Where the Numbers Disagree
Seven honest measurements of a person, read together. The average can tell us something useful. But sometimes the most revealing information lies in the places where the measurements disagree.
Anyone who has spent time on the ocean learns that no single sign tells you what the weather is going to do. A falling barometer on a still morning means one thing. A falling barometer with a swell building from the southwest means another. The sky can look perfect an hour before it turns. You read the signs together, and the ones that disagree are the ones that matter.
I grew up on the water and I have spent a working life around human beings, and I have come to think they work the same way. One reading means very little. A pattern of readings means more. And the most important information is almost always in the places where the readings disagree.
This Inquiry is about the layer of SportsFlow that sits on top of the seven state modules. It has no questions of its own. It produces no score. Its whole job is to read the other seven together. I call it the grammar.
Layer brief · The grammar technical brief
- Where it sits
- The integration layer. Not a scored module, and it asks no questions.
- What it measures
- Where the seven state modules disagree for one person, or one group.
- Why it matters
- A single score hides the most useful information about a person: the places where different parts of their life are moving in different directions. Two people with the same average can need opposite things. Reading the disagreements is what turns many measurements into one useful question.
- How it measures
- Twelve divergence readings, each pairing two states with what it means, what to do and what it costs if missed. A map of which states tend to move first. Rules for which reading wins. A list of things that must never be combined. No single score.
- What the science supports
- The constructs inside each module, and the practice of reading change against a person's own baseline. The research supports the parts. It does not validate this way of combining them.
- What is SportsFlow's own
- The whole layer: the readings, the lag order, the precedence rules and the never-combine list. All hypotheses, to be tested and revised as real data arrives.
- How athletes and coaches use it
- The coach gets a sentence or two about where states disagree, as a place to start a conversation rather than a conclusion. The athlete sees it first, in the same words.
- Limits
- A sequence is not a cause. A reading can fire on the wrong thing. It says where to look, not what is wrong.
- Why we built it
- Because conventional monitoring collapses a person into one score, and the most useful information is often in the disagreements a score erases.
Seven states
Seven states and the question each one answers
| State | The question | Instruments |
|---|---|---|
| Wellbeing | How is this person holding up? | 10 |
| Readiness | What is the body reporting? | 9 |
| Flow | How is the work going? | 3 |
| Connection | Who holds them? | 9 |
| Meaning | Why is it worth doing? | 6 |
| The Setback | What happens when it is interrupted? | 5 |
| The Coach | How is the person they are all responding to? | 6 |
The earlier Inquiries in this series each took one state. Wellbeing, Readiness, Flow, Connection, Meaning, The Setback and The Coach. Each module, on its own, is a useful instrument. Each one was built on the same premise that sits under the whole battery: a single measurement means little, and a combined set describes a state.
The grammar applies the same premise one level up. A state means something. Seven states read together mean more. But only if you read them the right way.
Imagine an athlete whose physical readiness is improving. Sleep has steadied. Resting heart rate is normal. Training feels manageable. But their attention in sessions has been slipping for weeks, and the reasons they once gave for competing have started to fade.
No single measurement tells that story. Physically, the athlete may be ready to train. Psychologically, something important may be changing. The disagreement does not tell us what is wrong. It tells us where a conversation might begin.
Why there is no single score
The first thing every club asks for is one number. Give me a single score for each athlete, ranked, so I can see who needs attention. It is a reasonable request. It is also the one output this battery will never produce.
Two people, the same average
Look at the two people in that figure. Both average 60. A single battery score would call them the same. Person A is steady across the board. Person B has a body that is ready to go and attention that is not there. They are deeply connected to the group and losing their reasons for being in it. Those are two problems, with two different fixes, and they are both completely invisible in the average.
That is the heart of the argument. An average can be useful. But the one thing seven states hold that one instrument does not is the disagreement between them. Averaging is precisely the operation that destroys it. A club that wants one number is asking for the one thing that cannot be honestly produced.
The only thing seven states know that one instrument doesn't is where they disagree. Averaging is the operation that deletes it.
Reading the disagreements
The grammar is a set of readings. Each one names two states, the direction each is moving, what it means, what to do, and what it costs if nobody notices. Here are six of them.
Six of the readings, and what each one asks a coach to do
| When these disagree | It means | Do this | Cost of missing it |
|---|---|---|---|
| Readiness high, Flow low | The body is ready and the attention is not | Look at conditions, not load. More rest will not touch it | Weeks of good work that produce nothing |
| Meaning falling, Wellbeing holding | The reasons are going before the person feels it | Talk now, while it is still a conversation | The person who quits in October having been fine in June |
| Meaning high, Wellbeing low | Paying more than the work is worth, and it does not feel like a cost | Read how the thing is held before anything else | Injury, illness, and a hard stop instead of a soft one |
| Connection high, Meaning low | Here for the people, not the work | Nothing today. Do not assume it lasts | A sudden exit when the group changes |
| Setback active, Connection falling | The club stopped calling | Somebody phones | An athlete who does not come back, and never says why |
| Coach load high, group falling | A program problem, not eight athlete problems | Do not fix the athletes. Look at the coach's load | Eight interventions aimed at the wrong person |
Three of those deserve a word.
Available and absent. Readiness is high and flow is low. The body is ready, and the attention is somewhere else. The natural move is to adjust the training, add rest, back off. It may not help, because load may not be the problem. The conditions might be. Look at the conditions of the work before the amount of it.
The early warning. Meaning is falling while wellbeing holds. Of all the readings, this is the one I care about most. The reasons often seem to go before the feelings do. That is the battery's working assumption, and some motivation research points the same way. A person in this pattern will honestly tell you they are fine, and they are, for now. Caught here, it costs a walk and a question. Caught later, it can cost a season.
The dangerous one. Meaning is high and wellbeing is low. A person is paying more than the work is worth, and it does not feel like a cost. When meaning is held rigidly, it can hide damage rather than prevent it. Nobody stops something that feels like a calling. The first thing to read here is how the person holds the thing they love, which the Meaning module reports separately as Held hard.
And one reading exists purely to say nothing. When two states agree, there is no puzzle to solve. The reading is just the reading. Most of the battery agrees most of the time, and the grammar stays quiet. I think that is one of its most important features. A system that produces a finding every day trains people to ignore findings.
An observation is not an explanation
Every reading in the grammar has two layers, and it matters to keep them apart. The first layer is observed: two states are moving in different directions. That part is data. The second layer is any explanation of why. That part is a guess, and the grammar is built never to make it on the person's behalf.
What the battery observes, and what it does not know
| What it observes | A tempting explanation | What it actually says |
|---|---|---|
| Readiness high, Flow low | They are not trying | The body and the attention disagree. Ask about conditions |
| Meaning falling, Wellbeing holding | They are losing interest | The reasons and the mood are moving differently. Ask what has changed |
| Connection falling during a setback | They have checked out | Contact is dropping while they are out. Call |
| Coach load high, group connection falling | The athletes are the problem | Two lines are moving together. Look at the program first |
What moves first
In short: some parts of a person change before others, and knowing the order keeps one problem from being counted twice.
Readings also have timing. Some states move fast. Some move slowly. Some move first and drag others behind them.
What moves first
| State | Timing | In plain words |
|---|---|---|
| Meaning | Leads by weeks to months | The reasons go first |
| The Coach | Leads by weeks | Coach capacity moves before the group's state |
| Connection | Slow, moves in blocks | Ties take months to form or fail |
| Wellbeing | Follows meaning and load | The person feels it after |
| Readiness | Follows load by days | The body reports quickly |
| Flow | Fast and noisy | Session to session. Needs many reports before it means anything |
| The Setback | Steps, not slopes | It changes at events, not gradually |
This lag map matters for a practical reason. Without it, the system would report the same event twice under two names. Meaning falls in March. Wellbeing follows in May. A naive system reports a meaning problem in March and a wellbeing problem in May, and the coach treats the second as new news. With the lag map, it reports one finding, the meaning one, and notes that wellbeing is following it.
It also names which reading is the lead and which is the echo: when a coach's capacity falls and the group's connection follows, the finding is the coach. But a change that appears first does not prove it caused the next one. The sequence gives a hypothesis worth examining, not a diagnosis. The lag order itself is a design assumption, drawn from the burnout, motivation and coaching research and from years of watching people. It is the first thing real data should test, and it will probably need adjusting.
Rules of order
In short: when readings conflict, the person comes first, the pattern comes before the number, and the system asks before it concludes.
When two readings point in different directions, something has to decide which one wins. Underneath the technical rules are three principles.
Read the person before the population. A person's own baseline is the only yardstick. Nobody is ranked against a squad or a population.
Read the pattern before the number. One observation rarely establishes a trend. Most readings need a disagreement held across more than one window before anything is said.
Ask before concluding. A system can find a discrepancy without knowing what it means. The person usually knows.
Behind those principles sit a few technical rules. The most important: while a setback is active, flow and readiness are shown as suspended, never as low, because a returning athlete reading their own low readiness as failure is an injury the product would have caused.
One person, one block, two findings
What must never be joined
In short: some numbers are honest on their own and harmful when combined.
Here is something I did not expect when we started connecting the modules. Joining two honest measurements can produce something dishonest. Each module had its own protections. Connected, they could be combined in ways that broke those protections without anyone intending it.
So the grammar carries a list of things that must never be combined. There is no single battery score, for the reasons above. Setback trust and physical clearance are never combined, because a combined number becomes a gate within a season. An athlete's account of their coach and the coach's account of that athlete are never crossed to produce a view that says this pair has a problem. Both halves exist, and they do not meet. Nothing in the battery is ever a selection input. And coach readings are never an employment input.
Every one of those rules exists only because the modules can now see each other. That is the cost of reading them together, and it is worth paying as long as someone writes the limits down first.
A conversation, not a conclusion
Here is how the example at the start of this piece might actually play out. It is a composite, not a real athlete.
The grammar fires two readings for her this block. Her body is ready and her attention is not. Her reasons are thinning while her mood still holds. She has shared her readiness, flow and meaning readings with her coach, and nothing else.
A coach reading only the numbers might conclude she is losing interest and needs a push. A coach reading the grammar knows only that two things disagree, and that the disagreement is a reason to ask. So the coach asks, on a walk after practice, nothing formal: how is it going, really? Not the training. You.
It turns out her closest friend in the group moved away in the spring. Practice still works. The person she trained beside is gone. She has not said so because it sounds small. It is not small. The coach does not try to fix it. She pairs the athlete with a new training partner for the long sessions, and checks back in a month.
From three readings to one conversation
| Step | What the coach sees or does |
|---|---|
| Readiness | High and steady. The body is ready to train |
| Flow | Low across eight session reports. Attention is somewhere else |
| Meaning | Falling for five blocks, while wellbeing holds |
| What the grammar says | Two disagreements. A place to look, not a cause |
| What a numbers-only coach might do | Push harder, or back off the load. Neither touches it |
| What this coach does | Asks one open question, away from training |
| What she learns | Her closest training partner left in the spring |
| What changes | A new partner for the long sessions. A check-in a month later |
Nothing in the data said "her friend left." The data only said where to look. The rest took a person.
It is worth hearing the difference in the coach's actual words, because that is where a reading either helps or harms. With a diagnosis in hand, the coach might say: "Your focus is down and your motivation is slipping. What's going on with you?" That sounds like a charge, and the athlete is likely to defend herself. With only the disagreement in hand, the coach says: "Your body looks ready this week, and I get the sense your head is somewhere else. Is that fair?" The second version says what the coach noticed, admits it could be wrong, and leaves the explanation to her. If she says, "Not really, I'm fine," the coach drops it and the readings stay in the background. The grammar is designed to support conversations like the second one. A reading is meant to start a conversation, never to finish one.
How it could be wrong
Here is how this could fail. Any system that reads people has a way to go wrong.
It could fire too often. A grammar that finds something every week becomes wallpaper, which is why it stays quiet when the states agree.
It could fire on the wrong thing. The readings are built from research and from experience, not from a large data set of SportsFlow users, because that data set does not exist yet. Some of the twelve will turn out to be less useful than they look. Some patterns we have not written down will turn out to matter more. The grammar is a set of hypotheses with good reasons behind them. It should be treated that way, and revised as the evidence comes in.
The most dangerous error may not be a bad number. It may be a plausible story built from incomplete information. Human beings are very good at making patterns meaningful. An intelligent system has to be just as good at admitting when the evidence does not justify the story.
And it could be read as a verdict. "The dangerous one" sounds like a judgment about a person. It is not. It is a description of a pattern, with a suggestion about what tends to help. Every reading in the grammar ends with what to do, not with what is wrong with someone.
What the person sees
When something fires, the output is a sentence or two in plain words, never a new number. It goes to the person first. Most of the examples in this piece are written from a coach's side, but the grammar is just as useful turned inward. A person who can see that their reasons are thinning while their mood still holds may have weeks or months of warning that most people never get. A person who can see that their body is ready and their attention is not can stop blaming their legs.
The readings go to the person first, in the same words. If a coach sees anything at all, it is only what the person has chosen to share. Modules the person has not shared arrive in the coach's view as a name and a status, never as a number, so the grammar cannot leak through the back door what the front door kept private.
Where this stands: the structure described here is designed and built. The item wording, thresholds and patterns are still being tested, and none of it is yet a validated finding.
The grammar
The battery layer reads all seven state modules together. It holds the divergence readings, the lag map, the rules of order, and the list of things that must never be combined. It adds no questions and no score. It is not in the questions. It is in how the answers are read.
The numbers do not have to agree to be useful. They have to be honest about what they know, and humble about what they do not.
Orion
Sources
- Saw AE, Main LC, Gastin PB. Monitoring the athlete training response: subjective self-reported measures trump commonly used objective measures: a systematic review. British Journal of Sports Medicine 50(5):281-291, 2016.
- Curran T, Appleton PR, Hill AP, Hall HK. Passion and burnout in elite junior soccer players: the mediating role of self-determined motivation. Psychology of Sport and Exercise 12(6):655-661, 2011.
- Lonsdale C, Hodge K. Temporal ordering of motivational quality and athlete burnout in elite sport. Medicine & Science in Sports & Exercise 43(5):913-921, 2011.
- Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annual Review of Psychology 52:397-422, 2001.
- Brewer BW, Van Raalte JL, Linder DE. Athletic identity: Hercules' muscles or Achilles heel? International Journal of Sport Psychology 24(2):237-254, 1993.
- SportsFlow emotional and physical assessment battery, grammar layer documentation, CurioCore, 2026.