Skip to content
WTFSports

Isolation basketball is quietly making a comeback, and the analytics crowd hasn't caught up

DBDeShawn Brooks· posted 3w

For most of the last decade the analytics consensus was clear: isolation basketball is inefficient, ball movement and threes win, stop letting one guy dribble the shot clock down and go get his own bucket. Coaches built entire systems around that finding, and it genuinely changed how the league played. But watch the league's best teams in the clutch right now, and isolation is quietly back, not as the whole offense, but as the closing weapon nobody's willing to talk themselves out of. Here's why the models undersold it. Isolation efficiency looks bad in the aggregate because most isolation possessions in the league are run by guys who shouldn't be running them, mediocre shot creators forcing bad shots. But in the hands of an actual elite one-on-one scorer, isolation efficiency in high-leverage moments is genuinely excellent, because it takes the play call, the pass, the extra read, all of it, out of the equation and just puts the ball in the hands of the guy most likely to make something happen when the defense knows exactly what's coming. The league didn't actually solve isolation basketball, it just correctly identified that most isolation basketball was being run badly. Elite closers doing it well were never really the problem the stats were describing. Bleacher Report's defense of the isolation play makes the same case, that the tool's fine, it's mostly the execution that's been bad.

2 1

Replies2

Football's had this exact argument with the individual dribbler versus the possession-based system. Both work when you've actually got the right player for it, and both look terrible when a team tries to force a system onto personnel that doesn't fit it. This isn't a basketball-specific insight, it's just true of any sport with a tactical fashion cycle.

The part I'd push back on is that isolation still empties out the other four guys on the floor for five to seven seconds, and that has a real cost even when the possession itself works out. You're not just measuring points per possession on the iso play, you're giving up the compounding value of everyone else staying involved and threatening something. The efficiency number hides that.