IBM Scores Tennis Serves With 21 Data Points. Not Just Speed. Obviously.
IBM's updated US Open app, part of a 35-year partnership, delivers personalized homepage updates without requiring navigation. The standout feature is a 'serve quality' metric that breaks down a serve into 21 different data points, analyzing limb and racket movements to generate a comprehensive score beyond traditional speed measurements. Jonathan Adashek explained the app surfaces relevant information immediately for each user.
This demonstrates dimensional decomposition. The mechanism is feature extraction: instead of measuring one obvious variable like serve speed, the system analyzes 21 underlying variables to characterize quality. The principle generalizes far beyond tennis. Any complex performance, when broken into sufficient components, becomes legible to both machine analysis and human improvement. Single metrics mislead. Multi-dimensional scoring reveals.
IBM built the app in partnership with the US Open, now in its 35th year. Jonathan Adashek described the personalized homepage and serve quality analysis features.
- Download the official US Open app during the tournament. Expected outcome: access to IBM's AI-powered features including the serve quality metric.
- Watch any match and locate the serve quality score for a player. Expected outcome: you see a single number that actually summarizes 21 underlying biomechanical data points.
- Compare two players' serve quality scores side by side during a match. Expected outcome: you notice that higher serve speed does not always mean higher serve quality, which is the entire point of dimensional analysis.