BASKETBALL AI
Welcome to this edition of the Basketball AI newsletter.
While March Madness is winding down, scouting and off season player development strategies are all top of mind along with the upcoming NBA and WNBA drafts.
In this edition, the articles will focus on AI driven scouting and player development strategies as well as research on draft optimization and roster construction.
In this issue:
“Five Layer Diagnostic” or New NBA “Leverage Score” Metric?
The Hybrid Scout: Merging the "Eye Test" with AI Precision
Ball AI: Leveraging AI for Remote Player Data Driven Development
"AI Intangibles" Checklist for Modern Scouting
Identifying Player Surplus Value
Let’s dive in.
LATEST DEVELOPMENTS
PLAYER DEVELOPMENT
“Five Layer Diagnostic” or New NBA “Leverage Score” Metric?
Image source: Nano Banana 2 / Basketball AI
Author: Dan Barto, Technical Director Basketball Innovation, IMG Academy
Overview: In the last issue of the newsletter, I used ChatGPT and Claude to create a proprietary data driven “Five Layers Diagnostic” analysis for Darius Acuff to pinpoint both his strengths and developmental challenges, focusing on how his skills translate to the NBA.
Here, I will apply the ”Five Layers Diagnostic” analysis and compare it to the new NBA “Leverage Score” Metric using Jayson Tatum’s return form injury as an example.
The Launch: In January 2026, the NBA and AWS launched “Leverage Score”, an AI system that uses counterfactual modeling to measure which players and possessions actually change game outcomes. The engineering is serious: a LightGBM model trained on three seasons of data, processing tracking at 25 frames per second, running 3–5 alternative scenarios per possession in near real-time. Credit is distributed across all 10 players on the court. It’s the most sophisticated broadcast analytics tool the NBA has ever built.
The Gap: Leverage Score tells you who changed the game. It cannot tell a development coach why a shot missed. The entire system is built on expected field goal percentage which is a a difficulty rating, not a diagnostic. It assigns credit and blame. It does not prescribe a fix. The coaching world shrugged at the launch because it wasn’t built for them.
The Case Study: Jayson Tatum vs. Atlanta, March 27. He put up 26 points on 8 for 24 shooting nine games into his return from a ruptured achilles. Five points on 2 for 11 in the first half. Twenty one on 6 for 13 in the second. Leverage Score would flag Pritchard’s 36 point bench explosion as the high leverage performance and label Tatum middling. That’s where the NBA’s system stops.
The Workflow: I pulled the full play-by-play, fed it to AI, and ran Tatum’s 16 misses through the Five Layers of a Missed Shot framework. Six steps, one conversation, 30 minutes: build a leverage map, isolate every miss with context, classify each by root cause layer, estimate recoverable points, generate a coaching brief.
The Output:
Seven of 16 misses traced to L2; Feet Into Ground (Energy Transfer). Five of those were within six feet of the rim. The Achilles recovery has disrupted his explosion and finishing power. Tatum confirmed it himself: “Certain things of explosion, attacking, getting downhill… the pace and speed of certain plays that just felt really normal.”
Four more misses flagged at L1 Eyes Up (Decision) pressing, forcing shots to prove he’s back. PLOT estimate: 9–14 recoverable points.
Three drill prescriptions: rim finishing progressions, pre-game activation protocol, shot selection constraints.
The Point: The NBA spent millions answering “who changed the game?” I spent 30 minutes with AI answering “why did Tatum miss 16 shots and what should his staff work on Monday?” Both questions matter. Only one makes a player better.
Leverage Score answers “who changed the game?”
Five Layers answers “how do we change the player?”
One is built for broadcasts. The other is built for gyms.
SCOUTING
The Hybrid Scout: Merging the "Eye Test" with AI Precision
Image source: Basketball AI
Overview: As the AAU circuit heats up, college coaches and assistants face the daunting task of evaluating thousands of prospects across simultaneous games. While the traditional "eye test" of honing in on a player’s motor, demeanor, and "feel" for the game remains the gold standard, the volume of talent makes manual scouting an uphill battle. The modern, seasoned scout is now evolving into a hybrid evaluator, using Artificial Intelligence to filter the noise and validate what their gut is telling them in real-time.
The Details: The integration of AI doesn't replace the coach's seat in the bleachers; it sharpens their focus. AI driven platforms process raw game film to provide immediate "Expected Value" metrics.
If a guard has a quick first step, AI tools can quantify their burst speed and lateral quickness relative to elite college benchmarks.
Automated scouting tools can track "gravity" which is how much defensive attention a player draws even when they don't have the ball
"Hidden" impact players who may not show up in a standard box score but consistently create high quality shots for teammates can be identified.
AI can handle the data heavy lifting so coaches can spend their time observing the intangibles like how a player reacts to a benching, their communication during a defensive breakdown and their coachability.
Why it Matters: In the high stakes recruiting environment, efficiency is the ultimate competitive advantage. Combining traditional scouting with AI reduces "observer bias" and prevents staff from missing out on late bloomers who might lack flashy stats but possess elite physical metrics. Ultimately, this hybrid approach ensures that offers (scholarship and NIL) are backed by both a veteran’s intuition and hard objective data, leading to better roster construction and long term program stability.
PLAYER DEVELOPMENT
Ball AI: Leveraging AI for Remote Player Data Driven Development
Image source: Ball AI
Overview: Traditional player development often suffers from a significant visibility gap, where coaches lose insight into a player's progress the moment they leave the team facility. While team practices are structured and monitored, individual "homework" sessions frequently rely on subjective self-reporting rather than verified performance data.
Ball AI addresses this disconnect by providing an AI powered training platform that transforms any standard hoop session into a repository of trackable metrics and shareable video. By digitizing individual workouts, the platform offers coaches a comprehensive view of player activity outside of scheduled gym time.
The Details: The system operates through a mobile interface where players position a smartphone to face the basket. Using computer vision, the AI automatically captures a wide array of shooting analytics, including:
Shot Outcomes: Automated tracking of makes, misses and overall shooting percentages.
Biometric & Spatial Data: Real-time analysis of release angles, shot speed and specific court locations.
Automated Documentation: Each session produces a full breakdown of performance trends and technical feedback.
The Coaches Dashboard: This serves as the central hub for remote management by coaches. The interface allows staff to assign specific shooting drills, establish weekly volume goals, and review video footage of every repetition to ensure mechanical consistency. This functionality effectively extends professional oversight into every individual workout a player performs independently.
Why it Matters: A primary challenge in coaching is maintaining accountability when a coach cannot be physically present. Ball AI replaces verbal "work logs" with objective data and video proof, allowing coaches to identify exactly who is putting in the work and whether that effort is yielding tangible improvement.
Ball AI technology addresses the rising "pay-to-play" barrier in basketball. With high end shooting machines often costing $5,000 and private trainers charging premium hourly rates. Elite level feedback has historically been restricted by budget. By utilizing Ball AI (hardware already in a player’s pocket), this platform democratizes professional grade structure and feedback for any athlete with a smartphone.
SCOUTING
"AI Intangibles" Checklist for Modern Scouting
Image source: Nano Banana 2 / Basketball AI
Overview: While traditional scouting relies on a "gut feeling" about a player's character, 2026-era AI tools can now quantify these behaviors by analyzing micro-movements and spatial relationships. When you review your automated clips, use this checklist to look for the "invisible" winning traits that AI makes visible.
Checklist:
1. Off-Ball "Gravity" & Space Creation
The AI Metric: Gravity Score / Defensive Attention.
What to Look For: Does the defense tighten or shift toward the player even when they don't have the ball? AI tracks defender proximity (in centimeters) to see if a player "commands a crowd," opening up lanes for teammates.
The Intangible: Selflessness and tactical importance. Off-Ball "Gravity" & Space Creation.
2. Defensive Processing Speed (Latency)
The AI Metric: Closeout Efficiency / Reaction Time.
What to Look For: How many milliseconds pass between an offensive pass and the defender’s first step? AI measures "decision latency" to identify players with elite anticipation who "see the play" before it happens.
The Intangible: Defensive IQ and mental alertness.
3. Recovery "Grit" & Second Effort Tracking
The AI Metric: High-Intensity Burst Volume.
What to Look For: Does the player’s speed drop in the 4th quarter? AI monitors "acceleration curves" to see who maintains their elite burst after 30 minutes of play. It also tags "second chance efforts" like sprinting to the corner for a contest after being beaten on a screen.
The Intangible: Conditioning, motor, and competitive resilience.
4. Shot Selection & "Shot Quality" IQ
The AI Metric: Expected FG% (xFG) vs. Actual.
What to Look For: Is the player taking "good" shots that miss, or "bad" shots that happen to go in? Evaluating the likelihood of a make based on defender distance and shooter balance.
The Intangible: Decision making and understanding of the system's "best available shot.”
5. Communication & On-Court Leadership
The AI Metric: Spatial Synchronization.
What to Look For: On defensive rotations, are all five players moving in a synchronized "shell," or is one player consistently late? AI heat maps show "defensive holes." A player who consistently fixes these holes is often the "vocal captain" directing traffic.
The Intangible: Vocal leadership and help side awareness.
Why it Matters: In the fiercely competitive world of college recruiting, relying solely on surface level statistics can lead to missed opportunities and costly roster mistakes. The "AI Intangibles" Checklist bridges the gap between raw data and a player's true game impact, offering objective metrics for behaviors that were once considered unquantifiable. This hybrid approach ensures that evaluations are comprehensive, backed by precise evidence, and ultimately leads to smarter and more resilient roster construction.
RESEARCH SUMMARY
Identifying Player Surplus Value
Image source: Basketball AI
Title: Scouting Anyone: Probabilistic Player Archetypes for Any League
Author: Marlin Myrte, Sebastian Buzzalino, Thierry Aymerich
Institution: 2026 MIT Sloan Analytics Conference
Publication Date: March 7, 2026
The Problem: Addressing the "Average" Player Inefficiency
Modern basketball decision making still relies heavily on rigid positions, subjective role labels, or hard clustering models that force players into a single category. This creates inefficiencies in scouting, lineup construction, and cross-league evaluation especially when roles are fluid, hybrid, and context dependent. Traditional clustering methods also tend to identify “average” players rather than the true stylistic extremes that coaches and scouts think in terms of. As a result, teams frequently misdiagnose player fit, overpay for redundant skill sets, or undervalue players who could thrive in a different ecosystem.
Methodology: From Abstract Clusters to Real-World Prototypes
The researchers utilize Archetypal Analysis (AA) and Archetypoid Analysis (ADA) to solve these issues.
Archetypal Analysis (AA): Think of this like a color wheel. Just as any color can be described as a mix of "pure" red, blue, and yellow, AA describes every player as a percentage of "pure" stylistic extremes. A player isn't just a "Center"; they might be 60% "Rim Protector" and 40% "Stretch Big".
Archetypoid Analysis (ADA): While AA creates theoretical "perfect" extremes, ADA anchors these roles to real-world benchmark players. Instead of an abstract statistical average, a scout can describe a prospect as "40% stylistically similar to Nikola Vucevic and 35% to Mitchell Robinson".
By applying these models to simple box score data and advanced play type metrics, the framework remains accessible to teams at any budget level.
Why it Matters: ROI through Stylistic Alignment
For GMs and coaches, this framework directly improves surplus value identification and risk management. It enables clearer separation between player talent and player fit, reducing costly acquisition mistakes driven by surface level production. Scouts can search globally for stylistic matches rather than positional look alikes, opening undervalued markets. Coaches can gain a insight explain why certain lineups fail despite “good players”.
Crucially, probabilistic roles align with how basketball actually works: players shift responsibilities based on teammates and schemes. This allows front offices to project how a player’s value will change after a trade, signing, or coaching change turning context dependency from a blind spot into a controllable variable with measurable ROI.
Actionable Takeaways:
Audit Your Roster Ecosystem: Map archetype distributions for your top lineups to identify redundancy, spacing compression, or missing complementary roles.
Upgrade Scouting Filters: Replace position based shortlists with archetype similarity searches across leagues to uncover undervalued fits.
Stress Test Acquisitions: Before trades or signings, simulate how a player’s archetype mix interacts with existing core players to forecast on/off impact.
Draft Optimization: Rather than drafting for "best player available," target specific archetype percentages (e.g., a "Stretch 4" with at least 50% "Pick-and-Pop" concentration) to maximize the efficiency of existing high usage scorers.
Research Paper Link: Scouting Anyone: Probabilistic Player Archetypes for Any League
QUICK HITS
🛠 Tools Spotlight:
Tools mentioned in this edition:
Leverage Score - Utilizing machine learning to predict win probability.
Ball AI - AI Powered basketball workouts.
Coming Soon to the Basketball AI Website! - A Library of the most useful Basketball AI tools…organized and categorized all in one spot.
📰 Everything else in Basketball AI:
Expert Insight: "Whether it's an AI companion that helps you understand a basketball game or a virtual coach that helps you learn new skills, democratizing this expertise is the most impactful thing I could accomplish. We’re teaching machines to notice the right details at the right time. Using terms like 'pick and roll' or 'lateral shuffle' to build technology that performs complex video analysis similar to how a human sports analyst does it."
— March 25, 2026, Gedas Bertasius, Computer Science Researcher and former small forward (UNC-Chapel Hill) on his research into AI models.
Expert Insight: "Coaches are now adding 'data analyst' to their roles. They are intaking data from dozens of sources—from wearable GPS sensors to historical trends—to figure out the optimal training plan. We’re seeing a landscape where data flows allow top-level coaches to increase practice intensity to ensure optimum stamina for the rapid-fire schedule of the tournament."
- March 2026, Mollie Brewer & Kevin Childs, AI researchers (University of Florida report on how the Florida Gators and other top programs utilized AI data processing to manage player load and peak performance specifically for the SEC and NCAA tournament runs.
📆 Upcoming Conferences, Events and Guides:
March 17, 2026 Last issue of Basketball AI
April 10-11, 2026 Connecticut Sports Analytics Symposium (CSAS)
April 14-16, 2026 Sports Business Journal (SBJ) Tech Week
April 20-21, 2026 International Conference on Athlete Performance
That’s a wrap for this edition. The next newsletter will be published Tuesday April 14, 2026.
Tell us what you thought of today's newsletter to help us improve it for you.
Until next time, Terry