll also occasionally do an updated post that includes season-to-date charting numbers. Finally, for newsletter subscribers, I plan to make available via a weekly email the game-by-game defensive charting data (along with the lineup combination (+/-) data).

Let me know if there are any questions about what’s going on with these metrics (or general questions about how credit/blame is assigned during the defensive charting process, etc.). Sorry for the pure data dump below; I promise that the next charting-related post will include more actual discussion about Carolina’s defense/individual defenders.

screen-shot-2016-11-20-at-3-20-39-pm

screen-shot-2016-11-20-at-10-05-00-pm

screen-shot-2016-11-20-at-10-05-26-pm

Leave a Reply

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