A recent study conducted by researchers at the University of North Carolina at Chapel Hill and Northeastern University has revealed that leading AI models, including ChatGPT and Google's Gemini, are significantly inadequate at analyzing professional sports. The research utilized a new testing framework called strategic video intelligence, or "SVI-bench," which involved 35,000 hours of sports footage and millions of annotated plays and reports. While AI demonstrated a 74% success rate in recognizing player actions, it faltered in causal reasoning with an average accuracy of around 40%. For example, when asked about an unusual basketball play, ChatGPT failed to provide a relevant explanation. In predictive tasks, AI struggled even more, performing as poorly as random guessing. The models' accuracy plummeted to just 5% when tasked with comprehensive post-game analysis, underscoring the gap between AI capabilities and the nuanced understanding required from human sportscasters. This study not only reassures sports commentators but also highlights broader implications for various professions that rely on understanding and interpreting complex events.