A patrol pattern chooses which route to walk and when, and the choice decides how well an observer can predict it. A fixed rotation of routes is known after one cycle, a rule never to repeat the previous route lets an observer guess right 1 time in 2 among 3 routes instead of 1 in 3, and a random draw at the briefing, combined with a start-time window, is harder to read but lengthens the worst gap between visits.
Part of Watch and Patrol Scheduling
A patrol pattern is the rule that decides which route, in which direction and at what time each patrol walks, with the aim of covering the ground without being easy to predict.
Predictability Is A Real Cost, And So Is Randomness
The parent article says never use the same route twice in a row, plan 3 or 4 alternate routes and rotate randomly, because a predictable patrol tells an observer when an area is unwatched. The aim is right. The rules deserve checking against the arithmetic, because two of them work against each other and randomness has its own price. No part of this is about outsmarting an enemy; it is about not announcing your absence.
Worked Example: How Well Can An Observer Guess?
An observer has watched the patrol on every earlier day and knows yesterday’s route. How well can they guess today’s route among R alternatives?
| Rule | R = 3 | R = 4 | R = 5 |
|---|---|---|---|
| Fixed cycle A, B, C, A, B, C | 100% once the cycle is seen | 100% | 100% |
| Random draw, repeats allowed | 1/3 = 33% | 1/4 = 25% | 1/5 = 20% |
| Never repeat yesterday’s route | 1/2 = 50% | 1/3 = 33% | 1/4 = 25% |
The parent’s “never the same twice in a row” is better than a fixed cycle but worse than a draw: it lets an observer who knows yesterday rule out one route. With 3 routes the guess succeeds half the time instead of a third. The cost of a draw is that the same route sometimes comes up twice running (with 3 routes, 1 time in 3), which is harmless if nobody is watching and lowers the observer’s odds if they are.
Time of day. If the start time is the nominal time plus a random offset of up to plus or minus j, successive visits to a checkpoint are separated by T plus or minus 2j, where T is the nominal interval. For T = 4 hours and j = 1 hour the gap between visits ranges from 2 to 6 hours. Randomising the start by 1 hour either way therefore lets the worst gap grow from 4 to 6 hours. Choose j by asking what the longest tolerable gap is, not by how unpredictable it feels.
The Common Patterns
- Loop. Out by one route and back by another, covering new ground each way. Each part is seen once.
- Out and back. The same ground twice, an hour or so apart; changes between the two passes show what moved while the patrol was out.
- Star. Several spokes from a base, checked in turn. Good for a few distinct objectives.
- Figure of eight. Two loops sharing a central checkpoint. Everything passes through the middle.
- Randomised checkpoint order. Same checkpoints, order drawn. The most unpredictable, the least familiar.
None of these is better in itself. What matters is the ground to be covered, the time available and the state of the people.
Method: Choosing And Running A Pattern
Step 1: Fix the checkpoints and the longest tolerable gap at each (Route Planning).
Step 2: Write 3 or 4 alternative orders that cover them.
Step 3: Draw the route and the direction at the briefing, not the night before: write the numbers on pebbles and draw one from a bag.
Step 4: Inform the posts the same day. The posts must know today’s route and window (Communication Between Posts); the draw is secret from outsiders, not from your own sentries.
Step 5: Vary the start within the window and note the start in the log (Log Keeping).
Step 6: Review the log for a pattern in your own behaviour. If the patrol always leaves within ten minutes of dusk, a watcher knows it.
Diagnosing Failures
- Two routes always chosen together. The draw is biased or people are skipping the hard route.
- Sentries mistake the patrol for an intruder. The day’s route was not passed on; fix the briefing.
- Gaps larger than tolerable. The offset j is too large for the interval T.
- Patrollers know only one route. They get lost on the others; walk each route in daylight.
Common Mistakes
- A rotation described as random. If you can write the order down in advance, an observer can too.
- Randomness in the wrong place. Unpredictable gaps in a route that must pass a cache every day defeat the purpose.
- Never walking the less-favoured route. The ground it covers goes unwatched.
What’s Next
- Learn next: Night Patrols — the same routes in the dark
- Combine with: Perimeter Patrol Routes — varying start time and direction on a household circuit