There are more and more games out there that let the players freely write up skills for their characters. The virtue of that is that you don't need to count on writing down every skill you have, and you don't have to look through endless lists of skills. There might be other advantages as well.

However, it poses four problems that has to be figured out (or more?):

* The extent of the skill: Are you allowed to take "Knows all languages in the world" as a skill? What about "scientist"? Some games use what I feel is a cop-out: "The GM decides what is appropriate"

* Grading the skill: Is "expert swimmer +1" better than "knows how to swim +1"? Does "computer hacker +1" only gives me a +1 improvement advantage over someone who doesn't have the skill, or does it mean BOTH that I'm a hacker and can try to hack computers (something not allowed by characters) AND it runs at stat+1? Or is it just the latter?

* Efficiency: Does the game give me x bonus points I have to maximize to have fun playing the game? Do I have to be careful of how I word my skills? Are the skills tied to charater efficency so I still have to go for a few "optimized" skills?

* Skill development: How do you create a development scheme applicable to any possible skill?

I know of a few games that use the "any skill scheme": Risus, HeroWars, Ghostbusters, Donjon... BUT I haven't actually played any of these. I'd like to both hear about how the games tackled these four issues, and what the drawbacks and advantages were.

I'm sure there tons of indie games that also deal with this issue. What are the "ways out" already explored?

I've played Risus. And the awe I have of that system is overwhelming. It suceeded in it's goal as well as I can imagine, and it remains the only system, to this day, that I have no criticicismns for. The concept that player skill selection brings is in that of a system where the conflict is undefined. YGG wouldn't seem to be that way. Risus is great because the examples from the source material are funny. This allows for the creation of the skill to be fun and not presumably practical in any way. Interesting things happen in this environment, so be wary. Example: My friend started out with the "Clique": Arrogant Millionaire. One of his tools was $1000000. This means that his character starts out with $1x10^6 and that the tool is needed to use his "Clique" practically. He spent a dollar at a vending machine, and was unable to use his "Clique". It was actually really funny. Loosing half of his dice because he only had $999,999 was histerical for me.

The entire essence of player derived skills is that of player freedom, which needs to be downplayed in competative RPGs. It works in Risus because some guy in a starfighter "Clique:4d" can loose to a barber with a comb: "Clique:3". It is something that every good game designer should be familiar with though.

Pale Fire wrote: Grading the skill: Is "expert swimmer +1" better than "knows how to swim +1"? Does "computer hacker +1" only gives me a +1 improvement advantage over someone who doesn't have the skill, or does it mean BOTH that I'm a hacker and can try to hack computers (something not allowed by characters) AND it runs at stat+1? Or is it just the latter?

Pale Fire wrote: The extent of the skill: Are you allowed to take "Knows all languages in the world" as a skill? What about "scientist"? Some games use what I feel is a cop-out: "The GM decides what is appropriate".

Pale Fire wrote: Efficiency: Does the game give me x bonus points I have to maximize to have fun playing the game? Do I have to be careful of how I word my skills? Are the skills tied to charater efficency so I still have to go for a few "optimized" skills?

Pale Fire wrote: Skill development: How do you create a development scheme applicable to any possible skill?

In games with select-any-skill mechanics, there is a natural tradeoff between skill specialization and generalization, in which specialized skills are more effective than generalized ones when the task falls within the area of specialty, and are less effective than generalized ones when the task falls outside the area of specialty. The broader the skill, the more situations it's effective but the less its maximum effectiveness is. (Unless it's a case in which breadth of knowledge is important, see below.) In play, I believe this is often applied automatically, perhaps even unconsciously.

There might be a way to formalize this. Such as:

Skill applies primarily to the specific task being attempted: 3 dice (or whatever the effectiveness mechanism is)

Skill applies to the specific task being attempted, but equally to other tasks of a related nature requiring different techniques: 2 dice

Skill is too specific to apply to the exact task being attempted, but the task it does apply to requires very similar techniques: 2 dice

Skill applies to the general type of task being attempted, but in no specific way: 1 die

Skill applies to only a small part of the task being attempted: 1 die

So for example:

Skill "rock climbing"; task "climbing a cliff"; 3 dice
Skill "climbing"; task "climbing a cliff"; 2 dice
Skill "acrobatics"; task "climbing a cliff"; 1 die
Skill "rock climbing"; task "climbing into a third story window"; 2 dice
Skill "climbing"; task "climbing into a third story window"; 2 dice
Skill "climbing"; task "leading a mountaineering expedition"; 2 dice
Skill "rock climbing"; task "leading a mountaineering expedition"; 1 die
Skill "science"; task "performing a Western Blot on a DNA sample"; 1 die
Skill "forensic science"; task "performing a Western Blot..."; 2 dice
Skill "genetic research"; task "performing a Western Blot..."; 3 dice
Skill "genetic research"; task "identifying the purpose of a strange laboratory"; 1 die (assuming the purpose is not related to genetics)
Skill "science"; task "identifying the purpose of a strange laboratory"; 3 dice.

Note the final example in which more generalized skill is more effective because it's one where broad knowledge of a range of possibilities is required. Similar examples: "knows all languages" would get 3 dice to identify an unfamiliar script but only 1 die to read any given instance, while "knows [this specific language]" would get 3 dice to read that language but 1 die at most to identify scripts other than that. (Two dice, if the script is linguistically close to the language known, no dice if it's totally alien.) "Expert in all martial arts" would get 3 dice to devise a counter to an unfamiliar style, but only 1 die to use the counter; while "Expert in [the specific weapon or style wielded]" would be exactly the reverse.

- Walt

What I get reading all this is that there are a couple of considerations.

First, Andrew, your suggestions are great for games with less mechanics than Christoffer's. He already has a system in place (for the most part, and as I understand it), that has ratings for skills, and does use points, etc. So he'll have trouble adapting. Because If you have decided to have ratings and balance of any sort, you do have to consider these things.

As always my solution is to look first at the mechanics and then label to suit them. That is, there is no taking Expert Juggling at level 1. This is simple, just do not allow any term that has a qualitative meaning as part of a skill title. So you can have Juggling, just not expert juggling. Yes, there will be borderline cases. Some might see Equestrian and more talented by default than Horseback riding. If players want to take certain titles just say that they have to stick to using that which has gone before (so if Equestrian is used first then everybody uses it for horseback riding), or just note to all that they mean the same thing. That Equestrian 2 = Horeseback Riding 2.

Because "balance" reqires this.

As to the subject of breadth, one could make a mechanic that worked with that, as well. For example I could take Spanish 3(1), and Languages 2(2), the first number indicating the effectiveness of the skill, and the second indicating the amount of time it would be useful. So, (using a pseudo system), when encountering Spanish, I get my skill use. When encountering another language, I roll my 2 dice against the rarity of the language in question to see if my skill even pertains before rolling the effectiveness.

But then this is probably superfluous, and using a system like what Walt describes is probably easier.

Yes, this requires a GM call. And, yes, all such systems will require judgement at some point. But this is no different than any other skill use. As Walt points out, can I use my Acrobatics to climb a mountain? I have to make some assessment as the GM. So tying to avoid that is impossible, and pointless.

The fact is that in the example games you give, and all such games the same "way" is used to ensure that the skills selected are proper. The GM is given examples, and then must decide whether a particular skill is suitable. All such systems will comment on whether or not (usually not) a skill such as "Knows all Languages" is appropriate.

And, as a person who has played such games, lo and behold, it works spectacularly well. Worrying about "abuse" will just get you the standard rant about how no system can avoid "abuse", and one should only worry about whether the game will be fun for players who employ it as it's designed. Does the game as you've written it challenge players to break it? Then you'll have problems. If not, you'll be fine.

Mike

Actually Mike, this question was more in a general vein than particular for my game. That's why I posted it here in the first place.
To keep things in the right forum and such, let's keep this thread general. I'll copy some stuff and start a new thread in Indie Game Design.

To get back to the more "general topic". Pyron presented a little bit of Risus approach and Andrew Accord's. Can we get more "reviews" of this type of systems?

If I understand HeroWars right you pick a culture and an occupation. These give you descriptors which default to certain rankings depending on the power level of the campaign. In addition you also get 10 free descriptors. All descriptors have defaulte values. After that, the default values can be raised for I think three of the key words or so. There are no stats, you simply use the rating of the appropriate descriptor.

Donjon (please correct me here folks if I'm wrong, only read a teaser of the rules) lets you make up 5 (or 3?) descriptors. These provide bonus dice - same as its rating (how do you select the rating?) - to the corresponding stat test if you can argue that you're using the descriptor to enhance that test.

What about the other games?

Pale Fire wrote:
If I understand HeroWars right you pick a culture and an occupation. These give you descriptors which default to certain rankings depending on the power level of the campaign. In addition you also get 10 free descriptors. All descriptors have defaulte values. After that, the default values can be raised for I think three of the key words or so. There are no stats, you simply use the rating of the appropriate descriptor.

Donjon (please correct me here folks if I'm wrong, only read a teaser of the rules) lets you make up 5 (or 3?) descriptors. These provide bonus dice - same as its rating (how do you select the rating?) - to the corresponding stat test if you can argue that you're using the descriptor to enhance that test.

What about the other games?

I'd agree with Mike here. If you try, a gamist can 'game' pretty much anything. Also, I don't think there would be that much a problem in practice.
'GM decides what is apropriate' is shorthand 'GM will check for compliance with your groups social contract'. You can have one group with evreyone has
general descritpors that covery many things, and another where all descriptors are quite specific.
Shadowrun has both systems. There is a fixed skill system, for skills with discrete mechanical effects and another skill system called 'knowlege skills', which players make up, sorta like descriptors.
For other sytems:Hero wars has a way that multiple descriptors can be used in an action, all past the first act as bonuses(probabilisticly). This compensates for over specificity. Also, I believe the opposing roll is modified a bit based an propriateness.
Donjon oth has a rule that no more than one descriptor can affect any one roll. Also, there only two types of descritors, Primary and Seconday. Since the differntion is fairly coarse, it's pretty easy to use.
Finally, it has been my experiance that a group tends to converge on some sort of standard for what sort of things are apropriate, it becomes
part of the social contract for that game. If different groups converge to different standards, it's no big deal. Do you know of any examples of this being a problem in practice? (Not a challenge, but I would like to know)

damion wrote: Finally, it has been my experiance that a group tends to converge on some sort of standard for what sort of things are apropriate, it becomes part of the social contract for that game. If different groups converge to different standards, it's no big deal.