Attribute/Skill system

[Egonblaidd]

Hi, I'm new to the community and to RPGs in general.  I feel like I'm introducing myself twice since I posted another thread, but I feel like that topic might be rendered moot by this one.  Basically, I've been working on a rule set, and I'm having second thoughts about my design.

Anyway, I've been rethinking the way I use attributes, skills, and dice rolls to resolve tasks and conflicts, and basically I just need to let my thoughts spill out so I can look over them in a more coherent fashion.  External input would also be helpful.  I also realize that it depends a lot in what my goals are with this, I admit I haven't quite worked that out yet.  I suppose I want things to be streamlined enough that people have fun, yet expansive enough that a player can do anything, like open a shop and become a blacksmith, or delve into politics.

Basically I'm wondering if I should change the fundamental principles on which skill tests etc. work.
Right now, attributes and skill levels are equivalent when it comes to a character's ability to perform a task.  An attribute of 7 and a skill of 5 is the same as an attribute of 5 and a skill of 7.  There are some subtle yet profound differences between attributes and skills, however.  Attributes never change after character creation.  Once your character is created, that's it, no way to ever get more attribute points.  Skills can increase indefinitely, your chances are never going to reach 100%, they just get closer and closer.  There are many skills which are augmented by a single attribute.  On occasion, a character will take an attribute test, so skills won't even be involved, while all skill tests use that skill's attribute.  Skill tests are resolved by a d100 roll, where the chance of success = 100 * (attribute * skill) / (attribute * skill + [some constant]).  Since this number is a little difficult to calculate, I instruct that it should be recorded on the character sheet next to each skill.  Attribute tests just have you roll a d10 against that attribute, so in the unlikely case that an attribute is 10+, success would be automatic (without modifiers).

While this works, and has some nice features, it also seems a little overly complicated and doesn't differentiate enough between attributes and skills except that attributes are better, but limited.  I'm thinking of something that would make a much larger distinction between the two, and make things a little easier computation wise.  One advantage of the above system, though, is that you know going into any test exactly what your odds are.

So here's the idea I was playing around with in my mind.  You get a number of d10s equal to your attribute.  Every task has some difficulty modified by skill (if applicable), and at least one of the die has to be higher than that difficulty (or harder tasks could require more than one die).  My problem with this is that it seems to be too granular; there's not a whole lot of variance if difficulty must range between 1 and 10.  The same argument could be made of attributes in my current system, but I doubt you'd want to toss more than 10 dice anyway so the attributes would be just as granular in either system.  Also, due to the granularity of the the difficulty, skills would either have a large impact (such as -1 difficulty per level) or they would have almost no effect.  For example, if difficulty = 10 * skill / (skill + [some constant]), then an increase in skill might yield no result until it reaches a certain level.  However, attributes could be anything and still never guaranty success, even on attribute tests.  I suppose one option would be to switch it; players get a number of dice equal to their skill and roll against their attribute (to try and get one of the die equal to or less than that attribute), but I shudder to think about a player scooping up 20+ dice to roll.

Also, I like the idea of being able to increase skills indefinitely.  Because, honestly, the guy who has been making weapons his whole life is going to be a lot better at it than you who just picked it up as a hobby and potential source of gear or cash, even if you can do it right 9 out of 10 times.

To see where I'm coming from, the only RPG that I'm actually familiar with the rules is Warhammer Fantasy Roleplay, from which I decided to adopt the d100 mechanic.  It seemed quite logical, but I've gone and complicated the whole thing.  I have done some research online, though, so I know about a lot of the different option that are available.  In fact, I discovered that none of my ideas were actually new, and I know that hasn't changed yet.

Anyway, thoughts?

[chance.thirteen]

These are only indirectly related. I too had issues with skills vs attributes in my fairly rules heavy designs. I look back at how I handled said issue in previous games.

In Champions, you buy a basic skill ability for a set cost, and then raise that ability by another set cost. The only variable is that the initial investment gets you a skill roll based on your Attribute. I ended up saying that making the roll indicated success, but that you could only succeed by an amount based on h9ow many points were spent on the skill.

In WoD or 7Th Sea type terms, you could say you roll Attribute+Skill level dice, but only keep up to your skill level as results.

Now, the reason for this is simple: It was a super hero RPG. The characters routinely had super human agility, intellect and presence, and I felt a need to mark the difference between being super agile, and being trained with how to do various athletic skills.

And that is the most important issue: what is so important about your design in your eyes? Why have Attributes and Skills add together for a pool? Why is Skill different from innate ability? What do you want to have this produce in your play?

[Egonblaidd]

I think I'll be sticking with my original system, or maybe a slightly modified version of it.  Using the Big Three and Power 19 I've managed to figure out what my design theme is, which is about choices and morality.  In a sense, attributes represent a situation that you had no control in shaping, and skills are your attempt to deal with that situation.  The attributes will make certain skills more viable than others, but any skill raised high enough will be able to compensate for a weakness in attributes, but it will require much more work to get there.  While a dice pool might be "simpler" from a gameplay perspective, a d100 system is much easier to calculate the probabilities, since that is exactly what you do when you figure out what number you need to roll to win.  The granularity is also much finer, allowing the difference between two character's aptitude to be much more noticeable.  In the end, a low attribute can be compensated for with a high skill, but there are a lot of skills, so you won't have the time to get all skills of that attribute high, and it's more efficient to focus on your strengths.

[Egonblaidd]

I've come up with a new question that relate to my Attributes and Skill system: advancement.  Here's how I originally wanted to do it.
You perform a skill test in order to perform some task.  This is a d100 roll, so let's say you have to roll X or less to pass.  You roll Y.  Now you take an Intelligence test, modified by your degree of success/failure, such that if, for a normal Intelligence test you must roll under Z, you now must roll under Z + Y - X.  If you pass, you are awarded a skill point (ten skill points advance a skill a level).
Example: To successfully pick this lock, you must roll a 46 or under.  You roll a 24.  Now you take an Intelligence test.  Normally, you'd have to roll a 60 or less to pass, now you have to roll a 60 + 24 - 46 = 48 or under to pass.
The idea is that your chance of learning new skills is influenced by the difficulty of the specific task, so by failing you increase your chances of increasing your skill.  This seems like a good idea to me, and I don't have a problem with this system.  My problem is that I rewrote my Attributes, and in doing so I split Intelligence into Intelligence and Cunning.  Intelligence is the ability to perform calculations and comprehend abstract concepts, while Cunning is more of a practical or street smarts.  My question is, which of these should be used to advance skills?  Or should both be used?  Should mental skills use Intelligence to advance will physical skills use Cunning to advance?  Or should each skill just use its associated Attribute to advance?  For example, Combat is an Agility based skill.  Should it use Cunning, Intelligence, or Agility to advance?
I realize it all depends on what I want to do, but this is one of those things where I don't think it matters so much, I just want to do something that will make sense to the players.  Right now I'm leaning towards Intelligence advancing mental skills while Cunning advances physical skills.

[Daniel B]

:-S The skill advancement is a bit math heavy for an RPG, isn't it?

From my perspective, the questions you're asking seem really, really abstract.

You say you realize that it depends on what you want to do with the system, but, not to be rude, I'm not convinced about that. What does "Cunning" or "Intelligence" mean in real-life terms? There's no physical sections in the brain set aside for either or both. In fact, they're just convenient labels used by other games for what they wanted to achieve in those games.

It feels to me like you're asking questions about structures that are inherently abstract. In order to help with those questions, I need a clearer idea of what you're trying to achieve by building these structures in the first place, for your own game. Otherwise it's just trying to make a science out of your own artistic piece, and only you can tell what's best for the art you create.

Dan

[chance.thirteen]

Some tidbits, both based on playing Runequest.

1) How often were you going to allow this sort of advancement. RQ had problems where players sought out chances to "get a check" on a skill, because you checked at the end of every adventure to see if the skills used improved. The reverse of this is if you intend every roll to have a chance for improvement associated with it, are you planning on keeping the number of rolls reined in? Or will my thief picking four locks in a row get four chances to raise his lockpicking?

2) RQ has a stat bonus based on several stats in their game. For instance, you might gain a bonus from int and charisma for a social skill.

2) RQ (the RQ I played a long while ago) used a percentile rating system. If you made a roll in play, you got a check. At the end of the adventure you rolled for each checked skill. To gain advancement you had to roll OVER your current skill, so the higher your skill the less likely the advancement. The group I played with added your stat bonus to your roll, so more taletned chaarcters advanced faster. Likewise the game always allowed advancement on a 96-100 roll. They gained 1d6 to the skill rating in all cases.

[Egonblaidd]

Actually, I think I was reading about the advancement system used in Runequest when I came  up with my idea of advancement, though it is kind of flipped around (it said you had to pass a test in RQ in order to advance, which meant that a higher skill would provide more opportunities to advance even if the chance of advancing for one opportunity was less with a higher skill).  I realize that I could have the problem of players seeking out chances to improve their skills (which kind of makes sense as real people would try and improve their skills), so I had thought that it might be necessary to impose some sort of limitation on advancement, i.e. only one skillpoint can be gained over a half hour in-game time, so you can't sit there trying to pick the same lock over and over again, or a failure must be within 4 magnitudes of failure (failed by less than 40), which would make it difficult for, say, the entire party to take a turn leveling up lockpicking before the thief actually picks the lock.  I haven't exactly worked this out, though.  The ideal, I suppose, would be to give a fixed number of advancements per session or something, but that doesn't exactly work with my system since people may use a lot of different skills, and the skills that are used should get a chance to increase since they may be needed more later.  For example, I don't think it would be appropriate for a group consisting of a student, farmer, priest, and mercenary to ALL dump their points into combat skills at the end of a session, but if they did do a lot of fighting then it would make sense (I'm not using a class system, but right now I'm using "backgrounds" as a means to distribute starting skills as with a template format).

I'm looking at my skill system and, besides the fact that I don't think I can cover everything, I think that it might be a little too complex.  I'm currently sitting at 146 skills (or subskills, I have skills grouped into 34 groups), and I'm thinking (a) what is that going to look like on a character sheet, and (b) what is that going to look like on my background (template) descriptions?  So I've been contemplating converting to a trait based skill system, like "Swordsman +2" or "Weaponsmith +1".  This would necessitate a complete overhaul of my resolution system, and I'm still thinking about what kind of system might work.  In any case, a trait system would lose the complexity and increase the diversity (maybe I'm a swordsmith, and not just a weaponsmith), but there'd be no way to regulate a trait based system since each trait would vary in power and ability ("Swordsman +1" is not as powerful as "Weaponmaster +1"), and you know that players would overlook aspects of their character until it actually came time to do a test that used a trait they had forgotten to give themselves.  I should probably do a little research on other trait based systems.

There must be some way to relate my skill system to my design theme, some way that can narrow down my options.  My design theme is about morality and making moral choices in difficult situations.  I don't immediately see how that would relate to the skill system, except that you use your skills to carry out your moral decisions, so if you have little chance of successfully completing one task due to a low skill then it would be better to consider a more likely option.  Someone suggested in another thread that I could just use the moralities as my stats, which is an idea I had considered, but I don't think it offers quite the breadth I was hoping for in a skill/attribute system.  Anyway, I'll think about things some more.  If anyone has any advice, I'd welcome it.  The fact is my current system is quite complex and math heavy.

[chance.thirteen]

I believe old RQ limited you to one advancement check per adventure, so higher skill had a great chance to get the check, but low chance of succeededing.

Also: think about what it is you want to see from your skill checks. Are they yes/no resolutions?

I myself have been thinking about it and my own preference seems to be one of these:
Skill is sufficient, no roll
Skill is sufficient, roll for extra success if it is available
Skill is sufficient, but another challenge comes up (test a different skill, or at least alter the stakes and retest the skill)
Skill is insufficient, choose another route.

If you are going to overhaul your resolution method, you may as well look and see how you intend to use it.

[Luke]

Hey Phillip,

Actually, the number of skills is largely unimportant. You can have a thousand skills in the game, but if the players are only using five each at a time, the overall number isn't a burden. I lost track of the number of skills my game has long ago. I add more with every book. It doesn't matter!

As for advancement, you can absolutely tie skill advancement into your morality system. You could write a rule that indicates players only get opportunities for advancement if they do the moral thing. THAT will encourage some moral behavior. Of course, you'd also need to weight the flip side. There'd have to be a reward for doing an immoral thing.

That's a key feature of the type of system you're looking for. You want to reward success and failure, each in their own way. Advancement is just one type of reward of many possibilities.

If you're going to use test-based advancement, which I think is swell, you need to place a limiter on how players can test.
The simplest limit for test-based advancement is one advancement per skill per situation. So it doesn't matter if you test to pick the lock three thousand times, you get one test for purposes of advancement.

-L

[Egonblaidd]

That works well for lockpicks (and something that I had considered myself), but what about combat?  It stands to reason that a person could learn quite a few things in a single skirmish, assuming they survive (although an advancement test every combat round might be a little much).  Although, I could slightly change the rate of advancement.  Right now I have a successful advancement test awarding a skill point, of which it takes 10 of to upgrade a skill.  I could change a success to simply upgrade a skill one level, or something, and only allow one skill point to be earned for any skill from a single encounter.  Anyway, that's something I'll have to sort out.

As for advancement being tied to the morality system, I'm less keen on that.  I expect the characters to go against their morality on a semi-regular basis, at least until they get a good feel on what they can and can't do and what they will and won't do.  As such, having immoral actions change a character's morality and impose a guilt penalty for some time seems like a more appropriate way of penalizing a character.  Advancement should happen regardless, if anything you'd almost think they would learn more by doing something that they typically avoided.

As for the number of skills, I think that depend on how skills are handled.  If skills are on a have/don't have basis a la WFRP, then you can have as many as you want and a player need only worry about the few he or she has.  If every character has every skills, but a different grade or rank or level of each skill, which is what I'm doing (sometimes, though, I think the other way would be so much simpler), then it makes for very cluttered character sheets and much wasted space in a rulebook detailing what skill levels what class starts with (though I'm using templates rather than classes).  In the end, though, players will still only worry about the skills they use, regardless of what the rank of those or other skills are.

Anyway, the best thing to do, probably, is to playtest different versions of my system to find one that works best.  I'm not quite to the point of playtesting, yet.  I've been trying to codify my rules in a rulebook at the same time I was trying to come up with them, so I started a new doc that is just number and statistics and the shortest, barest descriptions, and it seems to be better and faster.  I can write a rulebook once I get all my rules down, and probably playtested a bit.

[JoyWriter]

I've been contemplating converting to a trait based skill system, like "Swordsman +2" or "Weaponsmith +1".  This would necessitate a complete overhaul of my resolution system, and I'm still thinking about what kind of system might work.  In any case, a trait system would lose the complexity and increase the diversity (maybe I'm a swordsmith, and not just a weaponsmith), but there'd be no way to regulate a trait based system since each trait would vary in power and ability ("Swordsman +1" is not as powerful as "Weaponmaster +1"), and you know that players would overlook aspects of their character until it actually came time to do a test that used a trait they had forgotten to give themselves.

When I used to GM Shadowrun, I ran on the principle that defining skills yourself should include a tradeoff, diversity of application for quality of result. This worked smoother because the skills used were knowledge skills, and so a low level of insider knowledge was perhaps equivalent to a high level of outsider knowledge, with the outsider knowledge allowing inference of the stuff that was obvious for the insider.

The problem I realised with this is that it is only a trade-off if the player doesn't know what setting they will be in. And if they are not aware of the gamble, they could be left with skills they will never use. One solution I have considered to this is to look at the traits the players have put down and design the tone of the adventure accordingly. Now this can be considered "wreck the min-maxer" because an explicit part of the reaction would be to take broad skills like "master fighter" to be an indication that fighting was not meant to be the majority of encounters, and fitting in more of their more focused "minor skills". In the same vein I have considered having each player give 5 skills that their character considers non-intersecting, and use that as clues to design the potentials for events within the story.

Also I'd take Ron's advice on traits in "the pool"; the number sets the skill, always, the skill name just gives where it can be applied, and says something about the character. So swordsman might also tell you about military tactics, and weapon-master might tell you about forging.

A third idea is to have pre-made overlapping professions, more like your basic skills. You could set them so that "in the normal course of things" they have equal applicability, and then players can just pick between them, and sometimes apply more than one to the same roll.

[Luke]

Hi Phillip,

I think you missed the part where I said that both moral and immoral actions need to be rewarded and both success and failure need to be rewarded.

I strongly encourage you to play other roleplaying games, too. There are many games out that have the same premise as yours, that have tackled and solved what you're working on -- mine included.

Lastly, you're not playtesting? You should be playtesting. You should be playing your game once a week and making up the rules as you go.

Good luck!
-L

[Egonblaidd]

For the moment I think I'll stick with the skill system I've been working on, and when playtesting starts I can hammer out a trait-based alternative and see how it compares in playtesting.  The main reason I'm not playtesting yet is that I haven't finished all the mechanics.  I have no rules regarding the use of magic, or economic dealings, and only a somewhat sketchy idea how combat is going to work (I'm ditching hitpoints in favor of a more realistic and deadly wound system that penalizes resolution tests depending on the severity of the wound).  I guess what I'm saying is, I have an arm, a fairly meaty arm for something still in the preliminary design phase, but it's still just an arm, and I need the entire skeleton together before I can start playtesting.  Which means I should stop focusing on skills and attributes, I guess.
The other reason I'm not playtesting yet is that my gaming group has yet to actually show up to a gaming session.  I've been trying to get a group started for at least a month or two, and so far they haven't ever all shown up to a session.  All we've done is character creation (and their choice of characters doesn't make it easy for me to come up with plot, though it does provide some interesting possibilities).  But that's college life for you; anytime you have free time, someone else has some obscure class, or homework that needs doing, or someone is hanging out with their girlfriend.

As for your comment, Luke, about rewards for both success and failure in my morality system, I think it will get worked out alright.  I mean, if doing the right thing earned you, say, 100 gold that you could use to buy that sword you always wanted, and doing the wrong thing earned you... that sword you always wanted, then where is the motivation to do evil?  Rather, instead of a stock reward for acting immorally, I think the appeal for an immoral act should be inherent in the situation.  You want to stop this guy's evil plan, what you should do is expose his plot to the authorities, but you are given a chance to kill him and get away with it.  Do you take the easy way?  Or follow the harder path and risk failure and worse, sticking to your character's beliefs?  Also, I've been toying in my mind how Dark magic will fit into my game, and I'm thinking it will be closely linked to the moralities such that those who use Dark magic will be pushed to the extremes of each morality gauge, and in return will receive a large benefit to spellcasting.  Religion is also a factor in the morality system, though it of course pushes for right action rather than wrong action.  Holy magic requires the caster to be aligned to the morality of their church, which can be interesting if they don't know what that morality is (in metagame numbers, that is).

[Luke]

1) Stop worrying and start playing. Play with exactly what you have. If you all you have is skill mechanics, that play out skill challenges. Find one or two people who are interested in kit-bashing with you, sit down and hash it out. Write the mechanics as you go. Don't try to plan out everything ahead of time --- because you're wrong!

2) Oh Phillip, I remember a time nine short years ago when I fiesty young Luke was mumbling the exact same excuses... Which means, you're almost there man. You're almost through the cruft and the myopia. Soon, if you keep pushing, you just might have a game that expresses your vision. I'm going to give you some homework, though, since you clearly don't understand what I mean. "Rewards" is a game design term. It's not a dog biscuit. Advancement is just one type of reward. And advancement can be tied to morality. Go forth and play Dogs in the Vineyard by Vincent Baker. Turn your eye toward Fallout. How does that work? What's it rewarding? Go forth and play My Life with Master, by Paul Czege. How is the player rewarded for moral behavior? How for immoral behavior? And, if thou art a brave soul, go forth and play Sorcerer, by Ron Edwards. Grasp hold of Humanity and ring the life from it.

You've played on game and affected you deeply and inspired you to create a game. Playing more game is going to inspire you further and give you a more nuanced perspective and make your own game deeper.

3) Just because I can't resist: You're absolutely correct. Religion (and faith) ALWAYS pushes one toward the right action rather than the wrong action.

-L

[Egonblaidd]

Heheh, I guess I kind of blundered there.  Religion doesn't necessarily push people to do right actions, sometimes they misinterpret their religion and that pushes them to do wrong actions.  Also, it depends on the religion in question (though the only religions that will use Holy magic will be "good" religions, other religions will use a different form of magic).  But that's why I have religiosity as one of my morality gauges: if someone truly wants to do as their God or gods want, then they'll be more moderate and closer to the center of the scale.  If they are more concerned with enforcing their religion then they'll tend more toward Fanaticism, which is generally bad (it certainly isn't good, at least).  This can motivate wrong action, especially in an environment with more than one religion (which should make for some interesting situations).

Hmm, rewards...  Ok, how about this?  Doing an immoral act shifts your morality toward the type of action you performed and imposes a guilt penalty.  In addition, the player gains a Corruption point that is like the Morality points gained from doing right actions, except that the Corruption point can be spent on any action (not just those associated with that particular morality) that defies that character's moral beliefs for an automatic success, or a Corruption point can be spent to negate the guilt penalty.  Or better yet, the player can choose to either get the guilt penalty and no Corruption point, or forgo guilt and gain a Corruption point.  So players wanting to be "good" would endure the guilt when they perform a wrong action, but players that were becoming "evil" could pass off the guilt and instead get a bonus point.  But then there has to be a bonus to taking the guilt.  Hmm.  Maybe once the guilt wears off the player can choose to return his morality to its original state or keep it in its modified state (e.g. repent of his crime or accept that his original perception of right and wrong was misguided).  But is that enough?  Maybe.  By it's very nature, gaining and using Corruption points would prompt more shifts in morality, which would perpetuate the cycle until the character is as evil as they can get.  Maybe I need some kind of Conscience gauge as well.  Hmm, I'm going to have to think about this.