Defense wins championships.
You'll hear it time and time again, particularly with the NBA just three weeks away from the start of the 2022 playoffs.
And while the cliché does hold somewhat true -- every NBA champion since 2000 has ranked 11th or better in defensive efficiency -- we still struggle to measure what makes a good defense. Points per game allowed has given way to points per 100 possessions (the pace-adjusted measure we use to rank teams by defensive efficiency), but we still rely heavily on basic measures like blocks and steals to identify elite individual NBA defenders.
However, a decade into the NBA's player-tracking revolution, that's slowly changing. Both Second Spectrum and the league itself have introduced an ensemble of important new defensive metrics.
The league's internal effort is making rapid progress by combining player tracking data with sophisticated algorithms that codify the principles of NBA defense (e.g. denying passing lanes or staying between the offensive player and the basket) and calculate a variety of measures using AI and machine-learning techniques to extract meaningful defensive insights and stats, including identifying offense-defense player-matchup responsibilities throughout every possession.
Using these new metrics gives us a clearer picture of which players should make up the NBA's All-Defensive teams, some of whom might come as a surprise to the average NBA fan.
