Wednesday 6 November 2013

Loyalty Bonus

The only "core system" I didn't cover back in the early days of the blog when I was doing the introduction to AD&D posts was, or perhaps were, the twin topics of loyalty and morale.

The reason for that is that it's been a long time since I used them as written. In fact the loyalty rules contain a note to the effect that an experienced DM shouldn't really need them and probably the same goes for the morale section. But, as with so much, it's worth going back and re-calibrating every so often, so here's a summary of the loyalty and morale rules.

Twin Peeks
Looking at morale first, we're given a system to decide whether a particular opponent will stand and fight when the going gets tough. It's certainly worth noting that this starts at a base of 50% for creatures of less than 2 hit dice if they are intelligent. What if they're not intelligent? That's not stated but I'll come back to it below.

There's a list of fairly obvious to make a morale check - 25% of party killed, leader runs away etc. - and a set of possible modifiers. To make the morale check, the DM rolls the dice, adds the modifiers and subtracts the base morale of the figure. If the result is positive then the creature falls back, disengages, flees, or surrenders depending on how high the result is.

On the loyalty side, things seem fairly similar: there is a base score of 50% modified not by hit dice but by the charisma of whomever the loyalty is owed to, a set of occasions on which to test loyalty, and a bunch of modifiers.

Sadly, as is often the case, there is a pointless difference too: loyalty modifiers are added to the loyalty score before rolling, although they could be subtracted from the dice of course, and any score over the loyalty score indicates a fail. Although there is no grading of the degree of failure as there is for morale, I would suggest that the DM should take it into account.

Loyalty has a broader scope than morale. Loyalty can be tested by a bribe or by temptation when left to guard valuables and so forth. It's also subject to many more modifiers than morale and it's quite possible to end up with loyalty scores of 200% and more if a high-CHA character associates for a long time with henchmen or troops who are treated correctly. Simply being lawful good gives a bonus of 15%, and a racial preference gives another +20%, which means that paladins inspire a base loyalty score of 115%+ right off the starting blocks when dealing with humans who are of a compatible alignment.

On the subject of which, there is the question of compatible alignments and a strange table on DMG p 37. If the henchman (or whatever) has an alignment "1 place removed" from their liege then there is no modification. 2 places gives a mod of -15% and three one of -35%. But what is a "place"? The examples given are bizarre: LE<->LN is 1 place; LE<->LG is 2; but LE<->CG is 3. So what would LE<->CN be? I can only assume that the last row in the table should read "3 or more places removed" and that neutral is meant to be 2 places from LG and only 1 from CN.

It is possible for substantial loyalty penalties to be incurred, but generally they are all for things which a LG character would never do, so it's a fairly sure thing that if a paladin is leading a group of NPCs then those NPCs will come under "Fanatical - will serve unquestioningly and lay down own life if necessary without hesitation", probably even the chaotic evil ones (which is as good a reason as any to introduce a massive penalty for alignments which are opposite).

Loyal to whom?
So who is loyalty owed to? It's not defined specifically but the DMG does repeat the fact that the rules apply to hirelings as well as henchmen and seems to intend that the rules be used for any "associated character". I take this to mean those which are paid by or recruited by a character directly to to whom the liege character has paid specific attention to.

In a combat situation, I would also apply it to any friendly figures within a radius of the character's CHA stated as scaled inches (ie, 12" for a charisma score of 12), assuming that they are not being commanded by someone else and can be communicated with to some degree. In the case of an army on the field of battle, perhaps the normal (non-mercenary) soldier could have a score calculated from a base 0% instead of 50% and capped at 90%.

Morale is a much simpler system and hardly has any more detail than covered in the overview above, but it's worth looking at some typical scores.

A bog-standard man-at-arms has a morale score of 50% and a morale check with a 5% penalty is triggered when any figure has lost ¼ of its hit points. Since a man-at-arms can't have more than 7hp, this means that just 2hp damage gives a 55% chance that the figure will try to withdraw from combat to some degree. Orcs are the same, and gnolls have only a 5% increase on that.

A troll, on the other hand, has a base morale of 81%. Assuming that non-fire/acid damage doesn't bother a troll much, then even facing a superior force which is striking twice as often as the troll means that the troll will only fail morale 19% of the time and will never panic or surrender unless the attacking party are really showing no signs of wear and tear.

But still - 19% isn't peanuts and the morale system's tables allow for a rout to set in as each deserter grants a 15% penalty for those left behind, so three trolls will not last three times as long as one before retreating.

Morale makes a huge difference to combat, especially at low levels where it's needed most to soften the opposition that a party might face. Ignoring morale issues is certainly a cause of many low-level party wipe-outs.

So what about those unintelligent monsters? Well, it depends on what one means by unintelligent. My rule of thumb is that animal intelligence means a morale of 50% flat regardless of HD while true non-intelligence effectively means infinite morale, and certain particularly aggressive animals such as bull elephants, giant wolverines, water buffalo, and females with young, should get the normal hit dice bonus.

On that subject, what about loyalty and unintelligent creatures? Well, clearly animals can have loyalty but I'd say that most fungus can't. Most of the loyalty modifiers are, in fact, reasonably applicable to any creature that is capable of personalized memory of individuals and I would include the alignment mods (see below).

Conflict of Self Interest
The place where loyalty and morale come together is in combat. The DMG simply states that creatures which have loyalty use it instead of their morale score in morale checks (assuming that they have been ordered to fight, of course). That's fine but there's the chance that morale is actually higher than loyalty, what then?

The simple answer is that an NPC fights for themselves first (morale) and then for their leader (loyalty). If they don't want to fight then they will only do so if commanded to do so by someone they will obey. So, when morale breaks, check using loyalty.

For example, a party with three NPC men-at-arms (Larry, Curly, and Harpo) is in a dungeon and get into a fight with a substantial force of hobgoblins. The MaA have a loyalty to the party leader of just 40% and a morale of 50%.

On the second round of combat Curley is dropped to half their hp and makes a morale roll (at +15) of 70 - disengage and retreat. The leader orders Curly to stand firm and he makes a second check using his loyalty score (but with the same +15), rolling a 30. Curly's morale has broken but his faithfulness keeps him going.

On the third round, Larry is killed. Curly's morale has already broken so he rolls another check using his loyalty score (now at +30: 10% for a quarter of the party being lost, 10% for a friendly figure being killed, and 10% for taking casualties without inflicting any) with a total roll of 83, Curly throws down his weapon and flees. Harpo, meanwhile makes a check against his morale score (at +30% too; Curly's desertion won't have any effect until next round) and rolls 78% (disengage) and then 93 (surrender); taking the best of these means that Harpo attempts to disengage while Curly legs it.

If these guy's retreat is blocked (say, by other members of the party) then the only chance is to talk them down, making another loyalty check, but this will disrupt the party and another fail here may lead to fighting between the would-be flyers and the rest of the party.

There are modifiers to loyalty for alignment. I know from experience that the first thing a would-be Evil leader will claim is that they are not going to act evil and so should have the "neutral" modifier of ±0%. Well, as previously mentioned, alignment is an actual force in the AD&D world and as such I regard these modifiers as being unavoidable. People just sort of know, even if they can't quite put their finger on what the problem is.

Evil has many advantages in that it's always easier to be destructive than to be helpful and careful and the alignment modifiers to loyalty can be a strong boost to Good characters.

Meanwhile, Back in the Real World
At the table, of course, all this is baloney. There's no way any sane DM would sit and add up and subtract all the possible modifiers that might apply and still less track the changes (-1% per living enemy hit die?).

But the base scores are not too hard to quickly tot up and note on a stat-line for a monster or NPC and after that most of the intent can be captured by making reasonable guesses at modifiers and applying them ad hoc. If you roll 90 for morale when the party are low on hit points, it's likely that the NPC has decided to quit or even change sides if their loyalty breaks too. You don't need to worry about every little mod.

And in that spirit, you really don't need to use percentage scores for all of this either. The PHB mentions giving a +1 to morale in the bless spell and by +2 in the 4th level illusionist emotion spell. These clearly are signs that Gygax was intending to divide all these numbers by 5 and use a d20 instead. I suggest you follow suit and forget about the little ones like the hit dice modifiers to loyalty.

Some Spells
Banner (conjuration/summoning)
Cleric Lv3
Components: V,M
Casting Time: 1r
Save: None
Range: Touch
Duration: 1 turn/lv
Area of Effect: target's charisma+1" radius per level

Banner causes a spectral flag, pennant, banner, or otherwise culturally appropriate image to appear above the character touched. This expands the range of the character's charisma for the purposes of loyalty scores. Any friendly figure within that area is treated as having a normally calculated loyalty score to the target of the spell. If there is a conflict, whether because of another use of this spell or simply that the NPC already has a loyalty to someone else, then the higher charisma is the one which gains the NPC's loyalty with ties broken by the level of the charismatic character and finally by dice rolling. Once the spell ends, previous loyalties are restored.

The material component is a finely made duplicate of the image which is to be displayed. This must contain on it some identifying and inspiring mark associated with the target character.

This spell is generally only available to clerics of war gods and leader-type deities (e.g., Odin and Tyr; Zeus, Ares, and Athena; but not Thor or Pan).

Talisman (conjuration/summoning)
Cleric 5th level
Components: V, M
Casting time: 1 turn
Save: None
Range: Touch
Duration: 12hrs
Area of Effect: 1"/ level

Talisman causes the cleric's holy symbol (the material component) to radiate a sense of well-being and security to all those of the same faith as the cleric (same deity or same pantheon). It grants those within the range of the spell +10% (or +2) to their morale scores and -2 to those that they fight.

For Good clerics, the symbol also radiates a bright light such that monsters which do not like sunlight will fight at -1 even at night (vampires and the like are not damaged but do receive -1 to combat rolls). Evil symbols generate a gloom such that evil monsters can fight normally even in daylight (vampires and so forth included) and those without infra- or ultra-vision suffer -1 on ranged attacks only.

Protection from Evil/Good will ward out these effects as appropriate, assuming that the casting cleric is Good or Evil.

The area of effect is centred on the symbol and continues even if it is taken from the cleric, but no cleric can have more than one of these spells in operation at a time; later ones fail upon casting.

Magical items which duplicate the talisman spell power must take the form of holy symbols and never have charges. They are often associated with the tombs of saints (of all alignments).


  1. Nice dissection. I am pretty sure that +1 business refers to the OD&D morale system, mind.

    1. Well, that's what I get for putting my copy of OD&D on the top shelf where it's a pain to reach. I'll have to have a look at that again. Maybe converting bless into +5% is not enough (and maybe it is, that's the problem with using probability curves).