How the coach tests a question about your body

An investigation starts with a question in your own words - why am I so tired in the afternoons - and works by testing rather than guessing. This is how the first test is put together, so you can change it knowing what you are changing.

A hunch is a hypothesis

If you have an idea what it might be, the coach writes it down as a hypothesis with one more thing beside it: what would tell against it. A hunch that nothing could disprove is not a test, it is a belief. So coffee is behind my sleep is stored with my sleep stays the same for two weeks without coffee, and the test is built to produce one or the other.

No hunch is a fine answer. Then the days themselves are the test: the coach reads the better days against the worse ones for what differed.

The shortest test that could move the question

There are two shapes, and the setup picks by whether the factor can be left out.

• Leave it out. Coffee, alcohol, dairy, sugar, late screens - things you can go without for a while. For two weeks the factor is out and each evening you say a word on the thing you are investigating. If it eases, the hunch holds; if nothing changes, the factor is ruled out and the next one gets its turn.
• Log both sides. Sleep, stress, training, a medication - things you cannot remove. Nothing changes yet. Each evening you record the factor as it was and the thing you are investigating, and after two weeks the coach compares days with more of it against days with less.

Two weeks is the usual length because it gives enough good and bad days on both sides to compare. A week shows a strong effect and misses a subtle one. Three or four weeks is for something that only happens a few days a month, or moves slowly.

Why gluten is different

One factor the coach will not let you cut: gluten. Going gluten-free spoils the test that settles whether coeliac disease is part of the picture, and that test has to come first. If gluten is your hunch, the coach logs it rather than removing it, says why, and points you to your GP for the blood test. That refusal is more useful than the elimination would have been.

One test at a time

Only one test runs at a time. Two eliminations at once cannot tell you which one worked, and two logging tests double what you have to record each evening. If another test is already running, the question still opens and the new test waits for it to finish.

Where a doctor comes in

Logging and testing do not need a doctor. A diagnosis or a treatment does. The coach will not name a condition, and it reads every investigation against a list of signs that end it early and turn into a sentence about seeing someone this week: unintentional weight loss, bleeding, a new lump, fainting, breathlessness at rest, bloating that has not settled in three weeks, and a few more. That list is deliberately wide. A false alarm costs one investigation ending early; a miss costs more.

What it will not do

• It does not diagnose. What a test shows is the hunch held or the hunch was ruled out, never a name for what you have.
• It does not test a medication by having you stop it. A medication is logged and compared, and any change is a conversation with the person who prescribed it.
• It does not run a restrictive elimination - whole food groups, very low intake - as a test.
• It does not invent a hunch for you. Without one, it compares your days.

The question is yours to end at any time from its own page, and what each test showed stays on the file so the next one starts from what is already known.