Signal
← All posts

2026-05-06

What Your Recovery Data Is Actually Telling You (And What It's Not)

Wearables have made it easy to track recovery. They've also made it easy to misread it. Here's what the numbers mean and when to ignore them.

i started paying attention to my recovery score about two years ago. for the first few months i used it like a prescription. green means go, red means rest, anything in between means question everything you're doing.

that was probably the wrong approach. not because the data isn't real, but because i was treating a single number as a complete picture when it's more like one data point in a system that's a lot more complicated than any wearable can fully capture.

what recovery scores are actually measuring

most consumer wearables estimate recovery using heart rate variability, specifically resting HRV, usually measured overnight or first thing in the morning. HRV is the variation in time between heartbeats. higher variability generally indicates your autonomic nervous system is in a more parasympathetic (rest-and-digest) state, which correlates with better recovery. lower variability tends to mean the opposite.

the thing is, HRV is sensitive to a lot of things that aren't training. alcohol is the most obvious one. even a drink or two will suppress HRV noticeably. sleep quality matters more than duration. stress, travel, caffeine timing, hydration. all of it moves the number. your recovery score is a readout of your overall physiological state, not just how your legs feel from yesterday's run.

this is actually useful information. but it means a low score doesn't necessarily mean your training load was too high. it might mean you had two beers and watched tv until midnight.

when to trust it

the recovery score is most useful as a trend. a single data point doesn't tell you much. but if your scores have been suppressed for a week, that's worth taking seriously. you might be accumulating fatigue faster than you're recovering from it, and backing off makes sense.

it's also useful for catching things before they feel bad. there's often a lag between physiological stress and subjective perception of fatigue. your body knows you're depleted before your legs do. if your scores are tanking and you still feel okay on runs, that's usually not a green light. it's a warning before the wall.

the best athletes i know use recovery data as a prompt for a question, not an answer. "my score is low. what's actually going on today?" sometimes the answer is "nothing, yesterday's run was hard and i need to take it easy." sometimes it's "i slept terribly and drank too much, the training is fine." different responses, same number.

when to ignore it

context matters a lot here. during a hard training block, consistently lower scores are expected. you're accumulating load on purpose. if you're tapering before a race, you should see scores come up. if they don't, that's worth thinking about.

but if you're in week three of a build phase and your recovery score is orange every morning, that's not necessarily a problem. that might just be what a training block looks like on your body. the goal isn't to have a green score every day. it's to be building fitness faster than fatigue, and to recover fully when it matters.

i've had some of my best races after weeks of mediocre recovery scores. and i've had flat, dead-feeling workouts on days where every metric said i was ready to go.

the thing wearables can't measure

what doesn't show up in the data: motivation, mental freshness, life stress that isn't reflected in HRV, that feeling you get in your legs on some runs where everything just clicks. all of that is real and it matters.

recovery scoring is a tool for people who want to make better training decisions with more information. it's genuinely useful. but it's not smarter than you are. it's an input, not a verdict.

the runners who use it well treat it like one voice in the room, not the only one

Signal is an AI running coach for athletes training toward a goal.

Join the waitlist