One-Legged 3's, The Perfect Defenders & The Knicks One Trick Pony

In partnership with

Welcome to the NBA Librarian Weekly, where we curate and summarize the best NBA content of the week.

<1000 words each week. I consume so you don’t have to.

In Today's Edition:

1. The Beginning of the One-Legged Three Era

2. Analyzing the ‘Perfect Defenders’ for Each Player

3. Westbrook: The Problem or the Solution?

4. Sengun’s Defensive Impact

5. Celtics Need to Improve their DHO Defense

6. Are the Knicks a One-Trick Pony?

The Beginning of the One-Legged Three Era

In 2019, James Harden practicing the one-legged three sparked a ton of Reddit discussion, especially when he tried it in-game in 2020. Initially mocked as no more effective than his step-back, Harden quickly abandoned it. Last week, I highlighted a ‘Thinking Basketball’ piece on how players are inventing new shots to counter evolving defenses—enter the one-legged three-pointer.

We’re seeing these attempted more often, as this graphic shows. LaMelo Ball is absurd, hitting 21 of 53 one-legged threes—a better percentage than his regular threes. My guess? In a league hyper-focused on defending the three, these shots help players create more airspace, marking just the beginning of the one-legged three evolution.

Analyzing the ‘Perfect Defenders’ for Each Player

Defensive matchups in basketball often rely on tradition, like pairing players by height or general defensive ability. Ryan Miele’s project attempts to rethink this approach, identifying the ideal physical traits (like height and wingspan) needed to guard any offensive player and matching them to real NBA defenders.

How It Works:

• Matchup Impact Score: Combines key stats (e.g., FG%, fouls, blocks, turnovers) into a single metric to measure defensive effectiveness. Lower scores = better defense.

• Data Integration: Merged 5 years of NBA matchup data with height and wingspan stats from the current season to analyze defender impact.

• Machine Learning Predictions: Used a model to predict the best defender traits for minimizing offensive efficiency based on shot profiles and physical attributes. And matched them to actual NBA players today.

• Bayesian Optimization: Calculated the "perfect defender" for any player, like:

  • Trae Young: Best guarded by smaller, long-armed players.

  • Kevin Durant: Most effective defenders are slightly shorter with long wingspans.

In short, Ryan’s model uses data to take the guesswork out of defense, offering teams smarter ways to slow down opponents.