ClimbWatts

An open training log · 98 → 86.95 kg · 240 W and climbing

A real training plan, published as it is ridden. Built around FTP and power zones, fitted around rotating shift work. Climbing watts, one week at a time.

FTP
240 W
Weight
86.95 kg
W/kg
2.76
To 77 kg
10.0 kg

The rules: how I decide what to ride

A training plan tells you what to ride on Tuesday. That is useless when Tuesday is a night shift. What follows is the thing I use instead: four layers of rules that take a roster and a few numbers off a watch, and give back one decision — what to ride today. They are written down because a rule you keep in your head is a rule you renegotiate at six in the morning.

The principle: a rule you break is a broken rule

My first version said Body Battery under 65, do not ride. I broke it repeatedly, because somebody who wants to ride will ride. The rule was not too soft or too hard — it was the wrong shape.

Every rule here lets you ride. What changes is what you ride. There is no instruction to stay off the bike for a number on a screen, so there is nothing to argue with and nothing to cheat. The only exception is illness, and that one is at the bottom.

Layer 1: the roster decides what fits

Before any measurement, the shift decides the size of the day. This is the part every normal training plan gets wrong, because it assumes every Tuesday is the same Tuesday.

The dayWindowWhat fits
Day offanytimeAnything. Quality, the long ride, both if the week allows.
Afternoon shift
13:30–22:00
08:00–10:30 The best quality window in the whole roster. A full night of sleep behind you and a free morning in front of you.
First night of a seriesdaytime, work at 21:30 Counts as a day off — but see Layer 3, because what comes next changes it.
Early shift
05:30–14:00
17:00–19:00 Quality is possible but keep it under 75 minutes. Not where the long ride goes. You have been awake since 04:15.
Night, second onwards16:00–17:00 Endurance only, 90 minutes maximum, and only after the daytime sleep actually happened.
Day after the last nightafternoon Endurance only. No intensity — but riding easy is fine and helps put the body clock back.
12-hour shift
05:30–17:30
— Nothing. At most 30 minutes of spinning after food, and it does not count as training.

Add roughly 30 minutes of commuting each way to every shift above. That is not padding — it is the difference between a window that exists and one that does not.

Layer 2: the morning decides what survives

Layer 1 says a quality session fits today. Layer 2 asks whether it should happen. Five things, checked on waking — not at four in the afternoon, when the number only tells you that you have been at work.

Count them. That is the whole method:

SignalsWhat you ride
0The session as planned.
1The same session, minus the last interval.
2Same duration, all of it in Z2.
3+Thirty easy minutes, or nothing.

Two signals is not a punishment. Endurance on a tired body is still the best aerobic stimulus available that day. It is not the session skipped; it is the session done differently.

Why a count instead of thresholds that each veto the ride: one bad number is noise. Three bad numbers at once is information. A single gate turns every morning into a negotiation with one statistic.

What happened when I tested this against a week of data

The first version of Layer 2 used Body Battery under 65. I ran it against seven days of my own watch data — sleep, HRV status, resting heart rate, Training Readiness and waking Body Battery — to see what it would have told me to do.

DayWaking BBReadinessRule said (under 65)Rule said (under 50)
27 Sep2846Easy or nothingEasy or nothing
28 Sep3648Easy or nothingEasy or nothing
29 Sep6869Full sessionFull session
30 Sep8273Cut the last intervalCut the last interval
1 Oct5253Cut the last intervalFull session
2 Oct6077Cut the last intervalFull session
3 Oct6487Cut the last intervalFull session

Look at 3 October. Waking Body Battery 64, one point under my threshold, so the rule wanted me to cut the session — on a morning when the watch said Training Readiness 87, well recovered, and when I went out and rode 72 km with fifteen personal records in it. The rule was wrong, and it was wrong on three of the four best days of the week.

A rule that cries wolf four times out of seven is a rule you stop reading. That is not a failure of discipline, it is a failure of calibration: my waking Body Battery almost never clears 65, so 65 was not a warning line, it was most mornings.

So the threshold moved to 50, and the week scores as it should: two unambiguous rest days during the bad spell, one soft flag after a short night, and four clean days.

Calibrate it to yourself and do not copy my number. Take two weeks of your own waking Body Battery, and pick the value you only drop below on mornings you already knew were bad. That is your line. Do the same with resting heart rate: it is your own baseline plus five, not mine.

Layer 3: what is coming next

The two layers above look at today. This one looks forward, and it is the layer I did not have — which is exactly why I buried myself at the end of September.

Here is what those rules are made of. On 3 October I rode 72 km outdoors — three and a half hours, 525 m of climbing, the biggest day in months. On 4 October the first of four night shifts started.

The ride itself was right: the watch said well recovered, and my acute-to-chronic load ratio had fallen to 0.5 with the chronic load sliding from 646 to 525 over a week. That is not freshness, that is detraining, and it needed a big day. What was wrong was the placement — the longest ride in months on the last day before four nights of broken sleep, a week after HRV had bottomed out at 26 ms. Two days earlier and the same ride costs nothing.

Layer 4: a session you missed

The one rule that is not negotiable

A fever, or symptoms below the neck: no training. That one is not a signal to be counted against the others, and Body Battery has no say in it.

And if heavy fatigue lasts more than about a week despite the rides being dropped, it has stopped being a training question. See a doctor. No decision table on a website covers that.

Two goals at once: more watts, fewer kilograms

These pull against each other. Watts come from quality sessions, which need fuel and recovery. Kilograms come from an energy deficit, which limits both. You can chase both, but not both at full speed, and the order of sacrifice has to be decided in advance:

How large the deficit should be and what to eat are not training questions and are not answered here. What a ride needs in food and fluid, in general ranges, is on the nutrition calculator.

What this is not

It is one rider's decision system, built from one roster and one set of numbers, and some of it is reasoning rather than evidence. The thresholds are a starting point, not findings: mine came from watching my own session ratings for a few months and they have already changed twice.

It is not coaching, and nobody is watching your numbers but you. What it does do is turn a daily argument with yourself into a count you can do in ten seconds before you get out of bed — which is worth more than the precision of any single threshold in it.

You can see the rules being applied, and broken, on the plan. The longer version of why a rotating roster breaks normal training plans is on the shift work page.