Training on seven nights on, seven off
Seven twelve-hour nights followed by seven days off is the most extreme common roster. Here is a fortnight of cycling worked out around it, and why the first two days off are not training days.
Who works it. Offshore and remote-site work, a large share of hospital night contracts, and some control-room roles.
What makes it hard. Seven consecutive nights is a full circadian inversion, and then you invert back. The week off is genuinely free, but the first two days of it are spent repaying sleep, not training — and the last day is spent preparing to invert again.
What a fortnight actually looks like
This is not an example typed out by hand. It is the output of the same rule engine that builds the plan on the front page, run against a seven on, seven off (nights) roster. The shift decides the size of each day, then the hard session goes in the best remaining window, then endurance fills what is left.
| When | Shift | Session, and why it is there | TSS |
|---|---|---|---|
| Days 1–7 Mon–Sun | 12 h night | Rest A twelve-hour shift. Nothing planned: thirty easy minutes after food at most, and it does not count as training. | — |
| Day 8 Mon | Day off | Endurance 45′ · 45 min First day after the nights. Easy riding resets the body clock faster than lying still. Below 40 on Body Battery, skip it without guilt. | 31 |
| Day 9 Tue | Day off | Tempo 2 × 12 · 62 min The hard session goes here: the best window of the week — slept, with a free morning. Quality needs Body Battery of at least 50, over 6 h of sleep and HRV still balanced. Miss one and you move the session by a day rather than shortening it. | 52 |
| Day 10 Wed | Day off | Long ride 2 h · 120 min The week's longest ride. On a low-carb diet this is what moves fat, not the intervals. Below 40 on Body Battery, skip it without guilt. | 89 |
| Day 11 Thu | Day off | Endurance 60′ · 60 min Endurance fills the rest of the week. It is what the base is actually made of. Below 40 on Body Battery, skip it without guilt. | 42 |
| Day 12 Fri | Day off | Sweet spot 3 × 10 · 70 min The hard session goes here: the best window of the week — slept, with a free morning. Quality needs Body Battery of at least 50, over 6 h of sleep and HRV still balanced. Miss one and you move the session by a day rather than shortening it. | 65 |
| Day 13 Sat | Day off | Endurance 45′ · 45 min Endurance fills the rest of the week. It is what the base is actually made of. Below 40 on Body Battery, skip it without guilt. | 31 |
| Day 14 Sun | Day off | Rest Rest by design, not by accident — the week's hours are spent. | — |
Sessions are written as percentages of FTP, so they scale to whoever rides them — there is nothing here tuned to one person's numbers. The quality session lands on day 9, day 12; the longest ride on day 10. Change the roster and both move, which is the entire point.
The three things this roster decides for you
Write the night week off entirely.
Seven nights is not a week where training is compromised; it is a week where training does not happen. Accepting that is what makes the other week usable.
You have roughly four real training days, not seven.
Two days to recover at the start, one easy day before going back, and four good days in the middle. Four focused days is a complete training week — but only if you stop trying to make it seven.
Alternate the fortnight, do not repeat it.
With a whole week unavailable every fortnight, consistency comes from what you do across months, not weeks. Judge the load over four weeks and the pattern stops looking like a problem.
Do it with your own dates
The fortnight above starts on an arbitrary Monday. Yours does not. The week planner takes the shifts you actually have next week and places the sessions the same way, in your browser, with nothing sent anywhere.
The reasoning behind every placement is written out on the rules page — including the rule I got wrong first time and had to change after it failed against my own data. Eating around the same roster is covered on the nutrition page.
I work a rotating roster myself and publish the whole plan and its results, including the weeks that go badly. That is the only claim this site makes: not that it works for everyone, but that it is real and you can check it.