It’s March Madness and lots of player evaluations will be done. At the same time, it’s the beginning of the off season for high school and many college coaches. In this issue we explore a few AI powered player evaluation models and AI driven strategies for coaches and players.

In this issue:

Deep Reinforcement Learning for NBA Player Valuation: A Temporal Difference Approach with Shapley Attribution

Author Ben Jenkins - 2026 MIT Sloan Sports Analytics Conference

The Problem: Many AI models tell us an action was valuable, but often struggle to explain why or how to attribute that value when five players are working in unison. For a front office, this "Black Box" problem makes it difficult to defend a multi-million dollar contract based solely on an AI’s recommendation.

Methodology Details: A temporal difference learning with a distributional win-probability model (Shapley Attribution) is used to break down the total "Expected Points Added" of a possession to fairly distribute credit to all five players.

Why It Matters: This provides unprecedented clarity for salary negotiations. You can now prove that a "three-and-D" wing is actually generating 30% of their value through "Gravity" and "Screening Assists" that traditional stats miss. It allows front offices to build a "Value-to-Dollar" map for the entire roster, identifying which players are truly driving the team's efficiency.

Actionable Takeaways:

Author: Dan Barto, IMG Academy - Technical Director Basketball Innovation

Overview: I was lucky enough to watch Darius Acuff develop and dominate in his junior and senior year at IMG Academy. He separated himself in my mind as the best to ever wear the IMG jersey. Acuff is having a generational season at Arkansas and John Calipari made the case on Sirius XM that he be the number one pick in the 2026 NBA Draft.

The Details: To support this argument I went deep with ChatGPT and Claude to create the Darius Acuff NBA Transition Report, which delivers a proprietary data driven “Five Layers Diagnostic” analysis to pinpoint both his strengths and developmental challenges, focusing on how his skills translate to the NBA.

FIVE LAYER DIAGNOSTIC PRO WORKFLOW

Results Summary:

Why It Matters: For NBA teams, the report’s granular analysis is invaluable and the future for assessing talent. The Five Layers Diagnostic framework provides actionable insights for coaches and front offices, ensuring that Acuff’s elite skills are maximized and his vulnerabilities addressed.

Overview: The off season starts now for many coaches and is the best time to upgrade analytics solutions. Advanced player tracking and real-time shot analytics were luxuries reserved for the NBA and elite Power Five conferences. In the current "Small Ball" era of AI, High schools, local clubs, and small colleges now have access to computer vision tools that turn a standard smartphone or a single portable camera into a pro level data hub.

Details: The shift is driven by “Edge AI” algorithms that run directly on mobile devices without needing a server farm.

Why It Matters: A Coach can generate an automated scouting report on an opponent in sixty seconds. This levels the playing field against better funded programs. For players, these tools create a "digital resume." AI verified shot charts and highlight reels are becoming the gold standard for recruiting, allowing players to prove their shooting consistency to college scouts.

Action Steps:

Overview: The junior year is the most important year for college recruiting. Off season training needs to be high intensity and data driven. This 4 week training plan accelerates development by integrating elite level sports science and computer vision into a high volume reps, high quality data driven routine.

Plan Details: The 4-week "AI Powered Accelerator" plan combines on court skills with the AI to ensure every rep is tracked and optimized for maximum skill growth.

Why It Matters: Entering the junior year, competition for college roster spots intensifies. Using AI as a "digital coach" provides objective metrics that traditional training often misses. Practicing smarter and building a verified data profile proves to scouts discipline and technical foundation to play at the next level.

Overview: The Basketball AI Maturity Model (B-AIMM) is a proposed strategic framework designed to help basketball teams or organizations’ evaluate their current technical infrastructure. Rather than viewing "AI" as a single tool, B-AIMM categorizes a team’s evolution into five distinct stages: Data Aware, Analytical, Integrated, Predictive, and Autonomous. This allows teams to move past the hype and identify exactly where their "intelligence gaps" exist.

The Details: The model assesses where teams fall in the various stages. A "Data-Aware" team might simply track box scores, while a "Predictive" program utilizes computer vision to simulate how a specific defensive rotation will fare against a high-volume pick-and-roll offense.

The goal is to develop action plans that help teams move through the stages. Ultimately filling gaps, implementing efficient processes or standard operation procedures with clear goals for technology and tools ROI.

Why it Matters: In an era of data overload, the loudest voice in the room shouldn't be the only one heard. B-AIMM provides an objective benchmark for ROI. It prevents "shiny object syndrome" like buying expensive software that the coaching staff isn't ready to use and ensures that technology investments actually translate into wins.

Want to learn more about B-AIMM? email Terry Frederick

Tools mentioned in this edition:

Until next time, Terry