Have a little class, people.

[Jared A. Sorensen]

Back to a modern example, your Role is Computer Guy. If you decide to retrieve some data off of a machine you have multiple ways to do it. Do you sneak into the place, break into the machine and yank the harddrive? Or do you connect remotely and hack through their security? A valid class system would allow you to do either as part of fulfilling your Role through the game's Classes.

In my game, this ain't an acceptable character. The point is not to pin down the character to generic role with many possible variations depending on the circumstances. Computer Guy" is someone who does computer stuff, right? But my question would be, "Great -- but what is he DOING?"

One answer could be, "He's giggling, a half-smoked cigarette hanging from his lip, as he wrecks some dude's credit rating." So he's not just Computer Guy -- he's the unshaven guy who fucks with people via his cable modem. He's the...oh, I dunno..."Malevolent Hacker."

I think part of the "multi-class" character (ie: Computer Guy who could hack remotely, drop in Ninja style from the air ducts, charm the pants off some executive, etc. etc.) is fear of not being able to do stuff in the game. Like taking a "skiing" skill during character creation and then finding yourself in the jungle. It's frusterating...and many games feature a "feel free to alter your character concept during the game to fit the scenario" kinda rule to fix this.

[Clinton R. Nixon]

I think part of the "multi-class" character (ie: Computer Guy who could hack remotely, drop in Ninja style from the air ducts, charm the pants off some executive, etc. etc.) is fear of not being able to do stuff in the game. Like taking a "skiing" skill during character creation and then finding yourself in the jungle. It's frusterating...and many games feature a "feel free to alter your character concept during the game to fit the scenario" kinda rule to fix this.

This right here just gave me an idea, somewhat combined with the "metagame" thread going on right now, too. A high action game like octaNe or a spy game or whatever could easily be made where your abilities weren't things like a "skiing" skill, but instead abilities based off very broad categories like Movement, Destruction, Stealth, and Thinking. The point would be to never have your character de-protagonized because of, say, having that ski skill and being in the jungle. With Movement, you ski like a madman on the mountains, and in the jungle, well, you can climb trees or drive one of those Everglades hover-fan-boats or whatever. Point being - you can move.

[Jared A. Sorensen]

This right here just gave me an idea, somewhat combined with the "metagame" thread going on right now, too. A high action game like octaNe or a spy game or whatever could easily be made where your abilities weren't things like a "skiing" skill, but instead abilities based off very broad categories like Movement, Destruction, Stealth, and Thinking. The point would be to never have your character de-protagonized because of, say, having that ski skill and being in the jungle. With Movement, you ski like a madman on the mountains, and in the jungle, well, you can climb trees or drive one of those Everglades hover-fan-boats or whatever. Point being - you can move.

Dude. That's so right on -- look at James Bond! Speedboats, horses, skis, planes, helis, motorbikes, race cars...he can do it all. The trick is his intention: he's not a chase guy, he's an escape guy.

Hmmm...really good idea. I'd love to be able to use that in octaNe but hmm...how to do it? I think the idea needs its own system and unfortunately, I dunno octaNe is compatible with it. Very cool, though.

[Mike Holmes]

A high action game like octaNe or a spy game or whatever could easily be made where your abilities weren't things like a "skiing" skill, but instead abilities based off very broad categories like Movement, Destruction, Stealth, and Thinking. The point would be to never have your character de-protagonized because of, say, having that ski skill and being in the jungle. With Movement, you ski like a madman on the mountains, and in the jungle, well, you can climb trees or drive one of those Everglades hover-fan-boats or whatever. Point being - you can move.

See Sean Wipfli's Two Page Action Movie (2PAM).

Mike

[Ron Edwards]

Hiya,

What a great discussion.

First thing for me is to acknowledge Fang's writings about all of this, as his "Particles of character class" makes a lot of sense to me. Second thing is to acknowledge that Jared is spot on with his shift of focus to "role," in the active, in-play sense of the word - and to say that Fang and Jared are saying very similar things.

Looking at my own games, I've noted that they are stupendously "narrow" in terms of role. Yes, your Sorcerer character may have all manner of interesting descriptors for (say) Will or (especially) Cover, but every character is, god damn it, a sorcerer. Same goes for Elfs - you may play a genital-stage, anal-state, or oral-stage Elf, but it's pretty much just coloring around the edges in minor system-affecting ways. And soon, you will all see perhaps the acme of focused RPG design in Trollbabe, in which nothing is distinctive between characters, in terms of system - plus, they are all one "race" (trollbabes) and (gasp!) one sex.

InSpectres and Little Fears both demonstrate the same approach. So does Zero, if you'll believe it - what appears to be brutally-defined "character class" at the outset of play is exactly what actual play, over time, deconstructs. So does the original Pendragon (although with a different GNS orientation).

This design preference on my part is related to my intense interest in Narrativist Premise, and I can't say that such a focus would be appropriate for other design aesthetics. I do think, though, that offering a whole smorgasbord of Roles (rather than multiple nuances on one Role) has proven, in my experience, to be a poor default.

For instance, some discussion after fairly intensive L5R play revealed, to us anyway, that the game would vastly improved by removing shugenja as player-characters.

Now, does that mean that I think only single-focus games work? Not at all. The distinction between thraka and tala in Orkworld is a wonderful part of the game; the panoply of possible Roles (and I don't mean "occupation" or "race") is one of the great strengths of Hero Wars. I count both games as tremendously successful. The two classes in The Riddle of Steel seem all right to me, pending some play to learn more. (My apologies to Jake, but sorcerers vs. non-sorcerers are character classes in TROS. Squirm if you like, but they're there.)

So I conclude that the topic needs yet more workout. Why are "classes" unsatisfying in many cases? After all, a game isn't automatically better because it has only one role - or because it has many roles.

Let's break it down. Let's say we have a game which does work well with multiple roles - not Little Fears, not Sorcerer, etc. Why does the traditional approach to character class turn out - in many cases - not to work well for it? By contrast, does it ever work well, and why?

I strongly suspect that the answer lies somewhere in the parsing of in-game vs. out-of-game roles. I also have a sneaky suspicion that protagonism is going to be involved here as well, as was implied above.

Best,
Ron

[Gordon C. Landis]

Why do classes break down?  One observation I have is that classes are supposed to tell you what you do (often, what you do WELL), but in practice, it doesn't always work out that way.  The classic "but the thief can do more damage than my fighter" or "the magician can sneak better than my thief" kind of thing.  Classes set up an expectation of what things are going to be like, and rarely (at least over time) does play live up to those expectations.

"At least over time."  Maybe that parenthetical is important . . . maybe "classes", as often constructed and used, become a form of "playing before you play," of resolving things ahead of time that would be better resolved as the play evolves.  And when play evolves (as it especially does in Nar games), the inappropriateness of that "play before play" just becomes more and more glaring . . .

Gordon

[wyrdlyng]

Why are "classes" unsatisfying in many cases? After all, a game isn't automatically better because it has only one role - or because it has many roles.

True. From my experience I have found that the instance where "classes" fail is when a game tries to give focus the classes in on the point of the game but doesn't finish the job. For example, going back to D&D, the point of play is to kill things and level up. Now let's take the role of Fighter. If the role of the Fighter is to kill things with weapons then why dilute that with 4 different classes? You have a generic warrior and 3 specialized forms (Ranger, Paladin, Barbarian). Either create 1 generic Fighter class (a la the original D&D game) or create a class that be customized to fit variable concepts (Fighter with wilderness skills, fighter with holy powers, fighter with berzerk, etc.).

Let's break it down. Let's say we have a game which does work well with multiple roles - not Little Fears, not Sorcerer, etc. Why does the traditional approach to character class turn out - in many cases - not to work well for it? By contrast, does it ever work well, and why?

Let's throw Shadowrun around for a bit. You can say the the role of the characters is Shadowrunner but I argue that that's too broad of a role to apply to the entire game. That's like saying that Feng Shui characters have the single role of "Ass Kicker."

Shadowrun has no character classes but has optional archetypes. There are obvious roles in Shadowrun: Killing Guy, Driver Guy, Computer Guy, Magic Guy, etc. If two people chose the role of Killing Guy they are allowed to customize their characters enough so that there would be no real problem with deprotagonization (is that a word?).  The reason being that there are no hard classes to limit you. If each archetype listed in the book was turned into a hard class then you'd have pointless variations for variation's sake.

Yes, there is a big difference between the two. Shadowrun's roles are, to me, Out-of-Game roles and D&D's are In-Game roles. Shadowrun's roles are roles derived from Action films and TV (like the A-Team) and which are not explicitly stated in Shadowrun but are usually obvious to players familiar with Action films and TV. D&D's roles are created by the game and its point of play (HTH guy, trap guy, ranged guy, healer guy).

An example of a game with multiple roles which works well with classes is Alternity. The classes are broad enough to allow great customization and variation within a class.

A single role game with classes that works well is... crap, I can't think of any off the top of my head. Let's take L5R stripped of the Shugenja. The role would be you are a Samurai. The classes would be the different Clans. Each player's role would be the same but how they did that role would be differentiated by how each one would try to achieve their goals according to their Clan (Crab would beat you with a stick, Scorpion would blackmail you, Crane would get your daimyo to tell you to back off, etc.).

Did anything I just say make sense?

[Walt Freitag]

Classes as "playing before you play"

Yes, I think that's very true. This also goes back to Wyrdlyng's point that classes can easily be overloaded. That is to say, class rules tend to go beyond what's needed to define the role (or the niche, nice terminology Fang) and also attempt to encode effectiveness in that role.

Even when playing with primarily Gamist goals I found this annoying about AD&D. When my fighter character would go up a level, someone had already decided how his increased effectiveness would come about: more hit points, more weapon proficiencies, higher chance to hit, more attacks per round and so forth. Maybe I'd prefer to have greater attack range, increased ability to dodge, improved tactical perception, and quicker healing instead. Sorry. Those decisions have already been made.

Champions was a big improvement in that regard. Okay, sure, the Champions/Hero system has effectiveness attractors that all but force your character into certain optimal configurations. But at least you have the gamist pleasure of figuring out what those attractors are for yourself.

There's something else about classes that I have a problem with, something extremely basic that hasn't been mentioned yet. It's the timing. This is something I've been shouting (to no avail, of course) in the computer game design community for years about CRPGs: "I start up (or log onto) your game, and the very first question it asks me is what class I want to be. This decision is by far the most complex, the most interesting, and the most important decision I'll ever make playing your game. And when do you have me making it? At the very beginning, before I know anything about the setting, the situation, or the community. Does anyone else see a problem here?"

I think the same problem is common in RPGs. Classes exacerbate the problem because they always force more to be decided earlier.

- Walt

[Zak Arntson]

but instead abilities based off very broad categories like Movement, Destruction, Stealth, and Thinking.

I've took a "things you DO" approach to my half-baked Metal Opera. http://www.livejournal.com/talkread.bml?journal=zaka&itemid=46839 In it, you have three stats that work fine for something like James Bond (granted, you'd have to change the stats from Rock 'n Roll, Kicking Ass and Jury-Rigging ... to, I don't know, Suave 'n Debonaire, Stunts 'n Fighting and maybe something else).

Let's say we have a game which does work well with multiple roles .... Why does the traditional approach to character class turn out - in many cases - not to work well for it? By contrast, does it ever work well, and why?

In traditional gaming, I think a lot of the reason it breaks down is the traditional pre-designed scenario format. The designer can't realistically predict the Classes presented, so the adventure inevitably leaves out a role or gives too much influence to another.

When does it work well? I would venture that it works well with loosely pre-determined scenarios. Ones where the GM and/or Players can change the situations (upcoming or current) during play in order to support the different roles. octaNe falls into this category.

[:: Declare ::]

I haven't seen anyone mention this, so I'll throw it out as food for thought:

Seems to me in Narrativist games players can whip up a concept and then hit the ground running; in G/S games, the players might need some help wading through the crunchy bits, so as far as G/S-type rulsets are concerned, might not character class be a mechanic used primarily as a means for the players to quickly man-up and get in the game?

[leomknight]

Interesting topic. Kind of a flashback of all my roleplaying experiences.

It seems to me that, at the one end, you have the extremely rigid character classes like old D&D. You are a wizard, you can' t wear armor, etc. At the other end, there are the skill driven systems like Runequest, aparently without classes, but characters tend to have similar skills. Then there are general systems that tend to sacrafice detail for ease of play.

In old D&D, you were a character type, you played that role, and if you were say, a thief, and never picked pockets, but fought (badly) all the time, eventually you would be an umpty-ump level thief who can pick God's pocket, but still can't fight worth a damn. This is how my first long term D&D character went. Once a thief, always a thief. I think a lot of the early roleplaying games were created in frustration with the rigidity of the rules.

In D&D3 you now have the option of changing classes almost at will. The thief above could become a fighter or monk, and gain new, more useful abilities. The classes now remind me of the "package deals" in games like GURPS and Hero System. It's a start, a clear, easily described role, with clear abilities.

The biggest problem I see with classes, is that they must still be defined in terms of what that class can DO. If I'm a 15th level nerf-herder, what does that mean to me, as a player? It still has to be broken down. How well do I attack? Cast magic? Spot traps?

This is why I prefer skill based systems. I don't have to stop and define a role, just let the player choose his abilities. If he wants new or better abilities, most games allow this, at a price. This seems to produce happier players. They aren't shoehorned into a frustrating, over limited role. They may not be able to do everything yet, but they know they have a chance in time. I ran a long term Runequest campaign. True, all the characters had similar abilities, but there were differences in focus. One player really got into being the best swordsman. Another focussed heavily on magic. A third had lots of secondary skills for search and stealth. They did this by CHOICE, not by game designer's fiat. And if they wanted to change focus, they could, no muss, no fuss.

[Mike Holmes]

...in G/S games, the players might need some help wading through the crunchy bits, so as far as G/S-type rulsets are concerned, might not character class be a mechanic used primarily as a means for the players to quickly man-up and get in the game?

This is a good point. I think it was Mearls who pointed out that the Chineese menu method of character generation (pick one class, pick one race, pick armor, pick weapon) was one of the easiest ways to get the sort of people into a game that would otherwise balk at complex Character Creation mechanics. That being said, this could been seen as enabling a player who does not want to care much about the game to participate. Which is only good if you don't mind such players.

Mike

[Ron Edwards]

Hi Declare,

That is an excellent point. Let's see what I can make of it, with further breakdowns ... OK, spectrum time.

Far end: classic character classes. Class A does X, Y, and Z, and for all intents and purposes is a rules-set unto itself, including reward system and improvement details. Good ol' AD&D in those late 70s books does it this way.

Far-ish: Archetypes. Very similar to the above except with customizing modifications built in, and perhaps more across-class rules apply during play than within-class rules. Examples include Deadlands and Feng Shui.

Somewhere middle-ish: Chinese menu. The usual lists are "race/breed" and "occupation," with perhaps a few others like "magic," "personality," or "accessories" tossed in. Vampire, L5R, and about a bezillion others do it this way, or perhaps a little bit toward the next category. Hero Wars does it in an interesting way. D&D3E is mainly in this category although it pretends to be at the far end.

Near-ish: Build-your-own. Characters include intra-character categories like attributes, skills, etc, and some currency of character creation permits them to be built as desired (more or less). Champions, GURPS, Fudge, and a ton of others are in this category.

Near end: near free-form. Not only are there no uber-categories, but not even intra-character categories either. Make it up and spend whatever needs to be spent to describe it. The Window is about as tight against the end of the spectrum as possible.

I want to emphasize that all of these are potentially functional. Furthermore, lots of other independent variables can be factored in as well, such as the role of Fortune in character creation, and all sorts of Currency issues.

Best,
Ron

[unodiablo]

Hey Mike,
Thanks for mentioning 2PAM. :) I was about halfway through when I thought of it, then you were pretty much next post!

I don't have much to add to this conversation, but I have a question... Back in the day when we were still playing D&D, we used to use this old Dragon article that let you make your own unique 'class' by picking from the list of what any D&D char could do, and then you added up your total from those you chose and got a exp point modifier.

So you could dump the things you didn't want, and add other little desired tricks from other classes... (i.e. a wizard with pick pockets and cleric spells) Was this us trying to drift towards Narrativism, or at least away from Simulationism, or do you think it was just a way of giving more 'protagonization' ability to the characters?

Sean

[xiombarg]

This right here just gave me an idea, somewhat combined with the "metagame" thread going on right now, too. A high action game like octaNe or a spy game or whatever could easily be made where your abilities weren't things like a "skiing" skill, but instead abilities based off very broad categories like Movement, Destruction, Stealth, and Thinking. The point would be to never have your character de-protagonized because of, say, having that ski skill and being in the jungle. With Movement, you ski like a madman on the mountains, and in the jungle, well, you can climb trees or drive one of those Everglades hover-fan-boats or whatever. Point being - you can move.

Would it be bad for me to mention that this sounds like an alternate version of the ability break-up for Amber Diceless? You've got Warfare, which covers everything from using a sword to commanding a clone army, and you've got Psyche, which covers everything from mental combat to lighting a cigarette with magic, and the less-broad Strength (ability to take damage, raw physical strength, and weaponless combat) and Endurance (ability to heal, tie-breaker for all the other attributes).

And I'll note that the most common change made by GMs to Amber is to change the attributes. By changing the attributes, you chance the different niches that characters can aspire to. For example, in a very Oberon-centered game (Oberon is the father of all the Elder Amberites), I added "Favor" as an attribute, which represented how much Dad liked you, how much responsibility for Amber you were given, and how well-liked you were in court -- opening up the possiblity for the "courtier" niche, while someone like Benedict is off specializing in Warfare.