Troll versus elf

Blog » Troll versus elf

Posted on 14 Apr 2015 22:33

Game on. Our brave team is exploring some caves.
Suddenly, our scout notices a huuuge cavern.
Sadly, it is occupied by old, war-grizzled troll.

The fight starts. And… what now? How should a game master deal with it?

A light, agile elf should easily avoid troll's cumbersome attacks.
A troll, however, should he manage to hit an elf then he will easily turn him into an elven jelly.

Logical, isn't it?
This can be modelled as the situation below (which is how it is resolved in most classical RPGs):

Troll: attack 4, damage 10, defence 4, absorption 5, life 50
Elf: attack 8, damage 2, defence 8, absorption 0, life 15

If the players are using a single 6-sided die as a random component (denoted as 1d6), an elf will hit a troll virtually every time, damaging him for 0 to 3 damage per hit. Troll has a very low chance to actually hit that pesky elf (he has to roll 6 on d6), however, if he does hit, he will deal on average 13 damage (10 + d6), most likely killing an elf on the spot.

You could say that this is how it should be done. And in this particular case it even works!
However, the way skills are described here makes player's declarations kind of meaningless. The player only decides "I attack the troll", without any subtlety.

Let us extend our example.

This time an elf and a troll are fighting on an old, rickety rope bridge.
And this time elf's player approaches this topic differently, declaring an action:
"I want to use enemy momentum to make him fall off the bridge."

What now?

In theory, elf will of course hit a troll. But the mechanism shown above does not provide a proper way to tell if the 'trip the troll' action was successful. Or if it was even possible. Attack vs… defence? absorption? The mechanics doesn't say how to resolve it. The game master is at loss.

For us, this kind of mechanics is broken - it does not support our play style. Every time the game master has to improvise the rules, something went horribly wrong.

Therefore, we are using something different. Our character description says what a character can do and what is he / she like.
For example, sticking to a fantasy setting, we can say that an elf has a "war-dancer" (4) ability while a troll has a "berserker" (4) ability. Elf is "agile" (4), troll has "stone skin" (4)

Whenever elf uses his attack speed, reflexes, dodging ability etc in combat, he is using his full power - reaching 8.
Whenever troll is soaking hits or tries to smash enemy, he is at his 8.

Going back to the bridge example.
On this terrain an elf has an advantage. A troll can't fully use his tankiness and his size works against him. In addition, he has no skill to compensate for the hostile environment.
If that elf declares "I am trying to cut one of the ropes, so that the bridge tilts and I try to grab a rope to stay on it while the troll falls down", he will be using his "agile" trait. (he cannot use "war-dancer" because that was not a directly offensive action). Troll has no skill that would help him, resulting in elf having 4 and troll having 0. If elf succeeds, fight has just finished, which does not necessarily mean that the troll is dead.

If an elf wanted to maximize his numbers, he would need to change his declaration to e.g. "I feign an attack and dance past the troll to provoke the troll to do an attack that will result in him breaking the rope he's standing on with his club while (…)"

However, if the context is negligible, that whole fight boils down to one lucky dice roll and "I… yyyy… I try to cut him!". An agile elf war-dancer is '8' and a stone troll berserker is '8'. Without the context, this battle is boring. So, we have gained an interesting gameplay involving the player when the context matters, but we have lost the intricancies between the characters in the process. A tradeoff. Not an ideal situation.

So we are back to square one, as this is not really what we were aiming at.

It is worthy considering that this problem gets even bigger if we start to think about any type of non-combat conflicts.

Hektor is a tough inquisitor, he can interrogate anyone and anything. "If it sqeuaks, it will talk" - type of an inquisitor.
Sabina is a cute mascot. She can charm anyone and anything if she wants to. "If it smiles, it will like me" - type of a girl.

Hektor will always be able to interrogate Sabina, who will confess everything she knows and even more. On the other hand, Hektor will not be able to refuse her if she asks really nicely.

And this is a "troll vs elf" situation, but in slightly different context and more complex to tackle. Because in a way, physical combat is easy. It deals with something tangible - physical state. Someone will get hurt - the area of operation is "body", no matter the goal. Social conflicts, however, can be approached in many different ways. Sometimes a move in one direction will make other options unavailable.

For example, if Hektor wanted to interrogate Sabina, she could appeal to his better side he did not even know he had and no interrogation would happen.
However, if Hektor was properly suspicious about her, his inquisitor nature would win, leaving her womanly wiles useless against him.
But also Sabina could convince Hektor's superiors and he might have to stop the investigation. Or Hektor could threaten Sabina's friends to move away from help and not help her. So… how to resolve the situation like this: Sabina tries to convince her friend to help her even if Hektor earlier on convinced that friend to stop helping Sabina?

Not that easy to tackle.

One of the solutions to "troll vs elf" problem could be something between "attack, defence, damage…" and "agile war-dancer". Maybe we should use (borrowed from MMO) idea of "tanking" and "evasion". If we do that, agile war-dancer could give 6 to evasion and 6 to hit, while stone berserker would give 6 to damage and 6 to absorption.

In the fight without a context those mechanisms work similarly - 6 vs 6 (boring).

However someone having skills of both berserker and war-dancer would be able to defend from the enemy where they both are strong (damage vs. absorption) and attack where enemy is weak (precision critical hit vs… nothing). Now it is less boring. Now it starts making some sense - earlier on, without the distinction Hektor is (6) and Sabina is (6), they work the same way and their influence manifests in the same situations with the same results - the difference is just in color (that is, in descriptions).

Of course, we can expand on this idea.

A barbarian (more interesting than 'berserker') would ensure for example +1 to damage, +1 to hit (note: no defence), +1 courage (as in, 'willpower', anti-fear defence, against Hektor but not Sabina). If a character can have 6 skills and 6 traits describing it, the set and combination of those skills and traits would create a profile of that character.

This profile would naturally reward a player for doing things that fit the character, and particular skills would determine an exact way of how character's goals are achieved - therefore modifying the difficulty by the context of the game world.

Unfortunately, breaking all character skills and traits to a long list of statistics slows down the game if it takes place in real-time.

Thus, another approach to this problem; slightly similar, but not exactly.

You could roughly divide all skills into 3 categories (this requires deeper thought; at the moment, "3", but this will probably become 5): combat, social and "science"

Each of the categories can have skills that are subtle (war-dancer, everybody's mascot) and those that are rough (barbarian, inquisitor).

If we agree that defending against an attack from "the other" category is more difficult, elf vs troll issue becomes much smaller.

The biggest difference is in the story: elf will be able to trick the troll, but can't really hurt it, while troll will simply be unable to hit the elf.

Or, in different implementation, elf can critically hit the troll ignoring the absorption (direct hit into an eye) while troll will be able to smash an elf ignoring the evasion. The exact implementation depends on the goals of the mechanics, but you see the symmetry.


No matter which of those implementations would be selected, the result would be the same: with limited number of skills (traits) characters are starting to specialize. It is no longer possible to create a true one-man army - your character will always have a soft belly somewhere. Even if you did manage to cover all bases, you would be much weaker compared to someone who has fewer skills, but much more specialized; on the other hand, you might be able to use those weaker skills to hit an opponent where he is weak. And natural roles would appear in the team. Otherwise, without it, there is a "meta build, as in: the best": you take one social trait, one combat trait, one trait giving you scouting/perception, one trait giving you scientific abilities… in short, one trait per area of interest of your character.


Which solution is the best depends on needs of the players and the playstyle of the group. We plan to experiment until we find something that will work for us.

Regardless of what we choose, "an elf and a troll problem" is a good acceptance test for RPG mechanics. One of many we are utilizing and one of those every mechanics needs to answer somehow.

Like this entry?

O ile nie zaznaczono inaczej, treść tej strony objęta jest licencją Creative Commons Attribution-ShareAlike 3.0 License