Choiceology with Katy Milkman The Algorithm Advantage
33 snips
Mar 23, 2026 
Jennifer Logg, a Georgetown professor researching how people respond to algorithmic advice, and Dean Oliver, an ESPN data scientist who helped pioneer NBA analytics, join the conversation. They discuss the Celtics’ analytics-driven rebuild and how algorithms can outperform human judgment in prediction tasks. The show also explores when to trust machine guidance and the risks of biased or incomplete data.
AI Snips
Chapters
Books
Transcript
Episode notes
Celtics Rebuild Fueled By Analytics
- The 2007–08 Boston Celtics rebuilt by leaning on analytics instead of pure scouting intuition.
- Mike Zarin's predictive player model identified undervalued players like Kevin Garnett and Ray Allen, enabling the Big Three formation that led to a championship.
Analytics Reshaped NBA Strategy
- Algorithms changed NBA strategy by revealing undervalued plays like three-point emphasis and defensive switching.
- Teams shifted tactics leaguewide after analytics showed consistent long-term advantages that scouts' highlight-focused views missed.
People Prefer Algorithmic Advice Over Human Advice
- Jennifer Logg's research shows people incorporate more advice when it's labeled as coming from an algorithm than when identical advice is labeled human.
- In prediction tasks participants updated their initial answers more after algorithm-labeled advice, a pattern dubbed algorithm appreciation.
