Competing Abilities

[Manxome]

Potentially relevant for TNE, but also something I've just been thinking about recently...


It is a common convention in RPGs that characters get (preferably awesome) abilities, and that having these abilities makes them better at stuff. Having the ability to sneak attack or turn invisible or teleport is objectively better than not having that ability. And generally speaking, the value of an ability is based on two things: how inherently awesome it is, and how often it helps you.

Another common convention in RPGs is that characters have a limit to the number of actions they can take, and some abilities require an action to invoke. The fact that you are stabbing some guy in the face right now also means that you are not hurling a fireball, throwing the dwarf across the room, or opening a portal to hell at the same time. Thus, an ability used in this way carries an opportunity cost: the fact that you used ability X means that you can't also use ability Y (until next round). Having both abilities X and Y is still (hopefully) more awesome than either one by itself, because it means you get to choose which one you want to use, but if there's any overlap in the situations where they're useful, then picking up Y means you won't benefit from X as often as before, and so their combined value is less than the sum of their individual values. In general,

Abilities that have to be used in this way are often called "active" abilities, in contrast to "passive" abilities that work just because you have them. In general, the more active abilities you have, the less often you're going to use each one, and therefore the less benefit you get from each one.

The trait I want to call attention to here is that these abilities are "exclusive" in the sense that you can't get the maximum benefit from both at once. Getting lots of exclusive abilities is therefore less valuable than getting lots of nonexclusive abilities, all else being equal.

At this point I should point out that "active" and "exclusive" are not the same thing. You can have actively used, non-exclusive abilities, if the ability uses an action that other abilities aren't competing for, such as when it uses an action type you wouldn't otherwise use, or when you use it before the time pressure starts and its effects linger (and hence buffs let you break the game). You can also have passive exclusive abilities, if for example you have to choose one of several to activate when the conditions are met, or if you have to choose which ability's conditions you fulfill (e.g. weapon focus), or if they consume some other limited resource (MP), etc.

You can also have several groups of abilities using different types of resources (actions, MP, etc.) so that abilities within a group compete with each other, but don't compete with another group.


So why do we care? Well, as noted, RPGs like to hand out various kinds of abilities to characters, and they generally hand out more as you advance, and allow players some modicrum of control over the abilities they receive. And under this model, exclusive and nonexclusive abilities cannot be balanced against each other. This follows from the simple observation that receiving an exclusive ability reduces the value of all your other exclusive abilities and receiving a nonexclusive ability does not, which means that if your abilities are otherwise balanced, there is an optimal number of exclusive abilities for you to take, and every ability past that number must be a nonexclusive ability or you are hurting yourself.

How to solve this issue? Several options:

1. Every ability is nonexclusive. Even if some abilities use a resource (e.g. actions, MP) that others do not, you can potentially eliminate any meaningful competition by making sure that players always have more of those resources than can be used only by abilities (e.g. if special attacks are usable once per day and you can't get as many special attacks as the total number of attacks you'll make during a day, then the number of actions you can take isn't a meaningful limit on the usefulness of special attacks).
2. Every ability competes with every other. Even if some abilities use a resource (e.g. actions, MP) that others do not, you can potentially eliminate any asymmetric competition if there's a single resource that they all use and that is more scarce than the others.
3. You have both exclusive and nonexclusive abilities, but they're not interchangable; you pay for them out of separate pools, and foregoing abilities of one type does not let you pick any more of the other.
4. You have several groups of abilities, and every ability competes within its group, but not with other groups. (An easy way to turn a nonexclusive group into an exclusive group is to limit the number of abilities you can have "turned on" at one time.) Picking an ability reduces the benefit you get from every ability in that group, but not the benefit of other groups. At any given time, one group will have the highest payoff, but if you keep picking abilities from that group, the payoff will drop until another group is a better choice, and so if you receive enough (vaguely balanced) abilities, you'll eventually want to get abilities from every group, and you'll continue to pick from all groups as you get more.
5. Every exclusive ability comes paired with some special effect that scales in precisely the same way as the ability, and cancels the drawback of picking lots of competing abilities. For example, maybe learning more spells increases your total MP, so while you have more spells competing for it, you also have a larger pool to spread around (adrenal skills in Guild Wars basically work this way). Alternately, maybe knowing a single spell comes with some terrible penalty, but every spell you learn beyond the first reduces this penalty in some proportion that coincidentally cancels out the competitive effect of lots of spells sharing a single mana pool.
6. Every ability is worth precisely zero, so none of them are balance considerations in the first place. (The degenerate solution.)

That's actually more options than I originally thought there would be. Any that I missed?

[JonSetanta]

That's a fine discourse. Fine indeed.

Here's my counters and observations, noted in kind;

1. I've noticed that too, particularly in MMOs using separate systems like "fatigue points" for warriors and "mana points" for mages, and in all D&D incarnations. The downside to unifying all limited abilities to the same usages is that if a character obtains some kind of MEGABLAST ATTACK they may now do that attack X number of times in a day or battle, rather than on a self-scaling system probably beginning at the meek standard of "1/day".
I think I prefer your suggestion and rather that everyone use the same type of 'pool', but not sure at the moment.

2. Treating buffs and magic durations as equipment, such as in the card/tabletop dungeoncrawl game DungeonSiege, works extremely well as a regulating method.

3. That works too, but for the purpose of user-friendly mechanics I'd prefer any system that keeps the parallel pools concept to a minimum. For example, abilities as one-shots, charge-consuming, and at-will; no fatigue-burning, charges, or multiple mana pools at the same time.

4. Here's where I disagree, outright. Limitations should be maintained universally rather than by type of ability. It's a matter of record keeping since some or many abilities could be cross-group.

5. Nice. Final Fantasy 7 works this way, by increasing your mana (and conversely decreasing physical abilities and defense) with every spell-granting Materia item equipped. The characters themselves have very few unique abilities; most 'class features' are modular and granted by Materia, and using more Materia makes the character effectively a mage-type glass cannon (although they are always geared more towards the 'balanced' archetype)

6. ... what? That's awful!

And my tuppence:
7. Active abilities that affect hostile or allied targets that remain under the control of the user must be accounted for by limitation. The durations occupy a space that could otherwise be used to enhance oneself.
Or rather, beneficial effects with duration fill a slot of the target and become under the recipient's control.
For extra limitation, beneficial effects use slots of both supplier and target.

[K]

First, use more examples.

Second, we know how well #6 goes. 4e balanced all abilities against each other and we essentially have things being zeroed.

Third, drawback systems never work. People love to find ways to optimize away a drawback or just tank something they don't need. For example, 3e's dump stat system is a perfect example of trying to create competing priorities and failing.

That leaves us with 1-4. Each works in it's own way.

The only thing you didn't consider is the difference between real abilities and token abilities.

Real abilities work almost all the time. Fighter's melee, archery, magic missiles and the like work well enough and often enough that you can count on them to work in all but the wierdest circumstances.

Token abilities only work in rare circumstances. Fire immunity is a real ability when fighting a red dragon, but you could go ten levels and years of real life before it comes up again in your campaign.

In an ideal situation, you get some token abilities and some real abilities, and choosing one does not mean you choose one less of the other kind.

That being said, things need to come from one resource. 3e, in it's search for feature and/or power creep created such things as the Immediate and Swift actions, extra combat rounds in the form of AoO farming, and endless compendiums of spells that could be layered on top of one another for THE WIN.

Remember that complexity is not the same as variety. Complex resource allocation systems give the illusion of variety and choice, but they quickly fail to pay off the extra work involved.

[Josh_Kablack]

There's also the HERO System method

Players who wish to have multiple exclusive abilities to choose from can buy a multipower full of fixed slots and get basically 90% discount on the cost of each exclusive power after the first.

That's a gross oversimplification, but the point is that "give players a large discount to define an ability as exclusive" is another potential methodology.

[Roog]

Any that I missed?

I think that #4 is not technically a sloution to the problem. Simple forms of #4 are basically equivalent to the original solution. If #4 works as a technical solution, then you don't need groups at all, as players can use the tactics described in #4 in the original set-up.
Note: #4 is still a good idea, and still helps in improving the situation in the original setup.

8. Each active ability is a paring of a exclusive ability and an non-exclusive ability,with no way to get one without the other. (This is a more limited version of #3, but avoids beeing seen to have seperate pools.)

9. Discount competing abilities with a progressive discount, to account for their reduced utility. (Although this breaks the "all abilities are balanced" goal, and you may as well give all abilities point costs at this point.)

9a. Once a certain number of competing abilities are chosen, allow a single extra choice to give multiple abilities. This only works if the marginal utility drops very quickly (like it might do with specialised pools like #4) or the uncertanty in ability balance is of the order of at least +/- 33%.

10. Every non-exclusive ability comes paired with some special penalty effect that scales in precisely the same way as the ability, and emulates the drawback of picking lots of competing abilities.