I'm curious what sorts of things other people have come across when trying to make skill gain more "realistic". The mechanic I'm attempting to incorporate is having players only able to level up skills that they actually use in a game, rather than either leveling up everything at once or being able to just level up whatever they want during level-up. Players would keep track of the skills they use and the GM would need to keep tally to keep them honest.

I'm anticipating players being right honorable jackasses and trying every other minute to use skills that totally don't apply to the situation, but I didn't know if others had tried the same idea and given up. Does it actually encourage players to try more skill-based approaches to problems?

You'll forgive me if this is an often used mechanic in other RPGs, I'm coming from a Palladium Books-based group and Palladium's skill system is royally F'd in the A, so something like this seems absolutely mind-blowing to me.

Call of Cthulhu and Runequest did this back in the Stone Age.

Burning Wheel and Burning Empires refine the ideas advanced CoC and RQ.

Mouse Guard further refines and simplifies those ideas.

And Dogs in the Vineyard has a very fresh take on advancement through use.

Hi! Welcome to the Forge!

Crossposted! I wrote this before Luke replied to me.  He knows a little bit more about Burning Wheel than I do.

The kind of system you're suggesting exists in several systems, including Burning Wheel, Call of Cthulhu (and Runequest), and several others.  It's a neat approach, because characters develop "organically", rather than artificially.  It has a couple of problems though.  Firstly, there's a lot of book keeping.  You've got to note which skills you've used, and how many times.  Depending on the system you also have to note the difficulty of the roll, whether you succeeded, and so on.  Second, there's the issue of "grubbing for skills" as you describe.  To be honest, I don't actually think that's a serious problem in most situations.  Burning Wheel embraces that "problem" by making skills require a certain number of tests at a certain difficulty to advance.  Players go looking for chances to make really difficult, daring attempts, just to advance their skill.  The tradeoff in that system is the amount of book keeping.

I'd really recommend you read as many other RPGs as you can at this stage to get a sense of the scope of things out there.  There are a number of very good free games online that are worth a look.  I'd recommend The Shadow of Yesterday (http://tsoy.crngames.com/) and Fate (http://www.faterpg.com/) as two games that have some interesting ideas (though neither of them have the system you suggest, I think).

A very simple application of this is that the GM type allocates additional advancement towards the skill they felt were most important to the story/action or had the most room for improvement. An example is that when I ran Vampire, which uses a typical players spend the xps earned as desired more or less, and I kept about half the xps they earned and spent them on the skills I thought they should most improve, while letting them use the rest pretty much as desired.

Luke, Simon, thanks for all the great references. I've heard of most of those, but haven't had a chance to explore them yet, but now I even have yet another reason to read up on Mouse Guard, besides the awesome artwork.

The book keeping aspect I'm not so put off by (I can't say the same for players), but I'm hoping the more frequent (but shorter) character upkeep will work a bit more flexibly and (like you said) organically, rather than waiting until the end of a session for the GM to drop a big load of XP on the group and then making major changes/upgrades to your character in between gaming sessions. I'm also keen on the idea of going classless so that players don't feel like their characters only get things done a certain way. I'm hoping it will act in reverse of how classes work. With classes, a player picks a fighter and expects that their character hits stuff a lot with weapons because that's what fighter's do. I'm hoping with this that the characters kind of "grow into" whatever role they end up playing. So you're not a fighter because you picked it out of the book, you're a fighter because that's how you acted the entire game.

Along with that @chance.thirteen, the idea I'm going for is to keep the player's strictly in control of their own characters and keep the GM's job strictly on affecting all non-PC elements. The system you're talking about sounds like a neat idea, but it sounds like it takes some control away from the player. Also, with the group I roll with I can imagine the arguments that come up over how the GM assigns their experience. As a player too I know I'd get possessive of how my XP gets spent (at least as possessive as someone can get over imaginary things).

Note well that an advancement system like this cannot be bolted on or exist in isolation in a system. It must be firmly bound into the fabric and goal of the system as a whole, otherwise it will be laborious exercise in game breaking and book keeping.

In BW, advancement is tied to the obstacle system, Let It Ride, artha and the practice system. Without those interlocking pieces, it wouldn't work.

-L

One thing about classless system, with classes you have niche protection, each class does something with little or no overlap, with a classless system it's more then possible that everybody ends up more or less doing the same thing.

I have always been attracted to a feature of Chivalry and Sorcery ... 3rd edition? 4th? They have skills, and they have classes, but your character class merely gives you a relevant list of skills related to it and gives you a bonus learning and using said skills. They called it vocational skills and you also got free picks. In their model these could stack with the class picks for extra capacity, but I am not sure that is necessary.

In Runequest terms, you could allow someone to pick some number of skills, and just give them +5 or even +10 to LEARN the skill when those rolls are done. It's a quick nod without being tied to anything else in particular, other than whatever goal RQ had in mind with their skill checks and learning rolls system. It may just feel unsatisfactory.

Depending on the genre, too, remember that it's not hard to rationalize "passive" skill advancement. A fighter in D&D, for instance, probably spends his downtime performing weapon drills, maintaining his armor, etc. A modern day character might spend his downtime reading wikipedia. I don't want to say that performance based skill gain is not a good idea, but it may not be as necessary as you think. It can work well if the system is built around it, but it's not always good to tack it onto an existing system, like Luke said.

Paddirn, I'm with you on wanting skills to advance based on use. I don't know how others have handled it, but here's what I'm trying. I decided that the old adage 'we learn from our mistakes' might be a good model, so in Chronoplex, each time a character attempts a skill and is unsuccessful, the character gains at least one attempt point. Attempt points are just a tracking mechanism; they serve no other purpose. When a character accumulates five attempt points he may immediately convert them into one additional point of prior training for the skill base.

I also wanted to address the fact that characters do not all learn at the same rate, nor do they have the same aptitude in the same areas. I decided if the character's skill attempt is unsuccessful, then the character receives a number of extra attempt points equal to the higher positive value of the two talents used for the skill. (Each skill in Chronoplex has a primary and secondary talent; a bit unorthodox, I know, but I like it.) If both talents are zero or less, then the character receives his one attempt point but no additional points.

(Take this with a grain of salt, though, as I have not yet playtested it.)

Pendragon had a great simplified version of the Runequest system with less bookkeeping- IN Pendragon you try to roll as high number on a D20 without exceeding your skill, and a critical success is a roll that exactly matches your skill number (a bit like Blackjack). Basically every critical success with a skill meant you got to add a "tick" to that skill. In the "Winter Phase" (downtime) you rolled a D20 , if the roll was higher than your skill, you improved it by 1 (so the higher your skill goets the harder improvement becomes). You can only have one tick at a time on each skill.

You could tweak this so that the rather than rolling at the end of each adventure, the players could roll every time they get a chance to study. So having a teacher means you can roll your ticks more often and have a better chance of improving, but having a teacher without doing will learn you nothing.

Maybe that's more random than you want. There's less book-keeping than the D20 system.

I personally don't think failure alone should be enough. Perhaps fumbles - ie you have to majorly screw up to learn someething. Also you might want to only apply that in stressful/dangerous situations (so picking a lock to your forot door because you forgot the key doesn't count, but picking the lock to a jail cell with the guard on the way does).

Perhaps for opposed rolls cannot count unless the opponent is equally skilled, superior or has a major advantage. So the swordmaster would need to track down other swordmasters or give his opponents some major advantage (the Japanese swordmaster Miamoto fought duels to the death using a wooden practicing sword versus real katanas).

Making improvement points harder to get would mean less bookkeeping.

sockmonkey wrote:
...each time a character attempts a skill and is unsuccessful, the character gains at least one attempt point...

Fable the computer game was designed around a composite system; their concern was that growing difficulty and conflict stakes plus use-based advancement would lead to finer and finer specialisation (cause you've got to use your best skill), interfering with the ability to cause abrupt changes in style of character. Now this may be seen as a bonus; how many samurai/wuxia stories do you know where a character is trying to assimilate into peaceful society when all they really know how to do it fight?

But that kind of locking in was not what they were going for, so they combined a use-based advancement system with a combat (and combo) based experience system, so that you could choose what you wanted to your character to become even while developing them in the standard way.

Personally, I would associate such points not with combat but with the ends of quest arcs/meaningful points of denouement. You know, when a character is (in terms of dramatic structure) putting themselves back together after intense experience and coming to terms with the experience. At that point I would shift player attention to freebee experience points/advances, allowing them to still make tactical choices when their character is supposed to be chilling. Depending on the game these could be for teamwork, social conscience or self-sacrifice on the part of the character during the previous arc, or reflective and personal theme addressing behaviour, or for depth of interaction with the concepts of other players, or simply for success at assigned goals and depth of change to the world.