Sleep scores are a coach's problem, not an athlete's

The data on athlete sleep is good enough to change training schedules and nowhere near good enough to hand over each morning.

Two athletes, same squad, both sleeping badly. One is sleeping badly because the session is at half past six and the drive is forty minutes. The other is sleeping badly because he is nineteen and lives with four other nineteen year olds. Only one of those is a training problem, and the wearable cannot tell them apart.

That distinction is the whole argument. Sleep telemetry is genuinely useful at the level of the programme. If a squad's average sleep duration falls by forty minutes in the fortnight after a schedule change, that is an actionable finding and the fix is scheduling, not education. If the athletes who travel furthest are consistently short, you have learned something about your training venue. These are structural questions, they show up clearly in aggregate, and aggregate is where consumer-grade sensors are least wrong.

At the individual daily level the same devices are much weaker. Stage detection from wrist motion and pulse is an inference, not a measurement, and it degrades exactly where you would want it to hold up, which is on nights that were unusual. A restless night with normal heart rate can score well. A dead-still night with a slightly raised pulse can score badly.

And yet the daily figure is the one athletes see, because that is the product.

I think handing it over unfiltered does harm. There is decent evidence that expectation shapes next-day performance independently of what the sleep actually was; tell someone they slept poorly and they will produce a worse effort than the same person told nothing. A readiness score is therefore not a neutral observation. It is an intervention, delivered at six in the morning, by a company with no stake in the outcome.

The other thing coaches underrate is that sleep is the most fixable variable they have and they almost never touch it. Nobody wants to move the session. Everyone wants to buy another ring.

If I were running a group tomorrow I would collect the data, look at it weekly, never discuss individual nights, and spend the political capital on shifting the morning start by an hour. That single change usually beats every recovery product in the building.

The athletes will complain that the later start ruins their day. Some of them will be right. Fewer than you think.