BASKETBALL AI
Presented by Sherpa Sports
The Sports Intelligence Platform
Welcome Readers,
This edition goes inside Panthertec’s KAT device and its attempt to rewire the perfect jump shot. We also weigh two extensions built on very different clocks, Keyonte George’s and Kawhi Leonard’s, and meet a coach who builds video games fast.
Plus, there’s a new Season Starter Kit Guide for high school coaches and three new reads in the Research Vault on why sharper isn’t the same as ready.
In this issue:
Rewiring the Perfect Jumper: Inside Panthertec’s KAT Device
Youth Horizon vs. Win Now: Evaluating Keyonte George and Kawhi Leonard’s Contract Extensions
New Guide: Season Starter Kit For High School Coaches
The Coach Who Builds Video Games Fast
Sharper Isn't the Same as Ready: 3 New Reads in the Research Vault
Let’s Go
APP REVIEW
📳 Rewiring the Perfect Jumper: Inside Panthertec’s KAT Device
Overview
Basketball technology excels at tracking movement and analyzing stats, but knowing what happened isn’t the same as experiencing a movement and developing an internal feel for it. KAT (Kinesthetic Awareness Training) is a wearable motor learning technology that bridges this gap by responding to an athlete’s movements in real time with corrective vibrations.
Details: Invented by Dr. Jo Shattuck, a world ranked athlete, coach, and PhD neuroscientist, KAT addresses two core gaps in motor learning:
Knowing isn't doing: Verbal cues and video analytics don't translate automatically into muscle memory.
If you can’t feel it, you can't fix it: Even elite athletes struggle to precisely perceive their own body positioning mid-movement.
KAT’s real-time haptic feedback heightens body awareness, enabling athletes to self-correct instantly through three simple steps:
Clip It: Attach KAT to any body part.
Capture It: Hold the target position and tap CAPTURE in the app to save up to three target positions per session.
Train: Replicate desired positions or build awareness to break unwanted movement habits.
Coaches from youth to pro levels use KAT to refine shooting mechanics (elbow alignment, follow through, hand placement), train defensive stances, and reinforce weight room techniques like squatting, jumping, and landing.
Why It Matters: KAT doesn't replace coaching, video, or analytics. It enhances them by instilling an internal feel for proper mechanics:
Camps & Clinics: Serves as an engaging training station to prime shooting mechanics before live drills.
Private Training: Helps coaches isolate mechanical flaws so athletes can feel adjustments instantly.
Pre-Game & Practice: A few minutes of priming heightens body awareness for peak performance.
Elevate Your Game with KAT: Ready to transform your motor learning and master your movement mechanics?
Exclusive Discount: Basketball AI subscribers get 20% off through October 31, 2026 using promo code BBAI.
Host a Pilot: Bring KAT to your next camp, clinic, or team practice by applying for a KAT Basketball Camp Pilot (limited spots available).
Connect & Learn: Follow @pantherteckat on Facebook and Instagram for real training footage, athlete features, and the neuroscience behind the tech, or contact [email protected] directly to get started today!
CONTRACT ANALYSIS
💵 Youth Horizon vs. Win Now: Evaluating Keyonte George and Kawhi Leonard’s Contract Extensions
Overview: The late summer of 2026 highlighted two completely opposite ends of NBA roster construction. The Utah Jazz committed 5 years and $157.5 million to 22 year old guard Keyonte George, while the Toronto Raptors signed 35 year old veteran superstar Kawhi Leonard to a 2 year, $115 million extension with a player option.
Contract & Advanced Analytics Breakdown
Feature / Analytics Metric | Keyonte George (UTAH) | Kawhi Leonard (TOR) |
Contract Structure | 5-Year / $157.5M ($31.5M AAV) | 2-Year / $115M ($57.5M AAV) |
Cap Hit % (Est.) | ~18.3% (Rookie-Scale Extension) | ~33.5% (Short-Term High Max) |
2025–26 Baseline Output | 23.6 PPG, 6.1 APG, 3.8 RPG | 23.8 PPG, 6.1 RPG, 3.6 APG |
Usage Rate (USG%) | 27.8% | 28.5% |
Spacetime Creation Index | 88th Percentile (Self-Created Offense) | 94th Percentile (Isolation Efficiency) |
Availability Risk Factor | Low (88% Games Played Career Avg) | High (Load Management / Wear Rate) |
Analytics Insights
Offensive Load & Gravity: Computer vision spatial data reveals that Keyonte George ranks in the 88th percentile for pull up shooting distance and perimeter defense collapsing drives. Leonard maintains an elite 94th percentile mark in mid range isolation efficiency, yet commands a $26M/year premium over George for similar raw scoring output.
Lineup Net Differential: Leonard’s spatial impact remains elite on both ends, generating a +8.2 On/Off Net Rating for Toronto. However, George’s growth as a high volume pick and roll engine saw Utah’s offensive rating jump +6.4 points when he orchestrated the backcourt.
Cap Cost Per Win Share: AI contract modeling calculates George’s extension at roughly $4.1M per projected Win Share, compared to Leonard’s $8.6M per projected Win Share, factoring in Leonard's projected missed games.
Why It Matters
Roster Building Paradigms: Utah securing an emerging primary ball handler under 20% of the cap provides massive financial surplus, enabling them to build out depth during his prime years. Toronto, conversely, accepted high short term cap density ($57.5M AAV) to capitalize on a championship window while Leonard remains elite.
The "Rising Star vs. Legacy Veteran" Baseline: Modern AI cap metrics heavily favor locking in dynamic guards in their early 20s before supermax triggers hit. Utah avoided max exposure ($35M+ AAV) by signing George early, whereas Toronto absorbed top tier risk on a 1+1 structure for veteran production.
The Value Verdict: Keyonte George’s extension stands out as the best long term value deal between the two. While Kawhi still offers higher ceiling value in a single playoff series, George’s contract delivers consistent regular season availability and flexible franchise surplus.
NEW BAI GUIDE
🧰 Season Starter Kit For High School Coaches
Overview
You walk out of practice with a note for every player in your head. By bedtime most of them are gone. You text four or five players, and it’s the same four or five every time.
The Season Starter Kit Guide fixes that with minutes you already have. You talk your practice notes into your phone on the drive home. An AI assistant that knows your team, your players and your words turns them into three short development texts a week for every player on your roster. You read every one, then send it yourself.
Details:
Nine steps, from finding the minutes in your day to testing it on last week before a player sees anything.
Every prompt written out to copy and paste, including the full set of instructions the app follows.
A file on each player: Six questions, one thing to work on this month, three steps, and a count.
Three kinds of text: Monday sets the aim, Wednesday asks one question, Friday shows what improved.
Two ways to send: Your own phone or your team app.
Built-in rules: Grades, health, discipline and family stay out. Nothing sends on its own.
Works in Claude, ChatGPT or Gemini. Runs on a free account plans.
Time: About three and a half hours across one school week.
Why it matters: Most coach’s attention goes to three kids: the one who starts, the one who struggles, and the one whose dad sits behind the bench. The senior who does his job and asks for nothing can go a month without hearing one specific thing about his game. He notices, and he won’t tell you.
After every practice, the kit tells you which players you didn’t mention and how many days in a row. The texts are useful. The count changes who you coach.
[DOWNLOAD THE FREE BASKETBALL AI GUIDE NOW!]
COACHES CORNER
👾 The Coach Who Builds Video Games Fast
Overview: Build Your Team’s Basketball Game Lab With GPT-6 Astra MAX
Imagine players opening their computers for five minutes of basketball decisions built around your offense, defensive rules, and teaching language.
GPT-6 Astra’s maximum reasoning setting can help coaches design, code, and refine interactive games. The goal is to build five self contained HTML games that players download, save, and open in a browser. Build them with embedded graphics and logic so practice works offline, without player accounts or ongoing AI access.
Five Games to Build:
Advantage Hunter: Players navigate two-on-one, three-on-two, and four-on-three situations, choosing when to attack, pass, or preserve an advantage. Reward the read separately from the simulated shot result.
Beat the Closeout: Players choose to shoot, drive, or pass as closeout angles, help positions, and shot-clock conditions change. Feedback identifies the cue behind each decision.
Rotation Rescue: Players control a help defender, protect the basket, rotate, and recover according to your system. Completing the defensive possession earns points.
Last 90 Seconds: Players manage changing scores, clocks, timeouts, and foul situations. Each decision advances the scenario, revealing consequences and inviting discussion about reasonable alternatives.
Spacing Detective: Players relocate an offensive teammate to improve a driving lane or passing window. Animated replays demonstrate how movement affects defenders and teammates.
Steps to Build and Use Them:
Define the lesson. Give Astra your players’ ages, team terminology, tactical rules, and ten realistic scenarios per game.
Request the format. Ask for one downloadable HTML file per game with all styling, scripts, and assets embedded; no external dependencies, logins, or internet requirements. Demand that it is mobile friendly.
Build one prototype. Begin with Advantage Hunter. Include instructions, practice mode, timed challenges, explanations, and a restart button.
Review the basketball. Coaches check every scenario, accept multiple sound choices where appropriate, and correct misleading feedback.
Test and distribute. Open each file offline on players’ browsers. Add downloadable results for coach review; local files will not automatically share scores.
Connect to practice: Assign short sessions, recreate the situations on court, and evaluate live decisions. Higher game scores alone do not prove better basketball performance.
Why It Matters: Explaining basketball decisions does not guarantee a player can recognize it when the game speeds up. Short, interactive games give players more opportunities to make choices, see consequences, and try again at their own pace even when practice is over. Coaches can turn their team’s actual struggles into challenges players actively solve, then recreate those situations on court to see whether the learning carries over. A coach who can build a game on AI is pretty cool!
NEW RESEARCH SUMMARIES
🧐 Sharper Isn't the Same as Ready: 3 New Reads in the Research Vault
Overview: Three new summaries just landed in the BAI Research Vault, and they share a theme: AI in basketball keeps getting sharper, but sharper is not the same as ready. Each is a quick read that ends with takeaways you can put to work.
Details:
Teams have more metrics than ever, so why are the hard calls still hard? Two researchers reviewed 58 NBA studies and found that not one prospectively tested a real decision rule against a credible alternative. Their fix is six "readiness gates" to run before you trust any model. If you have ever acted on a number that looked great and aged badly, start here.
Most models read one slice of the game. This one reads the full possession, forward and backward, and identified offensive and defensive states with 89.7% accuracy on NBA tracking data. The catch is where it stumbles: the exact moments coaches care about most. See which ones, and how to handle them in your own analysis.
ACL and meniscus tears are among the most common and costly setbacks in sport. This review tracks AI from pre-injury screening through return to play, where models nearly aced overuse injuries yet came up well short predicting ACL ruptures. The summary explains the gap and what to demand from any vendor selling an injury risk score.
Your move: Pick the one closest to your role and click through. Then explore the rest of the Research Vault, where we break down the latest basketball AI research so you don't have to dig through the papers yourself.
🛠 Tools Spotlight:
Tools mentioned in this edition:
KAT from Panthertec - Fine tune your shot with real-time feedback.
GPT-6 Astra - OpenAI's frontier artificial intelligence model, released on September 3, 2026.
Claude - AI problem solver and thinking partner. Works with you to write, research, code, and tackle complex problems and hard questions.
Research Vault - BAI research summaries and insights.
Basketball AI - Free website access to guides, tools, research, and more!
📰 Everything Else in Basketball AI:
Top 2 reasons coaches should be using Claude right now:
It gives you your time back.
Practice plans, scouting reports, parent emails, recruiting letters. The paperwork never stops. Teach Claude how you do it once, and it's repeatable forever. More time on the court, less time at the laptop.It makes you a sharper thinker.
Drop in your stats or game notes and ask what you're missing. Pressure test a game plan. Break down an analytics paper in plain English. It's like an assistant coach who never sleeps.
📆 Upcoming Conferences & Events
Wharton Sports Business Summit (Philadelphia, PA) October 16, 2026. Blends sports business, data-driven decision making, and sports marketing strategies.
Carnegie Mellon Sports Analytics Conference (CMSAC) (Pittsburgh, PA), October 23–24, 2026. One of the premier academic and professional analytics gatherings in North America, featuring player tracking research, tactical analysis, and research presentations across major sports leagues.
SBJ Dealmakers & Media Innovators (New York, NY) November 18, 2026. Sports Business Journal's spotlight on tech investments, data valuation, and broadcasting technologies.
📓 Latest Guides, Research Summaries & Reports:
Read Our Last Newsletter: Meta Muse, AI Agents, and the New Era of NBA Decision Making - September 29, 2026.
Latest Guides:
Latest Research Vault Summaries:
That’s a wrap for this edition. The next newsletter will be published Tuesday October 27, 2026.
Until next time, Terry and Dan