Welcome
Issue One Hundred And Seventy Six.
This week's collection covers key topics in sports analytics, including the impact of swing speed aging in baseball, team chemistry in soccer, and career rebounding leaders in the NBA. Each piece highlights how specific data points shape performance and decision-making in their respective sports.
This Week's Lineup
Corbin Carroll Is Even Better Than Advertised
Not all outs in baseball are equal; players like Corbin Carroll, who avoid double plays and advance runners, provide more value than typical outs, while players like Aaron Judge, who hit into frequent double plays, cost their teams more runs, highlighting how different types of outs impact run expectancy.
Chemistry 101 — American Soccer Analysis
A model analyzing soccer team chemistry through cumulative pass data suggests that higher player connection correlates with increased expected goals for and reduced expected goals against, providing insight for coaches on team cohesion, squad management, and youth integration.
NBA Career Rebounding Leaders
A fun look at the NBA's all-time career rebound leaders include Wilt Chamberlain (23,924), Bill Russell (21,620), and Kareem Abdul-Jabbar (17,440), showcasing dominance in rebounding through their size, athleticism, and sustained success across long careers.
Swing Speed Aging Curve
The aging curve for swing speed in baseball shows a slight decline from ages 22 to 31, averaging a drop of 0.02 mph per year. After 31, the decline accelerates to 0.31 mph per year. Notably, a significant outlier, Jordan Walker, had a 2.4 mph swing speed increase between ages 21 and 22, influencing the overall average for that age group.
Towards Universal Soccer Video Understanding
The paper introduces SoccerReplay-1988, the largest multi-modal soccer dataset, containing videos and annotations from 1,988 matches, with an automated annotation pipeline. It also presents MatchVision, the first visual-language foundation model in soccer, designed to leverage spatiotemporal information across videos. Extensive experiments show that MatchVision outperforms existing models in event classification, commentary generation, and multi-view foul recognition, setting a new standard for sports understanding research.
Please Forward
Was this email forwarded to you? Sign up here >>
Unexpected Points Added is curated and maintained by Patrick Hayes.
I completed a Master of Applied Data Science from the University of Michigan; let's connect.
Send me an email with your feedback or questions. I respond to every single one.