What the Record Owes the Kid It Describes

Let's say we build it. Everything I've spent three essays describing turns out to be possible, and a twelve-year-old somewhere has a developmental record that actually works. It holds what she was trying to learn at every stage, what her coaches saw, what she said she was experiencing, what somebody tried, and whether it held up three tournaments later. It follows her when she changes coaches. It survives when her academy reorganizes. It's the thing I've been arguing junior tennis has never had.

Now imagine that record is very good, and ask a simple question about it. What is it rewarded for?

Because a system this capable will get extremely good at whatever we point it at, and nothing about building it well determines what it gets pointed at. If the thing it's measured on is revenue, it will become superb at knowing which families are ready to buy another tournament, and it will be right often enough to feel helpful. If the thing it's measured on is ranking movement, it will optimize for ranking, including in the stretches of a kid's development when ranking is the worst possible guide. Every system ends up rewarding something, whether or not anyone names it explicitly, and what it actually rewards is usually inherited from whatever was easiest to count.

Sometimes the Right Answer Is Less

Here's the test I'd apply, and it's uncomfortable because most of the industry would fail it.

A developmental system that's working should regularly recommend doing less. Fewer tournaments during a growth spurt. A month away from competition after a stretch that broke a kid's confidence. No lesson this week, because the problem isn't technical and another hour on court will just make everyone feel busy. A different environment entirely, because the one she's in is fine for someone else and wrong for her. Sometimes the correct recommendation is that a parent stop talking about tennis in the car for a while, which costs nothing and sells nothing and is occasionally the single most useful intervention available.

None of those is designed to maximize the next transaction. Several actively eliminate one. And that's the whole problem in one line: the developmentally correct answer and the commercially attractive answer come apart on a regular basis, and any system that can't say "less" is not a developmental system no matter what it calls itself.

I want to be fair here, because it's easy to make this sound like a morality play with investors as the villain. It isn't one. Junior tennis was very good at selling families more of everything long before private capital took an interest, and it managed that with no technology at all beyond a phone and a persuasive coach. The worry isn't that money corrupts something pure. The worry is that we're about to make a very old habit much more efficient, and efficiency has a way of hardening into architecture. Once a system is built around the assumption that more is the answer, "less" stops being a recommendation it can make.

A Note Is Not a Verdict

The deeper problem isn't what the record recommends. It's what the record says about a child, and how long it says it. Think about what actually goes into a developmental history. A coach writes down that a kid gets tight in third sets. At the time it's an observation, provisional, made in a specific week about a specific stretch of matches, and the coach who wrote it would tell you as much. But it doesn't stay provisional. It gets read by the next coach, who now watches for tightness and finds it, because people generally find what they're looking for. It gets read by the coach after that. Five years on, a kid who long ago stopped getting tight in third sets is still traveling with a sentence that says he does, and nobody remembers it was once just one person's read on one bad month.

That's the thing I keep turning over. We spent three essays arguing that tennis forgets too much. All of that is still true. But the fix for forgetting everything is not remembering everything forever, because a child is not a fixed object being measured. A child is a person in the middle of becoming someone, and the entire point of development is that who she is at twelve is not who she'll be at seventeen. A record that can't let go of the twelve-year-old is not preserving her history. It's holding her to it.

So a developmental record needs to forget on purpose, and forgetting on purpose is a design problem nobody in youth sports has taken seriously. Some observations should expire. Some should carry the context that produced them, so a later reader knows this was written the year she grew four inches, or the month she was playing up two age groups on purpose. Even that has a limit, though, and it's worth naming: the fact that an observation came during a growth spurt belongs in a developmental record, and the details of a family's private life almost certainly don't. Some should be contestable by the player herself, with her own account carried alongside the original rather than replacing it, because she's the only one who knows the inside of it and the only one who has to live with the description. And some things should never enter the record at all, which is a harder line to hold than it sounds when the cost of storing something is nearly zero.

Who Gets to Read It

A record built for one purpose changes character the moment someone reads it for another. Everything I've described is meant to help a coaching team make better decisions for a kid. Hand the same file to somebody making a selection decision and it becomes something else entirely. A college coach reading a six-year developmental history isn't looking for what to work on Monday. He's looking for reasons to say no, because that's what selection is, and a candid note about a kid's struggle with pressure at thirteen can be exactly the kind of thing that ends a recruiting conversation at seventeen. The observation was true and useful and written in good faith. Read four years later by a stranger with a different job, it does damage nobody intended.

The same goes for parents, and this is where I'll say something that will sound wrong. A parent does not need full access to everything in a developmental record. Not because parents can't be trusted, but because a good record contains provisional thinking, disagreements between coaches, hypotheses being tested and discarded. That material is useful to the people doing the work and corrosive to a parent trying to decide at eleven at night whether their kid is falling behind. Sometimes the most valuable thing a system can do for a parent is give them less, framed better, at the right moment. Every instinct in the technology business runs the other way, toward dashboards and transparency and more access as an obvious good, and I think that instinct will do real harm here if nobody argues with it. None of this is solved by a privacy policy. It's a question of design: who sees what, when, in what form, and with what obligation to the kid the whole thing describes.

Somebody Has to Be Able to Say No

If a system gets good at recommending things, the most important feature it can have is a human being with the standing to overrule it. That sounds obvious and it almost never survives contact with a real organization. Recommendations from a system holding six years of history carry weight. They arrive with an air of having considered everything. The coach who disagrees is now the person arguing against the data, and that's an uncomfortable position to occupy repeatedly, especially for a young coach, especially when the recommendation aligns with what the business would prefer anyway. Dissent has to be designed in and protected, or it quietly disappears while everyone insists the humans are still in charge.

This is where I keep landing on the difference between owning something and taking care of it. Ownership settles who controls an asset and who collects what it produces. It's a real question and the article that started this series is right to take it seriously. But ownership doesn't answer for consequences. A perfectly clear ownership structure can sit on top of a system that's slowly narrowing what a generation of kids is allowed to become, and every contract in the building can be in order while it happens. Stewardship is the other question, the one about who answers for what the thing does to the people inside it, and it's a question youth sports has far too little vocabulary for when the subject turns to data, technology, and developmental systems.

Tennis Is Just Where I Can See It

I've written all four of these essays about tennis because tennis is what I know and because the pipeline is unusually visible there. A junior player is identifiable across a decade, competes constantly, generates results at every step, and moves through a chain of coaches and programs that each hold a piece of her. If you want to watch this problem develop in a controlled setting, junior tennis is about as clear a laboratory as exists.

But nothing I've argued is really about tennis. The same architecture is being assembled around kids in music, in academics, in every youth activity organized enough to keep records. The question of what a developmental system should remember, how long it should hold it, who gets to read it, and whether it can recommend restraint is going to be answered in a dozen places over the next decade, mostly by default, mostly by whoever builds first. Tennis just happens to be where I can see it clearly enough to say something useful.

And underneath all of it is an older problem that has nothing to do with capital or software. One of the most powerful things in developing a person has always been sustained individual attention: somebody who watches closely, knows this particular kid, remembers what happened last spring, and adjusts. It works about as reliably as anything we know of, and it has always been rationed, because one person can only hold so many people in mind at once and there have never been enough hours or enough adults who do it well. Every system we've built since has been a compromise with that limit. What might be changing isn't that a machine can do the noticing. It's that the person doing the noticing could finally carry far more context into every decision than memory alone ever allowed, which is a smaller claim than it sounds and a more consequential one than anything in this series about markets.

Which is why the thing being built matters so much more than who ends up owning it. The infrastructure is coming either way. The question that stays open is what we decide it's for, and whether anybody insists on answering that before the economics answer it for us.

We know what happened to these kids. We're getting better at preserving what it meant. What we still owe them is a system designed to keep a record from becoming a verdict.

*This is the fourth and final essay in a series that began with a question Lisa Stone put to the ParentingAces community after sharing a Sports Brief article: what are the implications of this for junior tennis?*

Written with Reggie, Claude, Gemini, Otter, and 64+ trips around the sun.