GitHub Copilot, and Cursor are changing how developers write code. Microsoft 365 Copilot and Google’s Gemini in Workspace are aiming to transform how we all handle productivity. These systems don’t perform isolated tasks, but assist and augment our capabilities within our existing workflows. So, what does this “co-pilot revolution” look like when it fully materializes in sports?
Athletes: The Personalized Performance Partner
My thirteen-year-old is on his school basketball team. After each game, he asks me for an in-depth analysis of how he did. Instead of just a basic stat line or my limited observations, he could turn to an AI feedback tool.
I’ve been recording his games in hopes of using (or creating) a tool such as an Athlete co-pilot. After a particularly tough game where his team struggled against a zone defense, he might ask, “As the shooting guard, I felt lost against their 2-3 zone today. Take a special look at my positioning in the fourth quarter. Where should I have been moving when the ball was on the wing to get open or help my teammates?”
The Co-Pilot would would outperform me in several ways, including:
- Analyzing the game video I recorded, using computer vision to track his positioning, the ball’s movement, and the defensive shifts of the opponents whenever they were in that 2-3 zone
- Comparing his movements to proven strategies of attacking a zone defense, perhaps showing him clips of how others in the game and even more experienced players find gaps or create opportunities in similar situations
- Providing specific, actionable feedback: “Okay, I’ve reviewed the footage. When the ball was on the right wing, you often drifted towards the corner? (AI sourced highlight). If you had made a quick cut towards the high post or flashed to the short corner on the baseline, it could have opened up a passing lane or drawn a defender to create space for a teammate. Want to see a few examples of those cuts from the video, and I can suggest some drills to practice recognizing those openings?”
For a college or professional athlete, the AI Co-Pilot could scale up its sophistication. Consider a pro soccer midfielder who feels their performance has been inconsistent. They could prompt their AI Co-Pilot, “I need to understand why my pass completion rate into the final third dropped by 15% in the last three matches, compared to my season average.”
The AI could then analyze not just their passing technique from match footage (identifying if they were rushed, off-balance, or choosing overly ambitious passes), but also correlating this with their physical output from wearable sensors (was fatigue a factor?), their positioning relative to defensive pressure (were they receiving the ball in tighter windows?), and even the tactical setup of the opponents they faced.
The Co-Pilot might then present a summary: “Analysis shows that in the last three matches, you received the ball 20% more often while under immediate pressure from two or more defenders. While your decision to attempt forward passes remained consistent, your body positioning when receiving was often closed, limiting your immediate options. Additionally, your sprint distance in the 10 minutes prior to these passes was 12% higher on average. Let’s look at some clips of your positioning before receiving the ball, and I can suggest some drills for maintaining composure and vision under duress, as well as discussing potential load management adjustments with your performance coach.”
While there are a number of hurdles to make this a reality (ranging from ensuring high-quality, integrated data and developing truly context-aware AI, to addressing usability, trust, ethics, and accessibility), we are entering a phase where this is largely technically possible. A key unique aspect of the Athlete Co-Pilot is that this insight, for whatever questions they may have, is now directly in the athlete’s hands. Initially, coaches will undoubtedly play a crucial role in validating these AI-driven suggestions, and it’s precisely this human oversight and collaborative refinement that will sharpen these tools, making them progressively more accurate and trustworthy over time.
For Coaches: Tactics & Talent
Coaches, too, stand to gain an incredibly powerful assistant.
At the professional and collegiate levels, an AI co-pilot might offer real-time tactical suggestions during games, subtly delivered to a tablet: “Analysis indicates their pick-and-roll is creating weak-side open looks on 3 of the last 5 possessions. Consider switching to ‘Ice’ coverage to force the ball-handler baseline and simplify rotations.”
It’s not lost on me that, in order to preserve the purity of the game, leagues will not allow technology like this in-game for years to come. However, there will be other ways to leverage this intelligence. For instance, suggestions could be used pre-game based on previous match-ups and footage from opponents’ previous ten games with the same starting lineup.
But even for youth coaches, who are often volunteers juggling multiple commitments, AI could be transformative. Imagine an AI helping to
- Generate varied and engaging practice plans: When I coached my son’s rec basketball league, it was a challenge to come up with the right drills for the team. I often resorted to generic Jr. NBA drills. Given the widely diverse skillsets on the team, it would have been ideal to tailor the drills to each player to help develop their own strengths and weaknesses.
- Analyze simple practice footage: A coach could upload a short clip, and the AI could highlight common errors in basic skills (e.g., “Several players are consistently not getting low enough in their defensive stance”). Of course,another AI agent could pick up this analysis and push the analysis (along with coach commentary) to the player.
- Provide easy-to-digest summaries of opponent tendencies: For instance, a U14 soccer coach prepping for the Lightning with only a couple of matches information (a 3-1 win, 2-2 draw) and a note that “their #9 striker loves long shots” could get this from their AI Co-Pilot: “Coach, the Lightning average 2.5 goals and are offensively strong. With their #9 favoring distance shots, let’s prioritize closing down space outside the box in practice and alert our keeper. I can also search for public video of #9 to identify shooting patterns.” This turns fragments into a focused, actionable talking point.
This kind of intelligence doesn’t replace the art of coaching – the motivation, the mentorship, the meaningful connection. Instead, it enhances both coaching efficiency and effectiveness by shouldering more of the analytical and preparatory burdens, freeing up valuable time for coaches to focus on direct instruction, mentorship and player development.
For the Front Office: The Roster Architect
The Current State: Front offices rely on a powerful combination of tools and human expertise: intricate spreadsheets for cap management, siloed statistical models for player projections, invaluable qualitative insights from scouting networks, various video analysis platforms, and, critically, countless hours of human discussion, debate, and decision-making heavily guided by experience and intuition. It’s a system that has produced champions, but it often struggles with seamlessly integrating all available data streams and rapidly modeling the deep, multi-year consequences of complex decisions.
The Fourth Wave: The Pro & College GM and Coach’s central AI co-pilot for all things roster-related. This system would be built upon:
- A Unified Data Foundation: Integrating everything – player performance data (historical and contextually projected), detailed financials and contracts (with a dynamic, multi-year CBA rules engine), evolving asset values (players, draft picks), comprehensive risk factors (injury likelihood), and even carefully developed proxies for “softer” elements like scheme fit.
- An Advanced Modeling Layer: This is the “engine room,” using specialized AI models (distinct from the LLM interface) for the heavy lifting: generating those contextual performance projections, executing flawless financial calculations, providing algorithmic trade valuations based on historical market data, assessing multifaceted risks, and running complex simulations to explore a probabilistic range of outcomes for any given scenario.
- A Generative AI Interface: This is the “cockpit.” The GM or AD interacts using natural language (“What are our options for acquiring a playmaking midfielder under 25, with a transfer value below X, who has experience in a high-press system? Model the impact on our wage bill and squad depth for the next two seasons.”) or through a visual “sandbox” environment, receiving clear summaries, exploring scenarios, and drilling down into details.
Given this scenario is the subject of our next post, we’ll limit ourselves to two closing points:
- One of the most profound benefits of this AI co-pilot is its ability to model and highlight the indirect impacts that are often difficult to fully grasp today, like the ripple effects on other players or long term cap implications given upcoming free agencies, etc.
- It’s also crucial to emphasize that such a powerful tool isn’t an infallible oracle. It requires rigorous, continuous validation: ensuring impeccable data integrity, meticulously backtesting each individual AI model (for performance projections, financial calculations, risk assessments), and so on.
The rise of the AI Co-Pilot in sports signals something fundamental: a shift in how athletes hone their craft, how coaches devise strategies and nurture talent, and how front offices architect championship-contending teams. It’s about equipping every key individual in the sports ecosystem with an intelligent assistant that can process vast amounts of information, see patterns a human eye might miss, and help us explore possibilities we haven’t even fully imagined.
Of course, this journey isn’t a simple plug-and-play affair. Building these sophisticated co-pilots, ensuring their accuracy, fostering trust, and integrating them effectively will require planning, significant investment in both technology and talent, and a steadfast commitment to ethical development and data privacy.
We’ll always need to remember that the human element – the coach’s intuition, the player’s drive, the GM’s experience and leadership – remains at the very heart of the game. However, AI is rapidly becoming the sophisticated analytical engine that will amplify these human strengths, processing the complex data and running the intricate simulations needed to inform smarter strategies and accelerate development.