Continuing my notes on my own version of the traditional fantasy adventure game, next up are character classes. My approach here is to preserve the 3rd edition multiclassing kernel and the way players define their characters as combinations of different classes. Meanwhile I also want to utterly decimate the D&D idea of preplanning character advancement far into the future. Likewise, experience as the primary reward mechanism receives a critical look.
The character levels depict skill sets gained by the character from experience. A “starting adventurer” has around three levels. The average human gains perhaps one level per five years of life. The character creation rules will probably allow players to start characters at different levels of experience more in accordance with character backstory.
Each character level is associated with a character class, as one might expect. The most important part of the character class is the definition of its skill set, what a character practicing that class actually learns. This is all phrased in terms of the fiction, as the game doesn’t have a set skill list or anything like that. Each character class defines Core Skills and Secondary Skills: the former are whatever is actually crucial to practicing the societal role represented by the class (each class depicts an occupation or other social role, for the most part), while the latter are whatever other things a character might become acquainted with while living as a representative of their class. It’s important to note that these definitions are realistic, not formal: the class description is there to point to the fiction, but the ultimate decision over what’s core or secondary for a given class is based on evaluating the fictional setting. Thus, if a Fighter (or whatever, I’ll probably want to use a bit more specific classes) in your setting fights from horseback, then riding horses is a Core skill for fighters in your setting.
The way these Core Skills and Secondary Skills turn into skill bonuses for various Ability checks is simple: whenever a given Ability check comes under the purview of a given character class, a character gains +1 to his check for each appropriate class level with Core skill and +½ for each Secondarily applicable class, rounded down. I don’t particularly like division on aesthetic grounds, but in real practice it’s trivial to do while playing, so I’m going to stand it for now.
Keeping skills in check
When the above skill definition is combined with the earlier Ability rules, it’s easy to see that just like D&D, the relative significance of Abilities will diminish when skills go up. (This is especially the case when this system probably has somewhat higher level ranges than D&D.) Also, if we don’t put in any checks and balances, it will be relatively trivial for a character to become the kind of rules-warranted demigod D&D has on higher levels: simply pile on enough levels and you cannot be beaten.
I’m not in principle against “pile on enough levels and never be beaten”, but I’m also interested in exploring flat ceilings for skill values. Especially I’m intrigued by defining character tools and feats (or whatever I’ll call them) in terms of upping these ceilings. Consider: what if an unarmed character could only use a maximum of +2 skill adjustment in fighting conflicts? That’d mean that barring special dispensation (like some martial artist special ability) just having a high level wouldn’t help a character, he’d need to have tools to up that limit. If a sword allowed a skill adjustment up to +5, say, it’d also say something about the limits of conventional training: after five levels as a fighter, a character simply isn’t getting any better with a sword. He is as good as the sword allows him to be. Likewise, two master fighters fighting, one having a sword and the other unarmed, will not need any other special statistics to represent the difference: the lower skill ceiling of unarmed fighting tells all. Also, an interesting way of doing magic swords: if the “magic sword” allows you to use your full skill bonus, it is the more useful for the more experienced character, and doesn’t have as much of that annoying tool-centric feel as traditional D&D magic items do.
Special features of character classes are traditionally rather important in defining what a character class in D&D means. Old D&D was especially intriguing in how the class was often more important than level. I don’t think I can go that route here, but I’d still like to have some kind of special abilities doled out per class. Whatever it is, it needs to be compatible with the feat system, whatever that’ll be; I probably can’t survive without feats in some form, so they need to be in harmony with class-delivered special abilities.
One option I’ve pondered is to simply hand-craft each class to include interesting special abilities. Making classes is fun and making characters out of those classes is fun, too, so reducing classes to just their skill set component seems like a shame. On the other hand, there are lots of problems with this as well, starting with dead levels, leveling breakpoints and class dipping for special abilities. I can personally live with the last one, but the former issues are all rather intricate design-wise.
The other option I could imagine using is to unify feats and class special abilities into a system of learning checks where characters need to pass Ability checks to learn their feats. As different classes give bonuses to gaining different feats, classes can be interpreted solely as skill packages that just happen to favour learning particular kinds of special abilities. Balancing character classes would be trivial in this case, as a class would really just be a skill-set more or less usable in different situations.
Perhaps a suitable compromise would be to give each class level a recommended feat that may be learned without a learning check, or bypassed if the character wants to pick up something different. Or simply give the character a hefty extra bonus (+10 or so) for learning certain specific feats at specific level-ups. I like the simplicity of having all characters learn one special ability / feat per level, but the solution needs to balance character customization needs (especially the need to systemize odd-ball abilities and qualities that don’t really have anything to do with character classes) and the definitional nature of character classes. They way I did this in my D&D campaign was to have characters learn one fixed special ability from their class, while the player had the option to pick up a feat per level from in-fiction adventuring as well. That results in too many feats in the long term, though.
New classes and levels
I’d say that characters should only be able to switch classes via in-fiction training. Also, pretty much all classes in this game have in-fiction entrance and/or progressing requirements that bring unpredictability to character advancement: if your character happens to fail at an entrance check or doesn’t have access to a class in his current situation, he has to develop himself in some other manner. The player cannot plan his character development in advance, because he does not know in advance which character classes are available at which points in the game, and whether his character manages to get into the class in the first place.
As for when characters gain levels, I’m a bit torn on that point. I’m personally much intrigued by several different solutions, and I can’t rightly pick one over the others:
- Experience points are clumsy, but the principle behind having character development slow down with time unless characters take on more drastic risks (which is the original reason for scaling xp costs) intrigues me. I could imagine having “leveling checks” where the player rolls against the character’s current level, with bonuses from the challenge level of the adventure at hand. I’d probably make the leveling check difficulty go up faster than challenge bonuses, too, so as to force characters into increasingly suicidal situations if they want to push their level to legendary heights.
- Just getting one level per session is fine with me, really. Or one per adventure. Character levels do not need to be rewards for overcoming challenges, I am quite satisfied to have them be an abstract campaign arc indicator that slowly changes the dynamics of the developing campaign. There are plenty of reward possibilities without making experience one of them.
- On the above note, just having characters gain experience when they develop in their craft (probably over significant down-time in-fiction) in a totally simulational manner is fine with me, too. I like the idea that you could design a 50-year old character and simply calculate a high level for him, for example. D&D’s in-fiction distinction between life experience (old characters) and experience levels (adventuring characters) never quite sat right with me.
I’m probably happiest when I combine all of the above in some manner. We’ll see.