Forum changes: Editing of posts has been turned off until further notice.

Main Menu

Attributes or skills, not both?

Started by Logan, May 23, 2001, 04:49:00 PM

Previous topic - Next topic


I think I missed this rant, so somebody please fill me in.



Ron Edwards

OK, hold on everybody, I'm combing GO for this. I just saw it, too, so I know it's there.


Ron Edwards


Damn it! I couldn't find a fucking thing on GO today, no matter what I was looking for. So below are some transcriptions from the Sorcerer mailing list archives, my first set of discussion about the matter.

Logan, you will note that the following material is very much a sub-set of Currency, because we're talking about nuances of the category of Effectiveness. My objection to having both attributes and skills is that you end up with multiple ways to get a similar value (weak guy, high skill; strong guy, low skill) and these ways are almost never mathematically equivalent - thus you get the potential for optimization.

And for everyone, this was strictly a Narrativist topic - so that when I say, below, that a game "does best" to use only attributes or only skills, I'm talking about Narrativist goals only. This WAS on the Sorcerer mailing list, after all - our attentions were pretty focused.

***** Here's the original post ******

Does an RPG ever need BOTH characteristic scores AND skill scores? Sorcerer is built similarly to Amber, Over the Edge, Everway, and the old Fantasy Trip, in that any skills known or assumed to be known by the character are handled straight off characteristic scores. There is no formal skill list. (all of these games have at least an adjective or two to describe the character's field of expertise, providing some limits)

****** here's where it got resurrected **********

Here it is: does effective character design in an RPG need BOTH quantified skills AND quantified characteristics?

And let me clarify: yes, Sorcerer has "skills," in the form of the Cover. Yes, Amber has "skills," in the form of Warfare. Yes, Everway has "skills" in the form of specialties. Yes, Over the Edge has "skills" in the form of one or more attributes. BUT! None of them uses a NEW mechanic to quantify them -- skills in all of these RPGs are "another characteristic" and use the same characteristic-based mechanism to determine success or failure.

Same applies in reverse for lots of other games: Castle Falkenstein, Zero, and others all have NO CHARACTERISTICS, just skill lists. And yes, sometimes a given skill seems a lot like an attribute, like Brawn in Zero, for example. But again, in numerical terms, such things are treated exactly as "another skill."

My claim/suggestion is that an RPG does indeed do well to define a PC with one or the other method -- but that having BOTH (two separate sets of numbers, often at different scales, each of which are usually combined in some way to determine how "good" you are at something) is actually obstructive. I think it cannot help but contribute to mini-maxing and other abuses.

Exceptions (very weird, non-traditional ways to resolve actions using skills and characteristics): Talislanta, Shattered Dreams, Ars Magica (1st ed). More on these later if necessary.

** The following is from some ensuing discussion, when challenged for examples of how these "one-sided" systems could possibly function reasonably during play***

let's take Sorc for an example. There you'd have a guy matching his high Will against the other guy's Cover (e.g. Gambler). As a GM, I might provide a penalty on this, especially if it were on the gambler's turf, but it's certainly possible - the rules allow the scenario above to happen with no problem.

Same goes for (e.g.) Castle Falkenstein, a skills-only system. In that game, whatever skill you DON'T have at a stated level (Excellent, Poor, etc), you automatically have at "average." So the non-gambler guy might get a good draw of diamonds (= roll for mental performance, in other games; CF uses playing cards) and so hose the gambler. Or maybe he simply changes the confrontation by matching his high skill at Psychology to the gambler's average skill, which is pretty much what is described above.

See? The idea is that a characteristics-only system or a skills-only system does NOT prevent that scene from occurring. In fact, in practice, I've found it's the traditionally-designed games (GURPS being the most obvious example) in which the players can't improvise well and merely scan up and down their sheets to discover what they're "allowed" to do. The Sorcerer (Amber, Over the Edge, Everway) method or the Castle Falkenstein (Zero) method permit lots more imaginative and flexible ways to deal with threats, I think.



joshua neff

interesting...i had an idea a few weeks ago of running "fudge" in which there were no "abilities" as such, nor "skills", "backgrounds", "merits", "flaws", etc, just a list of characteristics in which any & all of these could be a characteristic (not a set list--players create their own characteristics), rated on the fudge scale, so you could have a character that looks like this:

quick-witted: great
scrappy fighter: good
contacts: superb
willpower: mediocre
occult knowledge: superb
avoid being hit: good
sharp tongue: great
keep mouth shut: poor
finances: mediocre
streetwise: good

now, correct me if i'm wrong, but isn't this basically how "hero wars" works?

[ This Message was edited by: joshua neff on 2001-05-23 18:21 ]

"You can't ignore a rain of toads!"--Mike Holmes

Ron Edwards

Yup, that's Hero Wars. Everything's an "ability," including Wealth and (very importantly) Relationships. Everything operates in the same way and on the same scale. And given the Augment rules, you can boost any ability by using any other, given the appropriate circumstances.



On 2001-05-23 18:20, joshua neff wrote:
interesting...i had an idea a few weeks ago of running "fudge" in which there were no "abilities" as such, nor "skills", "backgrounds", "merits", "flaws", etc, just a list of characteristics [ This Message was edited by: joshua neff on 2001-05-23 18:21 ]

In fact, Fudge -OUGHT- to work like that. The attempt to seperate skills and attributes is -very- clearly kind of a clugey add-on. I'm not sure why they thought it was necessary, as it is so obviously not central, and is very clunky in practice.

                     - James

Ron Edwards


I agree entirely about Fudge. Both times I ran it, that very aspect kept tripping me up. Its existence leads to all sorts of totally-unnecessary point-balance blather in the text. And what's up with all that "scale" stuff, anyway? Rhetorical question.

Back to our regularly scheduled program ...


Blake Hutchins

Castle Falkenstein also works this way, as has been mentioned previously.

5-29: And d'oh. Mentioned previously in this post. How'd I miss that? Apologies.


[ This Message was edited by: Blake Hutchins on 2001-05-29 20:16 ]

Mike Holmes

Note that I feel that point balancing is also the bane of simulationism. It throws things to gamist concerns and rarely helps to simulate anything well.

I've been thinking about it and what simulationist want to balance is something that we could term "Experience Share" or something. Like Story Share, but meaning real time that is spent by the player directing the actions of their character.

Interestingly, this leads me to all sorts of ideas as to how to improve the Simulationist game. For example, having default sorts of activities that the players can have their characters engage in when they aren't being engaged as players by the GM. A player could have a trading empire that they could have rules for administering when not playing actively with the GM (I'm sure I'll come up with better stuff as I think about it). Also, mechanics for introducing oneself into a scene are not only an interesting narrative mechanic, but might make a good simulationist one as well; make it so that players don't have to suffer non-participation as often.

That having been said, I like having sets of abilities that have different spans of control. For example, base attributes that are broad and skills that are narrow. I think that this is an intuitive and potentially effective method of describing characters. As long as you don't have a gamist motive to make characters as powerful as possible, I see no problem with non-power balanced characters at all. The fact is that a "choose your ability" system is more prone to imbalance as people can choose very broad abilities or very powerful ones. Superpowered flight (even at "poor" level) is a huge advantage if your the only superpowered character in the game. So while you're not concerned about single level abilities due to lack of gamist bias I am not concerned with it for multiple level abilities.

The trade off as I see it is in the math. Sure I may have to do an extra step somewhere along the line. But your characters may often end up without a descriptor for a large segment of what would be considered normal abilities. This is the advantage of a dexterity stat, a lot of actions can default to it.

Do you see my point at all?

Mike Holmes

[ This Message was edited by: Mike Holmes on 2001-05-24 20:03 ]
Member of Indie Netgaming
-Get your indie game fix online.

Ron Edwards

Hi Mike,

Sure, I definitely see your point. The entire discussion above, as I said, is predicated on having Narrativist goals, so it's perfectly reasonable to expect that other goals will lead us in another direction.

Now for Simulationist concerns - I can see that in this case one might well want a more granular, encompassing description of the character. I have some thoughts on this but they'll have to wait.

Back to Narrativist stuff that I'm pretty sure of.

Systems like Sorcerer and Castle Falkenstein don't leave "gaps" in characters due to their attributes/skills only approach. In the latter, for example, an unspecified skill is rated at "average," with very few exceptions that require extensive training. In the former, perception is handled specifically by any attribute (one of which is "Cover," or your job/lifestyle) that is relevant to the situation, so that perceiving an ambush in a dark alley is simply any combat-relevant attribute (e.g. Stamina), but perceiving a hidden book might be Lore.

Another, important point about such systems is that they (usually) only model CONFLICT during play, not competence in and of itself.

I think you probably understand this, but what I'm after in this whole thread is that "attribute" does not have to mean "innate ability" and "skill" does not have to mean "training." I'm talking specifically about keeping all aspects of Effectiveness, a very broad category, in the same numerical, non-transformable realm. Once this is achieved, the term "skill" or "attribute" actually becomes misleading.

That's why in Sorcerer, these things are called Scores - one is perfectly justified in thinking of them as "skill sets," or "training-modified innate talents," or whatever. Same in Hero Wars - everything is an Ability, period. Whether it's innate or trained or both is irrelevant in story-making terms, so it's irrelevant in terms of categories and mathematics as well.



Another objections I've always had to differentiation between attributes and skills is a practical one: in most games the two aren't handled in concert very well. When do I use my 'agility' attribute, and when do I use my 'acrobatics' skill ? This is a common problem in this sort of game.

However, I also feel that a problem is that this is an attempt for people to differentiate between nature and nurture, which we've seen (in other discussions) is a very difficult thing to do in the real world, and thus not neccesarily something that needs to be differentiated in a game. Further, even from a simulationist persepective, the bottom line is one ability at any given thing, regardless of where it came from.

                  - James


On 2001-05-24 14:48, Mike Holmes wrote:
Interestingly, this leads me to all sorts of ideas as to how to improve the Simulationist game. For example, having default sorts of activities that the players can have their characters engage in when they aren't being engaged as players by the GM.

Well Pendragon had its Solo Adventures.  A character could go hunting, or manage his lands, or court the princess while the rest of the group was doing other things.

But perhaps the best example of this I know of is the old En Garde game.  Players would script out their desired activities for a month and then go ahead and execute them completely independently of each other.  One player might head down to a local gentlemen's club for a little gambling or carousing.  Another might volunteer his regiment for front line duty.  These side quests were basically the entire game and completely replaced any centralized GM activity.

Mike Holmes


I do unserstand what you are talking about, and a unified field thoery of statistics appeals in it's elegance. One problem that I have with many of these systems is frequency with which you end up defaulting to that "average" score. While from a resolution in the middle/Narrativist POV this is probably fine, it is indeed a problem in a Simulationist game.

Interestingly, OTE doesn't work too bad (Simulationist-wise) because the scores are so much more broad in their scope. I find it very Simulationist to describe why a particular score that you have is pertinent to the die roll in question, as well. And it is easy in such a system to dial up the complexity when necessary. Still, even in OTE you end up with a lot of default 2D rolls. And, you are right, the granularity is not what it should be for a Simulationist design.


The "when to use which stat" problem is not solved by making just one level of statistic. There will always be questions about whether a stat is useable in a specific case. I believe that having large defaulty type stats and more specific ones actually makes the decision easier. If I have both acrobatics and agility in a single level system, then I do have to make a decision about which applies when. With the Ability/Skill type system, I look at whether the action can reasonably fit under the Acrobatics skill. Then if not it defaults to the Agility Ability (say that ten times fast) automatically. As I said, there will be ambiguous cases in any system.

Interestingly, that just brought me to an idea, you could have Skill Use Points that you could use to invoke skills in unusual or ambiguous places. Good for "Exporation Balance".


Interesting that you should mention En Garde. I was looking at playing an on-line version of the game because Max, AKA Balbinus suggested it. That is sort of the idea, but a bit more ground level is what I was thinking. But anyhow, something to do while not directly engaged by the GM, to keep the players interested and at least partially engaged. I was also thinking about journals and other such methods that we've seen, while trying to come up with something else completely original.

Thanks all,
Mike Holmes
Member of Indie Netgaming
-Get your indie game fix online.

Paul Czege

Hey Mike,

Also, mechanics for introducing oneself into a scene are not only an interesting narrative mechanic, but might make a good simulationist one as well; make it so that players don't have to suffer non-participation as often.

In what context could a mechanic for the player to use to introduce his character into a scene be considered Simulationist?


[ This Message was edited by: Paul Czege on 2001-05-25 15:07 ]
My Life with Master knows codependence.
And if you're doing anything with your Acts of Evil ashcan license, of course I'm curious and would love to hear about your plans

Mike Holmes

I was thinking about starting a whole thread about his topic. I'm considering messing around with the idea that Simulationists can't use authorial power.

Many won't want it, certain of my players have rejected it when I tried to give it to them before (ignored or forgot is maybe a better term). But in the game Feng Shui, there is an interesting rule. When a scene starts, the players are required to explain why their characters are there. This is a tiny example of authorial power. But we've argued that Feng Shui is Simulationist. And it is, IMO, but many people will pointto this sort of rule and say that it is a stray narrativist mechanic. But is it really...(that sly Laws, at it again)

Perhaps the Audience, Actor, Author, Director spectrum is too focused on the narrative style. Perhaps in Simulationist games we can give power to the players not to determine story elements, but world elements. There is a subtle difference, here. In the case of Feng Shui, a Simulationist will look at his character and decide on a reason for being in the opening of the scene. He might decide on something that is good for the story, even accidentally.

But a player like myself is more likely to choose a reason that fits the character and story be damned. After all this sort of power exists in very small quantities in actor stance all the time. Why did your character attack that man? He doesn't like that mans kind of appearance. That was decided on the spot, and is a simulation of that characters psyche. So what kind of power can be extended to Simulationist players exercising an approximately "authorial" level of power.

Well, they might be allowed to insert themselves into a scene, for example. This is simply a simulation of the fact that the character walked over, or decided to fly into town. Whatever. But instead of having to play out that portion of things which the player might find boring, the player uses the power as a shortcut to just quickly explain why he has arrived. The limit on this power is different than in a narrativist game. Presumably in a narrativist game the characters arrival must enhance the story.

In the Simulationist version, however, the limitation is that the characters arrival must be internally consistent. He must have had the time and resources to do so, he must be able to explain if asked why his character would reasonably show up: is it random and is that likely? Or did he have a reasonable motive for arriving. It may ruin the story, but if it makes sense for the character to arrive, the player can state that he has. In a way this power would just be an abreviation of the process whereby the player gets the GMs attention, explains the characters intent and they discuss the means of travel, or whatever. Instead the player just says my characcter takes a cab and go to Art's because he's lonely. And the GM says OK, Art you hear a knock on the front door.

And this can be extended further. You can give Simulationist players Directorial powers. They caould decide what sorts of adventures and exploits the characters would have presented to them, in perhaps general or perhaps more specific terms. They could create locales. As long as they didn't void some other part of the game world, and were therefore interally consistent, I have plenty of empty room on my maps that could be filled by the efforts of the players (got this idea from the Alyria forum, BTW, thanks to those there). and it goes on from there.

Who would like this game? Not the "Immersionists". They'd definitely be against it. But the world explorationist Simulationists, might like it. I know I'd like it. Sorry if I'm muddying issues, but this has just gelled in my brain suddenly and solidly.

Responses? Am I mad?

Mike Holmes
Member of Indie Netgaming
-Get your indie game fix online.