Rules for improve-by-doing skills

[GSHamster]

One idea to keep it in check might be to shift skill points instead of increasing them. If you spend time picking locks, your lockpicking skill increases, but another skill (perhaps the one used the least if you can keep track of it) decreases.

Thus activities shape your character, but you don't do silly things just to increase a skill.

 

[Sepulchrave II]

The "Skills in 5e" thread got me back thinking about an old subject.

What is you experience on rules/systems where your skills do not improve automatically by level, but improve only if the character uses them?

What RPG uses something like this? Is there any at all?

Have you used similar rules for something other than skills?

As other posters have noted, the BRP system of Runequest and CoC used this model; I can't say I'm a fan. It takes the granularity of 'advancement' to an extreme which I find unpalatable.

I think there is an analogy to be drawn between early edition "gp-for-XP" and later edition "increase any skills you choose (to a defined limit) as you level" - inasmuch as the gp-for-xp rule is simply best viewed as an abstraction of how much adventuring you've done. If you've accrued lots of gold, that means you've done lots of adventuring - hence, your level is higher.

Likewise, in later editions, if you've levelled, it means you've gotten better at a "bunch of things" - allocating those skill points is within the purview of the Player. Making skill advancement contingent upon specific uses of a skill enforces a degree of simulationism that is rather extreme - and this is from someone who is generally most comfortable within the 3.x ruleset. Nor does it allow for assuming "off-screen" use of skills. And I think it disempowers the Player. It's all a question of where you draw the granularity bar in terms of advancement.

Awarding XP for clever/inventive/extraordinary skill use is an abstraction which suits me better, as it feeds into the "getting better at a bunch of things" motif whilst empowering Player choice in the specifics of where the mechanical improvement actually happens (choices which are made when the character levels).

A better solution might be to have story-based awards which target specific, extraordinary uses of skills (e.g. the character gains 1 rank in the Bluff skill if he executes some amazing scam on the Baron). These might even allow a character to exceed normal skill limits - although some other hard cap would still be necessary.

 

[Li Shenron]

Say a rogue player simply decides to wait for nightfall and pick every lock he can find in each place he goes. Sure, some of them probably are the same, but the bottom line is you potentially have a player derailing the game to do something that is defensible as in-character and yet is metagaming and is really a problem.

Ultimately, instead of simply giving advancement for skill use, I think you'd have to give XP for meaningful or challenging skill use, but that has to be operationally defined for each skill, at which point there's just too much text devoted to this idea. I just think it's too hard to implement skill training by skill use or the like, which is why people are struggling to name PnP rpg examples.

Yes I had this scenario in mind already, but it could be handled exactly like you say: the challenge has to be meaningful.

I don't know yet the solution, otherwise I would not have needed to start this thread! But I do have something on my mind...

For instance, the system could be so that "easy" tasks would never increase your skill, thus making it pointless to spend time with them, unless you're actually interested in the outcome of using the skill i.e. you want the lock open because you need to get past the door / get the object in the box. Impossible tasks also would not improve your skill, so you would need to find "hard" tasks for the purpose of raising your score.

Just as a mental exercise... There might be a few categories of tasks, for example:

- "Trivial" tasks, automatically succeed without a roll, no chance of skill improvement
- "Easy" tasks, there's a chance of failure/success, but no chance of skill improvement
- "Hard" tasks, there's a change of failure/success, and a chance of skill improvement
- "Improssible" tasks, automatically fail without a roll, no chance of skill improvement

There can be a simple formula to determine which category does a task fall into, depending on your skill bonus and the task absolute DC. For example:

skill bonus > DC-2 : Trivial (in fact you succeed on a roll of 1)
DC-15 < skill bonus <= DC-2 : Easy (success rates varies from 35% to 95%)
DC-20 <= skill bonus <= DC-15 : Hard (you succeed on a roll of 15+, 30% or smaller chance)
skill bonus < DC-20 : Impossible (in fact you fail on a roll of 20)

As a first attempt, you may rule that the chance of improvement is identical to the chance of success (but is there in the first place only for "Hard" tasks).

This example should make it a waste of time to go around picking every lock in town or climbing every tree, unless you find one that is really challenging. If it's not enough, you can also declare that a similar lock or tree won't grant you the improvement twice (so even if you're in a forest with hard trees or a city with tough locks, you won't improve more than +1).

Or as an alternative, you can narrow the "window" for Hard tasks, e.g. make it so that it has to be a task with DC = skill bonus + 20, i.e. you succeed only on a roll of 20. This means, if you succeed then your skill bonus improves by 1, but immediately the same task ceases to be Hard for this purpose!

In Burning Wheel you don't test unless there's some penalty for failure.

And that's also something I had in mind. It's a good idea, but won't work for every skill, because skills such as Knowledge, Profession and more, just don't have penalties for failure except in a broad sense.

 

[LostSoul]

And that's also something I had in mind. It's a good idea, but won't work for every skill, because skills such as Knowledge, Profession and more, just don't have penalties for failure except in a broad sense.

Yeah. I always had trouble trying to figure out what a good failure was on a failed knowledge-type test. I could have gone "metagame" heavy - where the connection between the skill task and the failure isn't great so instead you play up failure of intent, eg. you're trying to escape some bandits, so you make an orienteering test to remember where the best path through the wastes lies; you fail, so your horse goes lame - but I'm not a huge fan of that, in general. (Not sure that "metagame" is a decent term for this sort of thing.)

 

[Ahnehnois]

Yes I had this scenario in mind already, but it could be handled exactly like you say: the challenge has to be meaningful.

Alright. Here's my (simpler) problematic example:

Character wants to improve his Jump (or athletics/acrobatics). Character sets up a jump across a significant distance over a moderate fall. He has to jump far enough that it's hard to roll. He falls far enough that he takes damage, but won't die. Each time his jump improves, he makes the task slightly harder. Soon, his character has a monklike +50 Jump.

What stops this from occurring?

 

[Bluenose]

Alright. Here's my (simpler) problematic example:

Character wants to improve his Jump (or athletics/acrobatics). Character sets up a jump across a significant distance over a moderate fall. He has to jump far enough that it's hard to roll. He falls far enough that he takes damage, but won't die. Each time his jump improves, he makes the task slightly harder. Soon, his character has a monklike +50 Jump.

What stops this from occurring?

In Runequest, the 1d6 damage which is the minimum you can take from a fall is applied to a random hit location. Only a small proportion of characters have enough hit points to be sure they won't break an Arm, Leg or Head with that much incoming damage. Natural healing rates for body parts reduced to zero or below are measured in weeks. It is not a safe practice.

Also, in RQ/CoC/BRP in general, high skills are harder to advance. When you check to see whether your skill advanced you need to roll *more* than the skills current value, which means it gets progressively harder to improve skills. Though I should also note that the more recent versions of RQ have moved away from this system, in favour of allocating a certain number of experience points that you can spend to make a skill improvement check or for other things, partly I suspect because of the problem with people attempting skills just for the chance that they'll get an experience check.

 

[sheadunne]

Alright. Here's my (simpler) problematic example:

Character wants to improve his Jump (or athletics/acrobatics). Character sets up a jump across a significant distance over a moderate fall. He has to jump far enough that it's hard to roll. He falls far enough that he takes damage, but won't die. Each time his jump improves, he makes the task slightly harder. Soon, his character has a monklike +50 Jump.

What stops this from occurring?

With most systems, you're limited to advancement during a set period of time, no matter how many times you use the skill during that period of time.

Using the leveling system in D&D you would only get to increase your skill once per a level, no matter how many times you used it during the course of the level.

That's the way it works in CoC. The GM decides the breaking point in the story where players have the opportunity to increase their skills, but it doesn't matter how many times they used it, they can still only increase it once, until the next story break.

 

[Ahnehnois]

[MENTION=49017]Bluenose[/MENTION] & [MENTION=27570]sheadunne[/MENTION]
Interesting stuff. I'm genuinely unfamiliar with this type of advancement and am not asking questions rhetorically.

The break point example is an interesting one. DM judgments are best to avoid if possible IMO, but something like that is at least feasible to implement; it's not that different from deciding when you can stop and level up.

***

Clearly, other rpgs have less forgiving health systems; in D&D damage is trivial, and it's easy to set up circumstances in which a character risks lethal damage from a failed check but doesn't really care, making it hard to enforce objective limitations on what constitutes a meaningful application of skill.

OTOH, the slower advancement is also an interesting point. D&D doesn't have that kind of tiered advancement where abilities get harder to improve the deeper you get into them, but there are a few good ways of doing that.

 

[LostSoul]

Alright. Here's my (simpler) problematic example:

Character wants to improve his Jump (or athletics/acrobatics). Character sets up a jump across a significant distance over a moderate fall. He has to jump far enough that it's hard to roll. He falls far enough that he takes damage, but won't die. Each time his jump improves, he makes the task slightly harder. Soon, his character has a monklike +50 Jump.

What stops this from occurring?

I think that's getting into "there's no roll needed" territory; you're just practising, so you'll get a +1 to your check after (whatever amount of time is necessary based on your skill).

 

[Dausuul]

Before debating the best way to create such a system, I think it's important to ask what the goal is. What are you trying to achieve by implementing an improve-by-skill-use mechanic?