What the Body Reports

Your body is always talking. The trouble is that most technology reads one sentence and calls it the whole conversation.

When I was ten, I had a stroke. I am not going to tell that story here, because it belongs to a longer piece. What matters for this one is what came after. My left side stopped reporting to me the way my right side did. It was slower, quieter, and sometimes it said nothing at all.

I rowed at Cal with that body. Rowing is a sport where the two halves of you have to agree, thousands of times an hour, at a level of precision that makes most people's shoulders hurt just watching. I learned very early that the body does not speak in one voice. I also learned that it speaks whether you are listening or not, and that it gets louder if you ignore it. Mine has a temper.

You do not need to have rowed to know this. Anyone who has trained hard, been injured or been ill has met a body that speaks in more than one voice.

I learned that the absence of a signal is not the absence of a condition. Sometimes the instrument is incomplete. Sometimes the language is. And sometimes the person living inside the body knows something the measurement has not yet learned to recognize.

This Inquiry is about the second module in the SportsFlow battery, called Readiness. It is the module that listens to the body directly: heart rate variability, resting heart rate, sleep, breathing, a map of where you are sore, and a few other physiological readings, pulled in from the wearables people already own.

It is also the module where most of the technology in this field goes wrong.

The problem with the morning number

Most people with a fitness watch have had this experience. You wake up feeling good. The watch says your readiness is 41. You feel worse immediately. Or you wake up feeling like a car accident and the watch says 88, so you go hard, and by the second interval you know the watch was lying.

The watch was not exactly lying. It was answering a question badly. The most popular readiness signal is heart rate variability, usually shortened to HRV. It is the small variation in the time between heartbeats, and it reflects how the calming, parasympathetic side of your nervous system is acting on your heart, one part of how the body recovers. It is a real signal. It is also a noisy one. A single morning reading moves with sleep, alcohol, a late meal, a fight with your partner, a full bladder, and whether you sat up too fast.

It is not a direct measure of resilience, emotional health or readiness to perform. And what it can predict for one person, such as how they will perform on a given day, is far less certain than what it measures. Trends in HRV have been linked to how athletes adapt across weeks of training. A single morning reading predicting a single session is a much weaker claim, and one I would not make.

The researchers who have worked hardest on this, Daniel Plews, Paul Laursen and Martin Buchheit among them, arrived at a plain answer: stop reading the morning, read the trend. Average the daily readings over a rolling week. Compare that average to the person's own normal variation. Treat a change smaller than that variation as noise. Plews and colleagues also found that you do not need a perfect streak. Three or four good readings a week gave results close to a full seven.

That is how Readiness works. Every physiological reading is compared to the same person's own rolling 28-day baseline. There is no chart that says what a 45-year-old woman's HRV should be. There is only you, last month, and you, this week.

A history before a vote

One rule in this module sounds fussy and is one of the most important. For the first ten days a person uses it, the body gets no vote at all.

Without a history, there is no baseline. Without a baseline, a physiological number means almost nothing. A resting heart rate of 58 is excellent for one person and a warning for another. So for the first ten days, readings are scored entirely from what the person reports, and the card says so in plain words: self-report only.

Ten days is a design choice, not a validated threshold. A longer history builds a better baseline, and whether ten days is enough is something that has to be tested against real-world data.

After ten days, constructs that have a defensible link to the body start to blend. The formula is simple. Seventy percent from what the person says, thirty percent from what the body shows. Self-report keeps the larger share on purpose, for the reason set out in No. 01.

Some constructs never blend. Self-kindness, gratitude, the sense of being part of a group. There is no honest wearable proxy for any of them, and pretending there is would mean guessing at someone's inner life from their pulse.

Same signal, different question

In short: the same number can mean different things depending on the question you ask of it.

Here is where it gets interesting, and where I think most products miss something.

A single HRV trend can answer at least three different questions. Am I recovered for today? Am I adapting to this training block? Am I under more strain than I can carry? Each question has a different picture of what good looks like.

Plews and his colleagues described something that complicates the simple story. In some very fit athletes, HRV can actually drop while resting heart rate also drops, a pattern known as saturation. A naive readiness score reads that as a bad sign. It can reflect very high vagal activity that the HRV measure no longer registers, rather than fatigue. The lesson is that a body signal does not have one meaning. It has a meaning relative to a question and relative to a person.

So in SportsFlow the same signal is scored against a different healthy range depending on which question it is helping to answer. It sounds like a small engineering detail. It is the difference between a number and an interpretation.

Two witnesses

The part of this module I am proudest of is not a measurement. It is a rule about what to do when the person and their body disagree.

Most systems handle this by picking a winner. Usually the machine wins. You say you feel great, your watch says you are wrecked, and the app gently suggests you might be wrong about your own life. I find that insulting, and the evidence does not support it either. The body and the mind are both witnesses. Neither one is the judge.

So Readiness treats them as two witnesses. If they agree within eight points, the reading is marked corroborated. If they disagree, nothing is said the first time, or the second. It takes three disagreeing windows in a row before anything surfaces. Then both are shown side by side, and the person decides what it means.

The eight points and the three windows are design decisions, not established thresholds. They were chosen to keep noise from becoming news. Whether they are the right numbers is one of the first things real data will need to answer.

Disagreement has more than one possible cause, and the wrist is not always the reliable party. A heart rate variability reading can move with alcohol the night before, a late meal, a short or badly timed sleep, the first day of a cold, or simply how and when the measurement was taken. Different devices also estimate it with different methods, so two watches can tell different stories about the same night. On the other side, a person's own rating is shaped by mood, expectation and what they think they ought to say. So when the two disagree, the module does not ask which one is right. It asks what could explain the gap. The first things to rule out are the dull ones: a poor night, a different device, an unusual evening. If a gap survives those, it is worth a conversation, because the signals cannot settle it on their own.

Sometimes the person is ahead of the body. They know they are carrying something that has not shown up in their sleep yet. The body may also be ahead of the person: the nervous system shows strain the mind has not yet admitted. Both are useful to know. Neither is a moral failure.

There is one pattern I want to name because it is easy to misread. A person starts reporting harder numbers about themselves while their body is clearly getting better. On paper it looks like decline. It may be the opposite. The standard a person uses to rate themselves changes as they pay closer attention, so the same state earns a harsher score. Researchers call one version of this response shift. The module is designed to flag this as possible sharpening rather than decline. Whether it can reliably tell the two apart is something only real data will show.

The body and the mind are both witnesses. Neither one gets to be the judge.

Guess before you look

There is a small sequence in the app that I think matters more than its size suggests. When a person checks in, they answer how they feel before they see any wearable number. The app does not show you your HRV and then ask how you feel about it. It asks first, then shows.

The gap between your guess and the reading is tracked as its own result. It is a rough, unvalidated proxy for part of what researchers call interoception, the ability to sense what is happening inside your own body. There is early evidence that interoception can be trained: in one nine-month meditation study, people's accuracy at sensing their own heartbeat rose steadily. There is some evidence that people who are good at it regulate their emotions better, though the research is still mixed. Most of us are worse at it than we think.

Guess-then-check turns every morning into a small practice. The hope is that, over weeks, the gap closes. The watch becomes a teacher you are slowly outgrowing, which I think is the right relationship to have with any piece of technology.

What the body is not

I have spent a long time around people who were sick, injured or recovering, and I have a healthy distrust of any device that claims to know how someone is. So Readiness has its refusals, like every module.

The coach's morning word draws on these signals, but the signals themselves stay with the person. A coach sees raw physiology only if it is explicitly granted, and a grant on a score does not carry the body data with it.

Readiness is not a diagnosis. A drop in HRV is not illness, and a good number is not health. It is a nervous system's report on the last few days.

And nothing in this module is used to decide who makes a team. A body that is having a hard week is not a worse athlete. It is a body having a hard week.

Where the morning word comes from

In short: the coach's morning word is built from your readings, but never shows them.

Readiness is also where the one-word morning call for coaches gets built: Ready, Hold or Back off. That word is the only thing a coach sees by default, so it is worth being precise about what goes into it.

It draws on the morning check-in, the physiological trend against the person's own baseline, and a slower reading called load tolerance. Load tolerance looks back over twelve weeks and asks how well this person has been absorbing the work they have been given. The thinking is simple even if the tuning is not finished: a rough night should be able to move the word toward Hold, and only a sustained slide in tolerance should move it to Back off. One bad morning is weather. Twelve weeks is climate.

The word also carries one line of action, written in plain language. Not a diagnosis. Something like: lighter session today, check in again tomorrow. And the athlete sees exactly the same word and line their coach sees. No second, private version for the adults.

What the wrist cannot see

In short: trust the trend on one device, not the exact number.

I want to say something unfashionable about wearables. They are remarkable, and they are not as precise as their screens make them look.

Different devices measure heart rate variability in different ways and at different times. A chest strap at rest in the morning and a ring averaging across the night are not measuring the same moment. Sleep stages on a consumer device are estimates built from movement and heart rate, not the brain recordings a sleep lab uses. None of this makes the devices useless. It means a baseline only means something on the device that built it.

So the practical advice is the same whether you use SportsFlow or not. Pick one device and stick with it. Measure at the same time each day. If you change devices, treat your baseline as starting over, because in a real sense it is. That is annoying. It is also honest.

My left side still reports late. Some mornings it doesn't report at all. I have stopped expecting it to sound like my right. What I have learned, over forty-odd years of living with a body that speaks with a stammer, is that the answer was never to trust it more or trust it less. It was to listen longer, and to stop treating one morning as the whole story.

The body is not a machine that produces answers. It is a living system offering evidence. Learning to listen means knowing what that evidence can tell us, and what it cannot.

Orion