Hmmm.  I need to post, it was probably the most well-attended roundtable at the IMGDC, and it’s been at least a whole month since I last touched upon it… eh, why not?  I’m going to just ramble about the class v skill thing a bit.

For background, Damion at Zen of Design has presented some excellent insights on this topic over the past half-year or so… links here and here.  There’s a forum thread over at on the topic at the moment, as well, for a selection of more player-centric viewpoints.

(Edit: broken links fixed)


For starters, let me present my definition of what both classes and skills are.  They represent two of the most common methods to define, represent, and control the range and power of a character’s abilities in a computer-moderated RPG, whether single player, multiplayer, or MMO.

Let me highlight a few of the points above.  Classes and skills represent just a couple of the most common methods for achieving the stated goal.  Other methods, including some that are not simply a hybrid of the two above, should be possible, and it is arguable whether any such system is truly required at all.

That is kind of where the computer-moderated proviso comes in.  Off-line, pen-and-paper RPGs have occasionally explored other methods of determining what a character will be allowed to do during the game.  Some have simply allowed the player to make the determination.  Others have restricted character abilities to that of the player him/herself.    The computer-moderated environment makes using numeric representations so very simple and obvious, however, that it is difficult to imagine breaking that paradigm.  (There are a few people arguably thinking about it, tho…)

For my part, I would like to make this definition model and the associated advancement a little less central to the CRPG experience.  My definition of the root concept of the RPG has always been “cooperative storytelling”, not “reach maximum level”.  Originally, skills and/or classes were meant to support the gameplay, not dominate it.   It would be nice to return to that perspective someday… but that’s not the topic for today.

Progression scale

As I hinted at above, classes and skills are not discrete concepts.  Most MMORPGs (and CRPGs in general) actually use some combination of the two, a hybrid system if you will, to help define the character’s abilities and limitations.  In essence, each system falls somewhere along a continuum, generally defined by how much diversity a single character is allowed to encompass, a “specialist” vs. “generalist” scale.  A crude visual representation might look like this…

Specialist Generalist

Class-based systems are usually near the specialist end of the scale, while skill-based systems generally end up near the generalist side, and hybrids can fall nearly anywhere in between. 

These are not diametrically opposed concepts by definition, either.  Multi-class systems, like DDO or Horizons for example, illustrate how blurred the distinction between classes and skills really is.   In many ways, skills can usually just be looked at as extremely narrowly defined “classes”, with unlimited (or at least copious) multi-classing allowed.

Strengths of Classes

As Damion has attempted to point out multiple times (see the links above), class-based systems do have certain advantages over skill-based systems.  The items he mentions explicitly are…

  • Player roles easy to balance, maintain, and expand
  • Able to make character choices without fear
  • Easy to advertise for group and guild play
  • PvP needs tactical transparency
  • Player roles should offer strongly varied experiences

I think it is generally safe to say that any system, skill-based or otherwise, should at least try to address the above bulletpoints in some way, or be able to explain why you have not.

Levels, XP, usage-based, etc.

These are additional topics which impact the discussion.  Levels and XP are generally considered to strongly relate to class-based systems, for example.  However, by expanding the definition very slightly, it is easily argued that skill ratings are simply levels, and points toward advancement of an individual skill are simply XP by another name.

Usage-based systems are a little more of a twist on the same old, same old… but usually not much.  A usage-based skill advancement system is arguably only a XP-based system that offers rewards on a more discrete, fine-tuned basis.  If the usage-based system only rewards “successes”, it still leaves open the possibility of receiving nothing for failing to completely “win”, for example.

The advantages of a level/XP-based design, again according to Damion at Zen of Design:

  • Quick evaluation of potential allies and rivals/enemies
  • Rewards devotion over skill
  • Needs a reason not to cancel
  • Players respond better to substantial improvement over minute improvement
  • If realism is your goal, system should not prompt unrealistic behavior
  • Players want continual rewards for their playstyle

Again, I think a designer wants to evaluate their systems against the above, and be able to address either how their system achieves the above, or at least why they don’t feel it is necessary.  I personally don’t think several of the items in this second set of bulletpoints in particular are necessarily desirable… but I do think it is important to be able to discuss why that is the case.

Disadvantages of classes/levels/etc.

I’ve focused so far on the advantages of the D20/D+D class/level/XP triumvirate… so why is there a continual debate at all?  In part because, based on the way they are most often implemented, the following limitations arise….

  • Classes can feel confining… why can’t my Warrior learn a couple minor spells?
  • Classes can be misleading… what does it really mean to be a Ranger?
  • Classes can become _too_ familiar… “this is a brand new game in early beta… why do I feel like I’ve done all this before?  Talk about deja vu…”
  • Levels can quickly divide friends into separate groups… “it’s only been a week… how the heck am I supposed to group my level 12 Spearcatcher with your level 42 Archmagi Supreme?”
  • Generic XP can create hard feelings amongst different play styles… (“Wait… so I did 2 massive quests, saved the entire frigging universe, twice I might add, and you got more XP in half the time grinding level 3 foozlewhatsits?!?”

None of the above is an insurmountable obstacle or unavoidable result of class/level/XP.  The same logic applies here as to the items listed earlier… these are topics you should be able to speak to as a designer of such a system, either as to how you’ve addressed them, or why you don’t feel it necessary to do so.  I’m sure people can come up with more issues if they try…

Not the only advancement concept possible

Finally, I just wanted to note that the Class/Skill paradigm is far from the only advancement methodology possible in the wide wide world of RPGs. 

Social advancement, either as a defined game system (EQ faction, CoH/CoV contacts, and Vanguard-style diplomacy, for example) or simply amongst the players without systemic support (guild leadership, etc.) is an obvious example.  Wealth and equipment are commonly implemented as a secondary advancement concept in most RPGs, MMO or otherwise. 

Some less commonly explored (or less emphasized) concepts might include divine favor, good/bad karma, “Force”-sensitivity, luck, collecting discrete knowledge elements (EQ spells, recipes in WoW), fame and/or infamy, and the like.

And those are just the ways of defining character abilities/limits strictly within a game system.  Tying back to the player him/herself, whether along the lines of puzzle games (“pattern-recognition”), first-person shooters (“twitch-based”), and/or other means, are also possible alternatives that could be weighted far more heavily.  There’s a lot of room for exploration left out there…

I’ll wrap this up for the moment… my next goal is to present an alternative system in the class/skill perspective, and evaluate it against the above bulletpoints. Real life rears it’s ugly head once again, however…