Rules for improve-by-doing skills

[Li Shenron]

What stops this from occurring?

First of all, the fact that skill bonus will still be limited by level.

Second, (also for [MENTION=58197]Dausuul[/MENTION]) the purpose of this idea is to have fun with seeing your skill bonus improve depending on what happens in the story, rather than just increase it up to your maximum when you level-up. So once again, a player that is just trying to find all possible ways to max the skill quickly, is a player who's not interested in this idea.

 

[Dausuul]

First of all, the fact that skill bonus will still be limited by level.

Second, (also for @Dausuul) the purpose of this idea is to have fun with seeing your skill bonus improve depending on what happens in the story, rather than just increase it up to your maximum when you level-up. So once again, a player that is just trying to find all possible ways to max the skill quickly, is a player who's not interested in this idea.

Interesting. So the goal is to de-emphasize player decisions at level-up, and focus instead on organic character growth?

I like the idea. My suggestion would be to base skill improvements on good rolls--e.g., if you get a natural 20 on a skill check, you get some number of points toward leveling up that skill. It adds excitement to the natural 20 and doesn't require a lot of bookkeeping. Then the challenge is a) balancing the skills against each other (some skills will naturally see a lot more rolls than others), and b) figuring out how to distinguish "real" skill checks from efforts by players to game the system. I don't mind having some DM judgement involved with b), but I'm not comfortable throwing the whole thing on the DM with no guidelines or support in the rules.

 

[Rune]

That's what I thought as well.

Thing is, I hate that system. Rewarding characters for what they actually do sounds nice, but it encourages players to game the system by swimming into walls, putting a heavy object on the mouse, and walking away for an hour.

In D&D, the risk is that a player would simply search every room to improve his search, constantly try to persuade NPCs of arbitrary things in order to improve his diplomacy, even drink poison every night to improve his fortitude save.

I'm honestly not sure what would prevent that outcome from occurring.

What I would do (if I were actually inclined to use such a system), is set a finite number of skills that can be checked for advancement per level.

Assuming that higher skills are less likely to be successfully advanced than lower skills (which I would absolutely do in this scenario), the player then has to make a meaningful decision about whether it is better to check for advancement to improve lesser-trained skills (thus becoming well-rounded), or risk no advancement in the pursuit of highly proficient (but necessarily more specialized) skill-levels.

 

[Li Shenron]

So the goal is to de-emphasize player decisions at level-up, and focus instead on organic character growth?

That's a nice way to describe it.

I should add one thing that I had in mind all the time but failed to mention... I had in mind that each character would still have a set number of chosen skills, so she would be trained in those chosen skills and untrained in others, and she would improve only in her trained skills.

So for example a player who would choose Lockpicking, Diplomacy, Kn:Arcana and Survival would use the improve-by-doing rules for these skills only. All other skills would remain untrained, and she would just essentially roll ability checks for them, without ever improving her scores.

This is not the only way of using this system, another DM could allow all skills to improve in this way.

The reason why I had in mind to use it this way, is because I envisioned each player being focused on a bunch of chosen skills, not unlike with the regular skill rules. So with this in mind, player decision is still there at character creation (and perhaps later on, if the general skill rules allow for example to add more skills to your list of trained skills).