Sleep — duration, and the regularity people forget

Sleep is where the adaptation you trained for actually happens: growth-hormone release, tissue repair and the consolidation of motor learning all concentrate in deep sleep. It is the highest-leverage recovery input you control, and the one most often traded away.

Duration matters, but it is not the whole measure. Two runners can both average 7.5 hours and be in completely different states: one sleeping 7 and 8 hours alternately, the other 5 and 10. Regularity of sleep timing and duration has its own effect on recovery and performance, independent of the total. That is why Runsei reports not only your average but how much your nights vary — and why a large spread is worth naming even when the average looks fine.

How Runsei reads it. Your recent sleep is compared with your own normal — a 30-day median excluding the most recent week — rather than with a 7-or-8-hour ideal. A naturally short sleeper should not be told they are under-recovered every single day, and a runner whose nights have quietly shortened over a fortnight should hear about it even if they are still "getting seven hours".

One bad night is not a crisis. A single short night costs less than a fortnight of moderate restriction, which accumulates quietly and shows up as flattened HRV, a raised resting HR and sessions that feel harder than the numbers say they should. If sleep is the thing that has slipped, it is usually the most productive thing to fix before adding training.