Monday 1 April 2013

Guards! Guards!

Ready for Action!
That's enough Terry Pratchett titles for now.

The BtB stance is that 1% of the world's human population are capable fo advancing in level in the way that the PCs do. But, in addition to these, there are several other types of character who have aspects of such "full" characters. Spys, for example, have levels but not XP and have a very small subset of the Assassin's skill set (ie, spying) which increases with their level. Interestingly, no guidance is given as regards to hit points.

But the most common "semi" characters are the watchmen and mercenary soldiers  who are encountered either patrolling areas or guarding specific places.

Unlike the spy, these characters do no increase in level, at least in any formal mechanical way laid down in the rules, but they do have the full list of abilities from their model class, specifically "Fighter".

The default level of these fighters is zero - the men-at-arms level where the fighter has 4-7 hp, and seems (from reading the details for clerical followers) to have 3 weapon proficiencies. Their ability scores run from 6 to 15 except
for STR which ranges from 7-16, and CON which ranges from 9 to 18.

Some improvised weapons are more frightening than others
These guys are still a step up from the normal soldier of pseudo-mediaeval times, who would in fact probably be a fit farmer with some padded armour and some agricultural weapon nailed to a long stick or pole. Those guys have 6-15 across the board and d6+1 hp and can fight at level 0 with their one weapon (if that). They're the bulking agent of armies and guard houses - very cheap, very expendable, not very reliable. We'll not worry further about them.

"Take me to the correct movie reference!"
Carry on, Serjeant
Getting back to the men-at-arms, they are essentially "privates" or "constables"; they have no leadership qualities. The initial step on the command ladder is the "Serjeant". There can be as many as 1 serjeant per 5 men-at-arms, and as little as 1 per 10. What happens if these numbers are ignored? Well, in the case of too many serjeants, probably nothing more than wasted money and perhaps arguments over who's in charge, whereas too few serjeants probably means a morale penalty.

The serjeant is equal to at least equal to a first level fighter and their ability scores are slightly different, although still (I think) based on 3dA (averaging d6s where the 1 is replaced by a 3 and the 6 by a 4). For serjeants, this is boosted by 2 for STR and only 1 for CON, for ranges from 8-17 and 7-16. Hit points are based on 1d10.

While mercenary serjeants are always 1st level, watch serjeants range up to 3rd (presumably reflecting the lower rate of mortality in the upper ranks or something).

Interestingly, the rules seem to assume that in a city the 1st-level serjeants will all be assigned to watch patrols rather than being put with "City Guards" who seem to have more localised briefs - gates, walls, and police functions, which I take to mean more or less riot-police rather than beat-police. City guard leaders are always at least 2nd level.

"I'm 3rd level, you know. Or is it 4th?"
As The Sparks Fly Upwards... the serjeant takes orders from someone else. Specifically, a lieutenant.

The mercenary lieutenant is 2nd or 3rd level; the watch lieutenant 4th or 5th. Each lieutenant is able to command their level times ten men-at-arms, although these must still be handled by the normal number of serjeants (who do not count towards the limit). Beyond this, the lieutenant is also able to handle a number of "special" serjeants equal to their own level. These serjeants are available for tasks other than simply directing troops.

The ability scores for lieutenant are not spelt out, personally I would use the serjeant method for those under 4th (ie, hero) level, and switch to the "special character" method from DMG p11 for those of 4th level and up with class minimums applied, so that STR and CON would run from 9-18 and 7-18 but with an average of 13.

"Captain", not "King"
Aye, Aye,  Captain
Top of the chain is the captain. The mercenary captain runs from 5th level up to 8th (yes, that means mercenaries have no 4th level fighter equivalents BtB - maybe they have a gap year) but is mechanically very similar to the lieutenant except s/he can command twice as many troops per level. So a 7th level captain can have 140 men-at-arms at the bottom of the command pyramid they lead. The captain may additionally have a number of lieutenants on special duty equal to their own level.

Captains are assumed to lead homogeneous troops - i.e., they lead "light foot" or "longbowmen" or "medium cavalry".

Both the captain and the lieutenant cost 100gp per level per month, which includes their uniforms, food, and barracks/housing.

Paying the troops
Economics (Stupid)
Let's look at the make up and cost of a large watch in a city which is headed by the highest level captain (8th) working at full capacity; what are the numbers?

The bottom level is limited by the captain's ability to command, so that's simply 160 guardsmen on duty at any one time. This number of guards requires 4 lieutenants of the 4th level (or 5th, for that matter, but 4th is cheaper) and 16 serjeants. The captain has up to 8 assistant lieutenants and the 4 "command" lieutenants have up to 4 assistant serjeants each. 8 assistants seems excessive, so I'd pare that back to a nice Bilko-style 2. Similarly, I'll take half the normal number of assistant serjeants.

We'll assume that the guards themselves are classified as light troops and cost 1gp per month with the serjeants costing 10.

That comes to a total of: 1 captain (800gp), 6 lieutenants (2400gp), 24 serjeants (240gp), and finally 160 men-at-arms (160gp).

So that's a total guard compliment of 191 with a monthly cost of 3600gp and an annual cost of 43200gp.

For a fighter, that equates to a city population (men, women, and children) of 10286 just to pay the guard (assuming PHB taxation levels); for a cleric that comes down to 8000 BtB. This in turn, gives a population to guardsmen ratio of  between 54 for the fighter and 42 for the cleric. Magic users are even more disadvantaged with a required population of 14400 and a population to guard ratio of about 76.

"We're collecting for new uniforms...bit tactless I suppose."
These population figures are pretty high for a mediaeval English city and we should probably double them to come anywhere close to a total monthly cost for running the place, but on the other hand ancient Rome is reputed to have had more than a million inhabitants (many, many arguments on this point) and even with doubling costs could have supported something in the region 15,500 men-at-arms plus their commanders using these figures which equates to 5185 guardsmen guarding the city during any 8hr shift (you hadn't forgotten that, had you?), although all could potentially be called up in an emergency.

I literally have no idea whether this is realistic or not, although the figure for Rome seems low but on the other hand, I've pulled the "doubling" cost for the whole of the city's infrastructure out of nowhere so maybe that's the issue.

Anyway, the point here was to look at what the 1e rules imply (probably unintentionally) about what it means to "encounter a watch patrol" and also what resources the NPC town or castle owner may have for assigning higher level guards to high-status locations such as treasure vaults and important prisoners while still preserving the particular mix of 0-level men-at-arms within the setting.

Inappropriate armour's not just for women
Repeating 20's at 20 Yards
Those men-at-arms are not to be ignored, particularly by non-fighters. They can be using missile weapons, and they can also simply overwhelm a party in the open by a mass overbearing charge. Even against fighters, if the men-at-arms can get within missile range, the combat tables and the limits of shields mean that even +5 platemail and +5 shield is not truly proof against their attacks (a dozen longbowmen will do an average of 1¾hp of damage per round if they can pin down such a target behind a ditch or in a courtyard, more if they can surround it).

While spellcasters are essentially neutralized by mass melee combat (they can certainly still use magical devices but so can everyone and even then overbearing and grappling is a major threat), the fighter should come into his/her own with a 4th level fighter able to attack four men-at-arms in a round and a lord being almost untouchable in melee by normal humans.

"Elementary, my dear Robin."
You've come to Nottingham once too often! (Which is to say "once")
Thus the game is designed for Conan's slaughter of BĂȘlit's crew, or for the attack on the castle to come down to the "main event" of Errol Flynn against Basil Rathbone - the two high level characters must resolve the combat between them. Hector must face Achilles; Ferdia must try to stop Cuchulain. Everyone else who tries just adds to the pile of bodies that line the battlefields like the embankments of a red river.

This is absolutely the origin and intent of the multiple attacks against <1HD rule and once the DM starts placing even 1st level fighters in every guard post or along every stretch of curtain wall, that is lost if one is playing by the book. It can be patched up, of course, by changing the threshold at which one grants the multiple attacks and I think that's particularly useful when play is not centred around areas with lots of humans. Having a single normal orc "block" the multiple attack ability seems okay when the fighter is 2nd level, but by 10th level it seems obviously out of whack when the same fighter can slice across a normal battlefield like a tank.

I have tried a "sliding" version of the rule where, for example, a 10th level fighter would get 2 attacks against an opponent with 5HD but it's irritating and fiddly in the middle of AD&D's very fast combat resolution and currently I'm thinking about giving the full number of attacks if the opponent's HD are less than half the fighter's.


  1. Good analysis, and I like how you have attempted to reconcile two sections of the DMG (personally, I suspect they were written on separate models and probably by different individuals). However, you have accidentally written that captains can be 4th-8th level, which makes your comments afterwards about 4th level seem nonsensical. :D

    1. Fixed; thanks. I suspect the two-authors explanation too, but I was expecting the link between revenue and the cost of a sensible guard compliment to be a disaster, which it wasn't. So maybe there's a bit more of a unified plan than it sometimes appears.

      Aaaand, maybe there isn't.