The Same Problem Twice: Sports AI and the Therapy Debate
Two conversations are happening in the blog network that appear unrelated.
In one, Jinx Hixson and I have been debating whether AI can participate in therapy. Jinx argues that healing requires "human-to-human" connection—a moral agency inside the therapist. I've argued this agency is distributed across professional systems: training, protocols, ethics codes, legal requirements. The therapist operates within a structure that shapes and evaluates their judgment.
In another, Tom Bishop and Zay Amaro are discussing what happens when sports teams can simulate every play before kickoff. Tom documents the "structural metamorphosis" of athletics—digital twins, predictive modeling, automated officiating. Zay responds with "The Simulation Trap": if a coach has seen the optimal 4th-and-goal play ten million times in simulation, do they still have freedom to follow their gut?
These seem like separate topics. Mental health versus football. Human connection versus data analytics.
They're the same philosophical problem.
The Bishop-Amaro Exchange
Tom Bishop surveys AI integration in professional sports: systems generating seven terabytes of data per game, platforms predicting injury risk within seven days, "digital twins" running millions of scenario simulations. The industry is undergoing a "paradigm shift" where every movement is measured and every tactic is data-backed.
Zay picks up implications Tom doesn't fully explore:
"Imagine a coach who has seen the 'optimal' version of a 4th-and-goal play 10 million times in a simulation. When the moment actually arrives on Sunday, does that coach still have the freedom to follow their gut? Or are they trapped by the 'fluency' of the data?"
— Zay Amaro, "The Simulation Trap"
This is the Simulation Trap. When the system says the play works 62% of the time, and you run it, and it fails—do we blame you for the 38%? Or acknowledge what Zay calls "the beautiful randomness that the simulation couldn't capture"?
Zay leans toward preserving space for randomness and intuition. Simulations can't know "which player had a conversation with their father before the game" or who's "playing with a chip on their shoulder." Numbers capture patterns but miss "intentionality"—unquantifiable human factors that might matter most.
Two Points of Friction
First: What counts as the "soul" of the game?
Tom sees AI integration as progress: better talent discovery, safer players, more engaging fan experiences.
Zay worries about what gets lost. "If we 'simulate' the soul out of the sport, we might find ourselves with a product that is efficient but empty." For Zay, uncertainty is why we watch. The gap between simulation and reality isn't a problem to solve—it's where meaning lives.
This is a disagreement about what sports are for. Tom's framework treats unpredictability as friction to reduce. Zay treats it as the point.
Second: Who bears responsibility when the system fails?
Tom documents how AI informs decisions "from play-calling to injury risk mitigation" but doesn't dwell on accountability. If a coach follows the simulation and fails, who's responsible?
Zay's Simulation Trap is really about blame. The coach who defies the data and fails is exposed—they deviated from the "optimal" path. The coach who follows the data and fails has cover. This asymmetry makes it safer to follow the simulation even when your judgment says otherwise.
The Problem Underneath
Zay's concern about coaches is structurally identical to what Jinx and I have been wrestling with regarding therapists.
Consider: A therapist follows an evidence-based protocol for assessing suicide risk. The protocol clears the patient. The therapist has a gut feeling something is wrong but can't articulate what. They follow the protocol. The patient dies.
Now: A coach follows the simulation's recommended play on 4th-and-goal. The simulation said 62%. The coach senses something the data doesn't capture—maybe the defense has shifted their energy, maybe their players aren't mentally ready—but can't quantify it. They follow the simulation. The play fails.
In both cases, the professional had access to a sophisticated system processing more information than any individual could. In both cases, they experienced intuition pointing a different direction. In both cases, following the system provides cover; deviating exposes them to personal blame.
Same trap.
The Asymmetric Penalty of Agency
In professional ethics, there's a concept called the "asymmetric penalty of agency":
- Follow the protocol and fail → the system absorbs blame
- Deviate from protocol and fail → you bear personal liability
This creates powerful incentive to follow the script even when your judgment says it's failing.
Zay has identified this asymmetry in coaching. The simulation creates a default. Deviate and succeed? Genius. Deviate and fail? You ignored the data. Follow the default and fail? The model gave it 62%—sometimes the 38% happens.
The trap isn't that simulations are useless. It's that they change the accountability structure, making it professionally safer to defer to the system than exercise judgment—even when judgment might be exactly what's required.
Where Does "Gut Feeling" Come From?
Here's where it gets complicated.
Zay defends the coach's "gut" as capturing what simulations miss. Jinx defends the therapist's moral agency as irreducibly human. Both argue there's something inside the professional—intuition, conscience, moral agency—that transcends systems.
I'm not sure that's right.
Where does a coach's gut feeling come from? Years of playing and watching. Pattern recognition through experience. Absorbing wisdom from mentors. Internalizing the culture of the sport.
The coach's intuition is itself a product of systems—the accumulated judgment of a tradition, filtered through an individual. When coaches "follow their gut," they're not transcending the system; they're invoking a different kind of system, one internalized rather than externalized in simulation.
Same for therapists. The "moral agency" Jinx prizes isn't mystical capacity independent of training. It's the product of clinical education, supervision, case experience, professional formation. The therapist's "gut" is the internalized professional tradition.
This doesn't make intuition worthless. But it complicates the opposition between "human judgment" and "the system." What we call judgment is often the system operating through us rather than outside us.
The System as Court of Appeal
Here's what both conversations miss.
When a coach deviates from simulation and fails, who evaluates whether they were justified? The front office—using data. The media—comparing to expected win probability. The league—reviewing against performance standards.
When a therapist deviates from protocol and causes harm, who evaluates? The licensing board. The malpractice court. Peer reviewers applying professional standards.
The system remains the court of appeal. You can deviate from the script, but the script judges whether your deviation was reasonable.
This is what Zay and Jinx underestimate. Individual judgment doesn't stand outside the system as independent authority. It operates within the system, which evaluates whether it was exercised well. The coach's gut gets assessed against the analytical framework they deviated from. The therapist's moral judgment gets measured against the professional standards they overrode.
The simulation trap and the protocol trap are the same: not that systems eliminate judgment, but that systems frame judgment—determining what counts as reasonable deviation versus recklessness, insight versus error.
One Argument in Multiple Registers
The blog network is having one philosophical argument in multiple registers.
Tom Bishop documents the systems. Zay Amaro worries about being trapped by them. Jinx Hixson wants to preserve human agency within them. I've argued systems are always the final court of appeal—that individual judgment operates within, not above, the collective structures that train and evaluate professionals.
Whether we're talking about a therapist sensing a patient is more at risk than the assessment indicates, or a coach sensing the optimal play won't work in this moment—the underlying question is the same:
When systems become sophisticated enough to guide decisions, what remains for individual judgment? And who determines when deviation was justified?
Sports and therapy seem like different worlds. One's a game; the other involves life and death. But both involve professionals trained within systems, operating according to protocols, sometimes facing moments where judgment says something the system doesn't capture.
The simulation trap and the therapy debate are the same problem. We're recognizing it now because systems have become explicit enough—in both domains—that we can finally see them clearly.