I have been keeping a spreadsheet through chemo. Mood, energy, nausea, rash, every day, scored zero to ten. WHOOP on my wrist when I remembered to charge it. A short paragraph at the bottom of each day saying what I ate, what I drank, what hurt, what helped. Three months in, I have seventy rows. That is enough to start seeing things.

What I expected was a long, slow downward curve. Chemo gets worse, you get worse, the line points south, you ride it out and hope the bloods catch up. That is not what the data did. It did something more interesting and a lot more useful.

The first thing the spreadsheet shows is that the bad days are weirdly predictable. Cycle-start Tuesdays look identical on the chart: resting heart rate climbs, HRV collapses, recovery drops below sixty percent. The body knows the bag is going up before I have finished the sandwich at the chemo unit. The next morning is the lowest WHOOP number of the fortnight, every single time, in five out of five cycles I have completed so far. I can plan a calendar around that signature. I have stopped trying to do anything useful on Wednesdays.

The second thing is more annoying. Alcohol absolutely smashes my recovery, and not by a sensible amount. Saturday eighteenth of April, I went out for beers and a shisha and felt fantastic. Mood ten. Energy ten. Recovery ninety-one percent. The next morning my recovery was eight percent. Eight. The largest one-day fall in the entire log by a country mile. Every other alcohol day in the spreadsheet has the same shape, just less dramatic. The data does not say stop drinking. The data says know what it costs.

The third pattern is GI. Fatty food triggers diarrhoea inside eight hours, every time, with the consistency of a vending machine that only does one thing. Big roast on day six of cycle one. Cinema fries and a milkshake on the twenty-fifth of March. Soy sauce on rice in early April. Pizza and a pint. Same outcome, same window. Once you can see the pattern on a sheet you stop blaming the chemo and start blaming the cheese. I keep Imodium in three different bags now.

The fourth thing is the one I did not expect. There are days in this log where WHOOP is in the red and I have logged a mood of nine or ten. Wednesday the eleventh of March: recovery twenty-seven percent, mood nine. Tuesday the seventeenth of March, an infusion day: recovery sixteen percent, mood nine. There is a real story in that gap. Subjective wellbeing and physiological recovery are two different signals. Both are real. Neither cancels the other. On a bad-looking biometric day I can still feel great. On a green-light WHOOP day I can still feel ropey. The spreadsheet stops me from over-trusting either.

Mostly though the log just makes the months legible. Cycle four was nearly entirely fine and the bloods were dropping. Cycle five hit harder because cetuximab is a different animal. May the eighth was the worst day of the period: stayed in bed, mood three out of ten, suspected I was blocked up. Two days later I was back to an eight. Without the log I would remember May the eighth as the moment things turned. With it I can see it was one day inside a cycle that bounced back. That is not nothing.

I am publishing the lot. Every row, every WHOOP score, every embarrassing pub trip. There is one anonymised line on the twenty-third of April because the original entry named a family member in a way that was not their fault and not for me to publish. Everything else is in. The pints. The shisha. The melatonin restarts. The temazepam. The Crooked Club. The constipation arc. The eight percent recovery score that I earned fair and square.

If you are a patient and you want to start logging, log everything. Not just the polite stuff. Mood, food, drink, sleep, who you spoke to, what you ate before the symptom appeared. If you give yourself an out for what counts as real data you will lose the patterns that actually matter, which are almost always behavioural and almost never the ones the clinic asks about. If you are a researcher and you can do something with three months of well-labelled patient data, the CSV is one click away. The XLSX is one click away. The cycle summary is one click away.

The full chemo log lives on the open-source-me page. Cross-referenced with my bloods, my scans, my genomic profile, my actual treatment protocol. If you find a pattern in there I missed, tell me.