Skill-based Systems

Okay, I threatened to rant about skill-based systems, so here it is: I don't like them.

My reasoning has nothing to do with skill-based systems in and of themselves, but rather has to do with the way they are used, which almost inevitably gives way to munchkinism (and, yes, I'm well aware of the irony of a D&D rules-lawyer ranting about munchkinism).

See, the complaint against class systems I've heard most often goes along the lines of "yes, hours of killing goblins is definately going to make me better at picking locks", or words to that effect. In other words, the complaint that, since the level gives benefits out in a package, some of the corner cases are really silly. It IS stupid that the book-worm arch-mage should necessarily be a bad-ass combatant simply because the system requires him to be high level to cast high level spells, but high level characters are all tough. And, again, the ninja grandmother falls into this category as well.

However, the other game design system posits a set of skills for characters and a pool of XP that can then be spent to increase the skills that the player wants to focus on. (Okay, there is a third advancement scheme, but it's less common. I'll get to it later.)

The problem with the skill-based system in practice is that players will inevitably determine the most "useful" skills. Where useful is defined based on the character they want to play. So, for the combat machine it will be the brawl and firearms skills. For the decker it will be the computer skills, and so on. Either way, there will be a small set of skills chosen, and every single XP that is awarded will then necessarily be placed into these core skills. Everything else will be ignored.

For me, the system that most shows this trait is Vampire, where in almost every game I've ever run or played, every single XP gained went towards the next discipline point, or to one of the skills used to activate those disciplines. In general, combat bunnies started with good skills, and then boosted their characters through disciplines. Occultists started with good occult skills, then added lots of disciplines, and so on. The only exceptions I can recall (and these really are the only ones) came about because the Storyteller was really tight with XP. As a result, it became pointless to try to boost disciplines, since the game would be over before you got anywhere.

Pretty soon you've got a combat machine who is unkillable in combat, but who is likely to crash a stationary car. The decker can solve the P=NP problem before breakfast, but can't tie his own shoe-laces. And the mage can cast an infinite number of wishes per day, but if you ask him to recount some of the basic points from those esoteric tomes he had to study to get this level of skill, he can't, because "Arcane Lore" is a skill that never actualy gets rolled. (Throw in merits and flaws, and the problems get even more rediculous, but that's another rant entirely.)

This would be fine for a few characters, since such specialists do crop up occasionally. However, when it's every character in every game that is so specialised, it becomes irritating. Furthermore, with such specialised characters it becomes difficult to design appropriate challenges. To challenge the combat bunny in a firefight, you need opponents who will slaughter the decker and the mage in an instant. A computer system that challenges the decker is impossible for anyone else. And woe betide the group that needs to rely on their back-up mage because John's kid is sick and he can't make it to the game.

Finally, there is no way to create a viable generalist in such a system, since the munchkins will just roll over you.

I did mention another advancement scheme. This is the one where a critical success on a skill roll naturally increases that skill. Over time, such rolls cause the skill to organically increase. Thus, the skills that get used go up, while the rest stay static. Most such systems also allocate a general pool of XP, which can be spent as above.

These games tend to be slightly better, but only just. The problem here is that a character is specialised by what he does, so instead of identifying the most useful skills and pouring all XP into those, the player just lets the system do it for him. That said, since most characters do make use of non-core skills, these do increase somewhat as well. Still, before too long you can get to the same state, with a group of specialists so skilled that they can't be challenged in their field, but are horribly vulnerable outwith that field.

Anyway, that's my problem with skill-based games in a nutshell.

How would I solve this? I'm inclined to think that skills need to be rated as "core", "non-core" and "incidental". XP is then awarded in three pools. 65% of XP could be assigned to core skills, 25% to non-core, and 10% to incidental skills. XP could be dropped down from more important to less important skills, but not moved the other way. So, core XP could be assigned to non-core skills, but not the other way around. This forces players to pay some attention to the less useful skills, and thus to build more rounded characters.

There are several problems with such a system. Firstly, the definitions of core, non-core and incidental skills are arguable. What if the player and GM disagree? Secondly, there need to be rules controlling how high these skills can become, since a clever munchkin would declare his selected "useful" skills as incidentals, and pour all XP into them, thus bypassing the system. Thirdly, there needs to be a mechanism whereby a core skill becomes non-core, or a non-core skill becomes a core skill. It is not unreasonable to expect people to change over time. Finally, this system is tending back towards class-based design, since all combat bunnies will have a lot of core skills in common, as will all deckers, mages, or whatever.

Thoughts?