Let’s try this: 20 knights (HD 1 AC 2 1d6 ML 8 with 90 hp) against 200 peasants (HD 1 AC 9 1d6 ML 6 with 900 hp).
At this point the knights must realise that this is harder than imagined. The next morale check is due when the knights have dealt another 395 points of damage. Since they hit on a 10 they have a 55% chance of dealing 3.5×5=17½ points of damage every round, so if the knights are never hit, this takes 395/0.55/17.5=41 rounds.
At the same time, the peasants hit on a 17 so they have 20% chance of dealing 3.5×9=31½ points of damage. The first hit is going to wipe out a third of the knights. This is bad news for the knights. They had one chance and blew it.
What happens if the knights split up? Let’s say they attach in four groups of 5 knights each.
What would have happened if the peasants had won initiative (50% chance) and landed at solid hit (20% chance, so a cumulative chance of just 10%), but they would have gotten an attack against each of the smaller knight units. As five knights only have 22 hp, that’s a 10% chance for each of the smaller knight units to be annihilated in one go.
It’s hard for me to say what a good strategy would have been. I have the feeling that in this situation the variance is too high and that the knights are better off waiting for a situation where the peasants try to split up.
Thus, the forces meet but no engagement is fought. The knights start trailing the peasant army.
I think I like this result.
Let’s try an example from an actual campaign, including player characters. I’m going to use Orcs vs Everybody Else, a scenario where we had a three way fight:
Let’s figure out how we’re going to run this. In the game the dwarves were trying to see whether they could get a good deal, or pull a quick one, or subjugate the survivors, so they’re going to delay.
Bandits are easy, we use our stats for humans. These bandits are poor. They have no armour, but they carry a throwing spear, a fighting spear, and a shield.
Elves are tricky because they use 1d6 instead of 1d8, and they each have a spell. Let’s say that a third has sleep, a third has magic missile, and the rest has charm person. As for arms and armour, they are lightly equipped with chain, bow & arrow, and a long sword. In melee, their horses also attack!
Dwarves are easy but the eye of terror is hard. I’m going to use the stats from the Advanced Edition Companion for Labyrinth Lord.
Orcs are also easy.
The party is hard. I’m just picking some of the player characters and retainers from the Status page...
Perhaps I’m going to simplify this as follows:
As they are “hiding” inside the orc unit, I’m not going to worry about the details, for now. What’s important is that we have at least one player character charismatic enough to push the orcs’ morale up to 9, just in case.
OK, I think now we’re ready.
Roll for initiative! Orcs & party win, then dwarves, then bandits & elves. The orcs & party decide to go against the elves, first.
On with the attacks:
I guess they massacred the elves that fell off their horses...
We still have two magic users to go, so they try to demoralise the bandits, too.
We don’t care about the bandits any more, they run for their lives before engaging in melee so nobody gets a parting shot. Their total losses is 27/4.5=6 bandits.
The orcs have suffered no losses. The part has suffered no losses. The elves lost more than half their number, and they’re still willing to fight, but since the dwarves see how things have gone, they stay put and the elves promise to leave and not to return, and the players let them go.
Initiative and morale is very important, obviously!
I guess the elves are still dangerous. A sleep spell takes out 9 orcs on average. A magic missile and two charm person spells take out another 2. But that still leaves a magic user with a fireball spell which is going to annihilate them... So better to stop now.