A watch rotation assigns named people to a post for fixed periods so that somebody is always watching and nobody is left alone with the whole night. One post watched round the clock is 24 watch-hours a day, which is 6 hours each for 4 people and 4.8 hours each for 5, whatever the length of each shift, so the number of people and the sleep they can protect set the limit, not the clock.
Part of Watch and Patrol Scheduling
A watch rotation is the roster that decides who stands which watch at which post, and it is judged by whether the people on it can keep it up without wrecking their sleep.
The Roster Is A Sleep Plan As Much As A Watch Plan
A rota that keeps a post staffed on paper but leaves its people exhausted produces a watch that is nominally manned and actually asleep. The parent article says as much when it notes that sleeping on watch is a system failure to fix before blaming individuals. A rotation therefore has to balance four things at once: coverage (someone is always there), alertness (nobody watches while worn out), sleep (everyone gets a real night most days) and fairness (nobody carries the small hours alone). They pull against each other, which is why the arithmetic comes first.
Worked Example: What The Staffing Table Really Says
The parent article’s table gives 4 to 5 people for one post watched round the clock, 8 to 10 for two posts and 12 to 15 for three. Each post needs 24 watch-hours a day, so for P posts and N people the watch load per person per day is 24 x P / N.
| Posts | People | Watch-hours per person per day |
|---|---|---|
| 1 | 4 | 24 / 4 = 6.0 |
| 1 | 5 | 24 / 5 = 4.8 |
| 2 | 8 | 48 / 8 = 6.0 |
| 2 | 10 | 48 / 10 = 4.8 |
| 3 | 12 | 72 / 12 = 6.0 |
| 3 | 15 | 72 / 15 = 4.8 |
Every row is 4.8 to 6.0 hours a day: 6 hours each for 4 people on a post, 4.8 hours each for 5. The parent’s labels, “barely sustainable” for one post and “comfortable” for three, cannot come from the watch load, which is the same in all three. What differs is how much one absence hurts. If one person falls ill: with 4 people the other 3 take 24 / 3 = 8 hours each, a rise of one third; with 15 people on three posts the other 14 take 72 / 14 = 5.1 hours, a rise of about 7%. Larger groups are steadier, not lighter.
The same table also says nothing about when in the day the hours fall, and that is where sleep is lost; see Sleep Cycle Management.
Kinds Of Rotation
- Fixed. Each person has the same slot every day. Simple, and one person gets the small hours for good.
- Rolling. Slots shift so that everyone eventually stands every hour. Fairer, harder to run, and hard on the body clock (Rotating Schedules).
- Paired. Two people on each shift (Two-Person Shifts).
Method: Starting A Roster
Step 1: Count the posts and the hours each must be manned, using the threat-level postures in the parent article.
Step 2: Count the people who can stand watch, after children, the sick and people with other critical jobs.
Step 3: Compute the watch-hours per person per day as above. If it is more than the group can keep up, lower the manned hours (a lower threat posture, or mechanical and animal warning at night) before shortening anyone’s sleep.
Step 4: Write the roster out with names, slots and who relieves whom, and pin it where every sentry can see it.
Step 5: Review in a fortnight. Ask who is not sleeping, whose log entries have become copied, and who dreads a slot.
Common Mistakes
- Counting posts, not people. A post you cannot staff is better left to alarms than staffed by exhausted sentries.
- Using the roster as a punishment. Nobody should draw the small hours as a penalty.
- Never rewriting it. People fall ill, leave, and learn what the roster costs them.
What’s Next
- Learn next: Two-Person Shifts — watches that are never held alone
- Combine with: Sleep Cycle Management — protecting rest inside the roster