How Many Soldiers are Needed to Patrol/Defend a Medieval Wall?

Question

I have a walled city in a medieval RPG setting. It's out in the middle of a desert and has an impossibly efficient means of acquiring water and producing food off of farmland inside the walls (magic is involved), so its geographical location and internal self-sufficiency make it a very daunting target to attack. However, because of the monsters that infest the world, the walls need to be patrolled during times of peace in order to keep the monsters out; and during the rare times of war, there need to be enough soldiers to effectively defend the walls against determined human attackers.

The walls are between 20-30 feet (6-9 meters) tall, there's no moat or ditch outside the walls; and the walls are, at a bare minimum, 12 miles (20 kilometers, 20,000 meters) long. Assume that the existence of magic doesn't dramatically change the realities of combat, and that you need roughly the same number of defenders to defend the same chunk of wall against the same number of attackers.

With this in mind, how many soldiers would be needed to patrol the walls, and how many would be needed to defend the walls?

EDIT: I suppose I should clarify what they would be defending against. Assume that the defenders have to protect against an army of, say, 20,000 men, or one man per every meter of wall. The walls provide an advantage, but that's a lot of wall to target for various tricks, ruses and attacks.

EDIT2: The attackers are actively attacking the walls, and there is nothing of value outside them, so the defenders have no reason to give up the advantages of the walls. Since the city is self-sufficient, trying to starve out the defenders won't work. The only option for the attackers, then, is to storm the city, either through brute force or trickery.

EDIT3: Of the monsters that try to get in, the ones that are most able to scale the walls are probably about human sized, and have comparable levels of stealthiness, and aren't much more dangerous than an armed soldier... but they operate exclusively at night.

Answer 1

Patroling

The normal patrolling of a wall can be parallelised: substantial fortifications are invariably broken up by regular watchtowers and the guards in one watchtower are responsible for patrolling the the wall between themselves and the next watchtower over. How far apart the watchtowers are dictates how challenging a job that is.

How long would it take a sneaky monster to climb your wall? A world-record-level human can climb 15 metres in less than six seconds, but that is up a wall that, by fortification standards, is pretty poor (as in has a lot of strong holds). Your generals need to (deliberately or inadvertently) choose a 'minimum climb time', which is the time that, as long as any given stretch of wall is observed at least that frequently, any climbing monsters will be seen before they reach the top. Let's say five minutes for that.

Humans walk at a preferred speed of about 1.4m/s, but wall patrol guards will need to stop regularly to actually look out over the wall, so let's say maybe 1m/s patrolling speed. So in five minutes a patrolling guard will cover 300m of wall. Guards should certainly patrol in pairs at least, and could only realistically be on duty a third of the time, so that patrol actually consists of six men, or 50m of wall per guard, continuously guarded. In terms of how big an army you need to maintain that level of watch, the proportion of the whole army comprised of frontline guards (the tooth-to-tail ratio might be as low as 50% when you factor in both officers and support staff (armourers, grooms, quartermasters, trainers, etc). So that suggests maybe 800-1000 men to continuously patrol 20km of wall.

Clearly, you can swing this number wildly by changing the patrol frequency: double the minimum-climb-time, half the guard, and so forth. This quantifies just how much benefit there is to be gained from making the wall more impenetrable. Increasing the time to half a day (cough GoT Wall) makes the ratio as low as 3.5km per guard, which makes guarding a 500km-long wall with under a thousand men just about feasible (although the Night's Watch is far from an efficient fighting force).

Defending

When defending a wall, the game changes completely, but remember that the patrolling game must also continue as well, so the patrolling manpower is an absolute minimum. Here we care about the numerical advantage that the wall (and any other factors like quality of training, nutrition, morale, etc) gives the defenders over the attackers. Crudely, if the numerical advantage is 5 and the defending general puts one soldier on the wall opposite every five attackers on the ground, the defenders should win. Those defenders need to be in addition to the patrolling guards who need to keep patrolling in order to detect any new incursion attempts.

What's a realistic number for the numerical advantage? The exterior terrain you've described is pretty formidable: an army without very strong logistic support is going to collapse pretty quickly without natural water and food, so the most sensible offensive strategy is probably to march up and attack straight away, especially given that a seige is impractical. So the army will be tired and will not have had the opportunity to construct extensive artillery or siege equipment, they'll have had to bring any rams, ladders or towers with them from some distance. I don't think a numerical advantage of ten to one is unreasonable, in which case a reserve force of 2000 frontline soldiers would be sufficient. Of course in reality when the attackers move in you would also call up all your armourers and officers (so the T3R of that 2,000 can be much higher) and any civilian militia as well, but they would be contingency.

TLDR: a standing army of 3000 would be more than sufficient to continuously patrol a 20km wall, and probably to defend it against a 20k attacker, as long as it was well trained and equipped and had solid morale.

Answer 2

Wall dogs.

This answer is only for the monster question.

Your wall is patrolled by wall dogs. These dogs actually spend their time at the foot of the wall, roaming in packs. They are trained and come for food to their food stations. Each pack has a section of wall of about 1 km.

The wall dogs are really for the monsters. A monster which comes out of the desert at night is going to be detected by the dogs and they will be all over it before it starts climbing the wall. They will be barking too (the ones that don't have a mouth full of monster) and human handlers will come to help. Because of the dogs you only need a couple of humans every km.

Dog senses are better for detecting monster incursions. Dogs are much cheaper than humans and they are more expendable.

Wall dogs will be brought into their kennels in case of a siege. They are fearless but would be wasted against a bunch of armed men.

---

thinking more - this is D&D! So if you encounter the dogs, there will be a chance that a non-dog creature will be in their pack, working with them. The non-dogs will have been raised with dogs and will think they are dogs. I like displacer beasts for this.

Answer 3

Monsters and Armies are different problems.

With individual monsters, you probably want to catch all or most individuals. A leak rate of 20% or 30% is likely unacceptable. So you need enough troops on the wall, 24/7, to catch them. Call it a double patrol every 50m, plus a small squad every 400m or so -- enough to arrive at the midway point in a minute, even running in armor. Quadruple that for 24/7 duty and we're talking about 6,000 troops.

You need fewer troops on a nice, sunny day, but the garrison should be able to cope with rain and snow. Go with the worst weather assumption.

Against an army, you have the benefit of inner lines. The purpose of fortifications is not to be impenetrable, it is to buy time for a small number of defenders until the main defensive force arrives at the trouble spot, and to increase the effectiveness of the defenders. If a few scouts climb the wall, so what? They could just as well disguise themselvers as traders and bribe the gate watch.

An army of 20,000 with baggage and a siege train would do well to do 15 km in a day. By contrast, the defenders need one or two hours from the center of the city to the wall. So it should be possible for the defenders to match any attempt of the invaders at concentrating their force. The key question, then, how many people do you need on the walls to block the invaders below?

Look at the numbers for Chateau Gaillard, or the 1187 siege of Jerusalem (even if that ended in a capture).

Answer 4

Perhaps not directly applicable to the city you have described, but the burghs might be a good data point for medieval manning levels. The burghs were a network of fortifications through Wessex and later England in the 10th century. Farmland was outside the walls, but the farmers would retreat to the walls to protect against raids, and the burghs were also used as strong points during wars.

The Burghal Hidage states:

For the maintenance and defence of an acre’s breadth of wall sixteen hides are required. If every hide is represented by one man, then every pole of wall can be manned by four men. Then for the maintenance of twenty poles of wall eighty hides are required.

A hide was a patch of land of no fixed size used for administrative purposes. A pole is about 5m. So the requirement is a little less than one man per metre to defend the wall (and the hide should also provide enough income to pay for the maintenance, in addition to other obligations such as bridge maintenance). These men would be local fyrd - not full time soldiers, but still fairly well armed and trained. The walls they are defending would be a combination of earthen ditch, stone facings or wall, and wooden palisade, all together 2-3m high.

That's not to say you would expect to see the garrison spaced out 1.2m apart all the way round the wall before a fight - probably they would be denser facing the enemy, much sparser round the back, and there would be a reserve who didn't start on the wall at all. Rather it is a rule of thumb for the adequate manning of the wall.

So I would say that a 20km wall requires 16000 men to man it properly. That said, a force of 20000 men would have very little hope of success against say 10000 atop a wall, so you can get away with under-manning it. But if a larger enemy turns up and you have 10000 defenders, then the wall is too long, and you will benefit from shortening it. Perhaps by falling back to a shorter inner wall after making the attacker pay for gaining the outer one.

Answer 5

In peacetime you need sufficient guards within running distance of where a climber, or pack of climbers, will summit the wall after it is spotted.

You need two shifts of night guards and all the infrastructure to support them.

Work done on the wall to make it harder to climb (precision stone cutting, installing an iron facing, just making it taller, etc) pays a major benefit as fewer guards are now needed.

When besieged you need archers (or ballistae) with fire-arrows (or bolts) sufficient to destroy the invaders supply of wood. without wood there will be no siege engines or tunneling sappers. and no hope of a successful siege.

Answer 6

What shape is your city?

It could be irregular, rectangular, square, octagonal, round, or have a lot of other shapes.

The more regular and equal sided the city is, the more land can be enclosed for the same length of wall. Thus there can be more soldiers supported by the city to defend the wall.

If the city is perfectly circular and has wall 12 miles or 20 kilometers long, it will have a radius of 1.909 miles or 3.1831 kilometers, and an area of 11.4488 square miles or 31.8310 square kilometers. That would be the maximum possible area within walls 12 miles or 20 kilometers long, and will have to support a sufficiently large number of defenders using realistic medieval farming techniques, and/or futuristic food producing technology, and/or magic.

I discuss efficient food production technologies in my answer here:

Giving Tolkien Architecture a Reality Check: Dwarvish Kingdoms1

which is based on ckersch's answer to:

How many people can you feed per square-kilometer of farmland?2

My answers to this question:

How much space would a city need within a ringwall to survive for an indefinite period of time?3

And this question:

Build an impregnable fortress in the middle ages with modern technology4

Discuss the design of an impregnable fortress or city in the middle ages and how to support the number of defenders needed.

Experts in medieval warfare should be able to calculate the number of defenders needed to defend against particular threats.

Answer 7

Patrolling depends largely on visibility.

Two blokes can walk up with a ladder, put it up against the wall and get over in half a minute. Someone more athletic could throw a grappling hook over with similar effects. So basically you've got to have fairly continuous visual coverage of the approach. In daytime, one guard per mile might be fine. Then heavy rain falls, visibility a hundred yards and you need a lot more people. Thick fog, or a moonless night, mean you have to start relying on patrols and hearing.

A medieval army is extremely unlikely to storm defended fortifications at any odds. That would be suicidal. More likely they will undermine the walls, reduce them with siege engines, use rams to go through the gates, or try other methods (fiery projectiles, plague rats, or, more simply, agents inside the city).

Answer 8

Patrolling:
The rule for a detection patrol is very simple. (assuming your absolutely need to stop 100% of intruders)
You figure out the quickest possible intrusion. Then you space your patrols at intervals of 1/2 that time. This ensures that a breach is detected even it one patrol is missing/delayed/stupid.
Example: if it takes an enemy 10 minutes to breach your wall, then a patrol every 5 minutes is needed. If it will take your monster 10 seconds to breach the wall, then you need a patrol every 5 seconds, and someone needs to fire the wall architect!

With a relatively easily breachable "Wall", for example a chain-link fence, this effectively means you need continuous surveillance from watchtowers suitably spaced to cover every inch of perimeter.

In real life, walls do not serve as impregnable barriers, but merely as obstacles. They hinder and delay intruders, both when entering and leaving. Make them look for easier targets. Prevent them from feeling secure once they have entered, as escape is also hindered. Make it harder for stolen items to be carried off safely. Most defensive walls actually fall in this category, I suspect yours will too. If you make it so that a monster trying to enter is more likely to get killed than get away with a prey, that monster will die off of starvation. Even if it does get in sometimes. Harsh, but that's the real world.

As for siege defense: A barrier wall for keeping intruders out is very unsuitable for army-scale defense, and vice versa. A good army defense wall will have multiple sally ports, and tons of access stairways/ramps on the inside, and ammo points, and evacuation routes. All of which make its utility as an intrusion wall compromised.

I strongly advise that the "anti pest" wall be further out from the military wall. Much lower/smaller/cheaper. And very regularly patrolled, or even permanently occupied. The space between your army wall and the nuisance wall is a great place to house and train your army!