Sports & Media · A major professional sports league
Schedule optimization and ratings prediction
+8%
TV ratings above historical average
- Challenge
- A legacy scheduling system was slow to solve and produced schedules that left value on the table.
- What we built
- Ratings prediction models trained on preseason and historical viewership data, and an optimization algorithm purpose-built for the league’s scheduling problem, delivered in an interface the staff run themselves.
- Result
- TV ratings 8% above historical average
- Schedule penalty score reduced 50%
- Ratings predictions 93%+ accurate