Attribute/Skill Hierarchy for a Universal System

[Egonblaidd]

Vordark's threads about Genesys have gotten me thinking to how I would design a universal system.  I think it would be a good exercise even if I can't/don't make one capable of competing with the pros, and besides, I just like designing things.  Anyway, I'm starting off with a mechanics idea that basically organizes all attributes and skills into a hierarchy.  I haven't really thought much about skills, but I have an outline of an attribute hierarchy.  Basically, each attribute/skill has a number of attributes/skills below it on the hierarchy, and the rank of an attribute/skill is the average of ranks of the attributes/skills below it (example: if Body = 7, maybe Athletics = 6, Dexterity = 8, Senses = 7, see below).  For example, on the top hierarchical level for attributes, I have

Body
Mind
Social

Three very distinct and simple attributes, easy to remember, not complicated to use.  Each of them has three sub-attributes on the level below it.  Each of these has three further attributes below it.  So my complete list is

Body Mind Social

┌ │Athletics │ │ │ ┤Dexterity │ │ │ │Senses └ ┌ │Intelligence │ │ │ ┤Cunning │ │ │ │Will └ ┌ │Charisma │ │ │ ┤Appearance │ │ │ │Cool └

┌Strength ┤Endurance └Speed ┌Coordination ┤Reflexes └Agility ┌Vision ┤Hearing └Smell ┌Creativity ┤Logic └Memory ┌Perception ┤Wit └Planning ┌Self Control ┤Determination └Spirit ┌Charm ┤Influence └Empathy ┌Beauty ┤Command └Intimidation ┌Composure ┤Confidence └Self Esteem

As you can see, the full set is quite complicated, but that's exactly it.  This design has inspired two possible, but unfortunately, mutually exclusive applications (you could possibly do both, see a few paragraphs below).  The first is that when some kind of resolution is required, the appropriate level of hierarchy can be applied.  A rough example might be a fitness examination.  You could test Strength, Endurance, Speed, Coordination, Reflexes, and Agility individually, or you could just test Athletics and Dexterity.  Or even just a single Body test (a fitness examination would probably at least involve a Vision test).  One task might require a good overall Dexterity (which could be resolved through several tests of the Dexterity sub-attributes, or one Dexterity test) while another might specifically require a good Coordination.

Another advantage (which should always apply, regardless of which of these two applications you use) is that the presence of sub-attributes helps to describe what an attribute entails, such as Athletics being about heavy lifting, stamina, and running.  There will be, however, some confusion, for example, what is the difference between Dexterity and Agility, or Cunning and Wit?  Here I define Agility as movement speed, for example, the time it takes between starting to throw a punch and when it connects with your target, and Wit as a sort of mental reflexes, how quickly you can respond to mental input, such as in a conversation.  There may even seem to be some conflicts, say, between Cool and Will, but you need only understand that someone can be stubborn and hot tempered, or calm and submissive in order to see the difference.

The other application, and the one I see as being the more practical of the two, is collapsibility.  Want a simple game?  Collapse all attributes into Body, Mind, and Social.  Want a hack 'n' slash?  Expand Body, Athletics, and possibly Dexterity.  Then your attributes are Strength, Endurance, Speed, Dexterity (if you don't expand it), Senses, Mind, and Social, similar to most standard RPGs.  Is magic important?  Expand Mind, and possibly its sub-attributes.  Do you want to play a game about social interaction?  Collapse Body, and possibly Mind, and expand Social all the way through.  Do you like highly complex games (like me)?  Expand everything.  You can have whatever attributes you want, and of course you can add, change, or delete an attribute if you want.  Maybe you don't see Senses as being useful in a hack 'n' slash, so you remove it.  Maybe you want to add Psionics to Intelligence for a sci-fi game.  Or maybe *gasp* you want to add another hierarchical level, because it isn't complex enough for you.

This inspires an interesting design philosophy: make everything in the system collapsible.  You can play using just one Combat skill, or you can expand it so that it encompasses different weapon types, or even fighting styles or individual techniques.  You can play with one hit location and two health states, alive and dead, or you can have a dozen hit locations with twenty ranks of wound severities (gratuitously lethal wound to the right index finger!).  Perhaps one sword is the same as any other, or perhaps a wide variety of swords exist, from arming swords to rapiers, to broad swords, to long swords, to claymores, to katanas, etc., and furthermore each item, including swords, has different levels of quality that modify its properties.  Basically, there would be a base system (first hierarchy level), with optional levels of complexity to each system (subsequent hierarchy levels).  Do you want 10 skills, or 500?  More specifically, if skills are modular like attributes, which of the 500 (hypothetical) skills are relevant to the game you want to play, and which ones can be collapsed down to simpler hierarchy levels?  Is one Security skill enough, or do you need Lockpicking, Hacking, etc.?

This also ties in with another idea I had, that the rules would actually be different for different players.  (This might actually allow for both above applications to be used without being mutually exclusive.)  Are you a warrior?  You use more complex combat rules.  Are you a wizard?  You use more complex magic rules.  A wizard CAN still fight, but with simpler combat rules there is less options and less things he can take advantage of.  Basically, in any RPG the more in depth a particular part of the rules is, the more important that aspect of gameplay is to that RPG.  In the same way, a player can use more complex rules for things they are interested in.  Maybe the fighter CAN try and bluff his way past a guard, but a rogue would have a whole lot more ways to do it and loopholes to take advantage of.  Using simple combat rules, you could only strike an enemy in a generic hit location and cause some sort of general penalty, while a warrior could target specific hit locations and inflict a wider range of injuries, capitalizing on an opponent's weak points, for example, disabling an enemy's weapon arm.  A wizard might have a higher Logic than, say, Creativity (and so, his Logic is higher (maybe) than his Mind), and therefore can exploit that advantage by using Logic based skills, while a warrior might only have a Mind attribute (that includes both Logic and Creativity), and therefore can take no such advantage.  Just a thought about how this might work.  It would, of course, be easier to use the same rules for everyone, but this is definitely an interesting idea, I think.

Something else that can be done with the attributes listed above is to design a damage system for each aspect.  Many RPGs, especially standard ones, have a physical damage system.  Some, like WFRP (Warhammer Fantasy Roleplay), have a system that governs insanities.  By designing separate, but related systems for physical, mental, and emotional damage, the potential for a universal system increases (you can have a game about killing monsters or political debates, damage systems would exist that would allow for either type of play; getting a rival to lose his temper in an important debate could devastate his political career, allowing you move up the political chain).  It makes it even better if you can model all three systems in the same terms, perhaps as a single system where the effects on each group are listed (for example, you could use the same damage ranks and simply list the effects of a wound of that rank for physical, mental, and emotional damage).  The more similar the systems are, the easier it will be for people to understand them (the less there will be to learn).

At this point, I don't think I can provide more than a hierarchy level or two for skills, since skills will depend on the setting (fantasy? sci-fi? historical? modern?).  One option would be to do as Vordark is doing with his system, selecting a generic setting and tailoring the system to that setting, but such that it is easily modifiable to other settings.  Another option would be to provide a skill for everything in every genre (which obviously doesn't make a whole lot of sense, since people can always invent new settings with something you don't have), so you might provide Hacking and Lockpicking skills, but in a fantasy setting it would be collapsed into a single Security skill that would be used pretty much just for picking locks.  Personally, I think it makes more sense to keep setting-specific things, like a skill list, short and limited, and then to provide modules that expand on the base system and provide a specific application.

Anyway, as I said, I just like designing things, and ideas like this intrigue me.  Has something like this already been done?  Does anyone else think this is an interesting idea?  Feel free to use any aspect of this if you want.  I have a bad habit of starting things but not finishing them, so by all means, put my ideas to good use.  Also, I realize my choice of attributes is not necessarily ideal, some of them are more "filler" (especially under the Social attribute) just so I had an equal number for everything.  Also, and I don't know if this is good or bad, but some actions might require tests of more than one unrelated attribute, such as a Vision test to see if you see something, and, if successful, a Perception test to see if you notice what you saw (imagine trying to spot a particular person in a crowd, for example, seeing and noticing aren't the same thing).  Or Wit would probably be used a lot in Social skills.  Anyway, I kind of wish we had a Theory section on the forums, because I'd be there a lot, I think (there is one, but it's been closed down, I think).  One thing I'll say is I'd hate to be in charge of designing character sheets for this.  Character sheets would virtually have to be custom-made for each game.

[Ayyavazi]

Hey Egon,

Just to let you know, my original system actually used almost exactly what you are describing. And the ideas about different characters having access to different rules is an exact parallel of my own thoughts. How about this for a twist: What if the skills were just a lower level hierarchy of the stats? So from a stat might spring say five skills, which spring into another two each, or three, or whatever is necessary. This way there isn't a complex interaction between attributes and skills, just your ranking in whatever it is you are doing.

And yes, I think it is a good idea, so long as you leave design room for expandability beyond what is initially created. So, instead of releasing supplements with more skills and detail, just include a good chapter on how to create your own and work them into the system. Put the creation tools in the players hands (they'll hack the game anyway, they might as well know what they are doing). And tell them how things affect other things. For example, give mechanical examples of how adding a whole layer of combat skills encourages combat more, and what kind of statistical advantage is gained. This way, players will understand that making the rules needlessly complex will affect their game in negative ways, not positive ones. This assists them in having a better experience with the rules. Then, once all of that is done, people can start releasing their own supplements in addition to your own. You might even be able to start releasing different supplements to address the same issue, for example, two western supplements that focus on slightly different applications of the rules.

You could even experiment with different die rolling methods that yield similar odds and predictability. For example, compare White Wolf games, Sorcerer, Dogs in the Vineyard, D&D (any edition), Ninja Burger, and Warhammer Fantasy Roleplay. As long as players know which die system and numbers yield the appropriate odds, they can even tinker with the key rolling mechanics.

Just a thought.

--Norm

[visioNationstudios]

We're nearing the completion of a universal system that works similarly to how this breaks down, but the issue of expandability, as mentioned above, is nonexistant.  7 Stats are the end all, be all rolls for anything else done.  Skills (determined by the player, rather than choosing from a list) provide a bonus percentage to the stat they directly fall under.  Further specialization can come at a cheaper price than raising skill levels, but both are viable options.

Example:  Players X, Y, and Z needs to get over a wall.  Each have Strength at a 60%.
Player X has no applicable skill in this instance.  He rolls on his base Strength.
Player Y has the skill Climb at level 1, which gives him +5% to his Strength, letting him roll on a 65%.
Player Z has not only the skill Climb, but an additional skill, which he calls "Scale Sheer Surface", both giving +5%, so he rolls Strength on a 70%.

There are a myriad of ways that these can be combined, and the point costs are such that no build style is better than another.  So yeah, I would encourage you to work less on an exhaustive skill list and instead, come up with ways that encourages the players to be creative on their own.

[Vordark]

I like it!  I think you're right about possible confusion or overlap between some of the leaves of the tree, but depending on how task resolution is designed, you could make rolls based on a merger of one or more leaves.

At this point, I don't think I can provide more than a hierarchy level or two for skills, since skills will depend on the setting (fantasy? sci-fi? historical? modern?).  One option would be to do as Vordark is doing with his system, selecting a generic setting and tailoring the system to that setting, but such that it is easily modifiable to other settings.  Another option would be to provide a skill for everything in every genre (which obviously doesn't make a whole lot of sense, since people can always invent new settings with something you don't have), so you might provide Hacking and Lockpicking skills, but in a fantasy setting it would be collapsed into a single Security skill that would be used pretty much just for picking locks.  Personally, I think it makes more sense to keep setting-specific things, like a skill list, short and limited, and then to provide modules that expand on the base system and provide a specific application.

Given a choice between the two methods, I would seriously consider doing what I did (which it seems like you are).  Pick a setting everyone is familiar with, then come up with an example set of skills that GMs and players can use for inspiration.  There's a few reasons I took this approach.

A large skill list is nearly impossible to keep in the GMs head.  Call of Cthulhu can get away with it because the skills are right there on the character sheet.  Ditto for D&D 3.5.  Ditto for old WoD.  But a generic engine, by definition, is going to have different skills applicable to different games.  So each GM is going to have to go through the skill list in the core rules, pick out what he likes for his setting and take note of them all.  And odds are you're either going to forget a skill (meaning the GM has to create it anyway) or the GM is going to disagree with your vision of the skill (meaning he's going to hack it anyway).  I took the easy way out with Genesys and said "Here are some skills.  Use them to create your own.  I'm not lazy, it's a feature!"

Another trap I found myself in during the early days was creating skills that were only slightly different from other ones.  For example:  Persuade for convincing something to do something for you; Bargain for trying to haggle prices.  I still skim through my skill list (which sits at 41 skills, if I recall correctly) looking for ways to pare it down.  The first list I had was nutty!

[Vulpinoid]

This is reminding me of the wide and narrow skill sets in Mayfair's Chill from the early 1990s.

If you don't want something to be a key feature in your game, you just use the wide version of the skill field...but if you want to play up a specific genre, then you can expand out the skill field into a range of specific skills that really examine the depth of the activities concerned.

Some games might have "science" as a skill field, more technical gmaes might divide this up into "physics", "chemistry", "botany", etc.

Same applies for combat skills depending on how complicated you wanted a melee to become.

From the times I played it, the system seemed to work pretty well.

V

[Egonblaidd]

The first list I had was nutty!

You mean like my list of 145 skills for my heartbreaker?  Yeah, I can see how it might make sense to pare (pair?  pear?  banana?) it down a bit.  I don't exactly see the point of combining skills and attributes, as you were saying, Norm.  It seems like it would make more sense to either do away with skills or do away with attributes.  As far as I know, there isn't really a difference between "skills" and "attributes", but having both present indicates that the two groups interact with one another.  I mean, if, say, Acrobatics is a sub-attribute of Agility, then it's still an attribute and not a skill.  On the other hand, I could reterm everything as skills.  Athletics and Dexterity are plausible skills.  It would be simpler to have "only skills" or "only attributes", but I think having both skills and attributes lends itself to greater possibilities.

As for the interaction between skills and attributes, that might be a good place to present several alternatives.  Personally, I don't like the idea of simply adding the two values together.  Gaining a point in a skill is equivalent to gaining a point in an attribute, as for as that one skill is concerned.  I like the idea of some sort of multiplicative relationships, with fixed attributes and unlimited skill advancement.  That way, a low attribute will always gimp you, but with enough hard work you can counterbalance it with a high skill.  First off, by saying attributes are fixed and skills can advance without limit, you've just made a very clear distinction between attributes and skills (as I said above, in a general way they are the same thing).  Also, it immediately indicates that attributes are much more valuable.  A multiplicative relationship clarifies this.  If the relationship was additive then you could just increase your skill by a couple extra points to account for a low attribute.  If it's multiplicative, then a low attribute means you will need more ranks of a a skill in order to have the same proficiency as someone with a higher attribute.  I guess the point is, if I have both skills and attributes, then I want them to actually be different, mechanically.  Warhammer does it an interesting way, too.  All rolls are about attributes (which include "Weapon Skill" and "Ballistic Skill"), but trying to use a Basic skill that you don't have cuts your chance of success in half, and Advanced skills can't even be attempted.  Skills are on a have/don't have basis, although skills can be acquired multiple times to get a bonus.

Anyway, the only skill I can think of that makes sense yet is Knowledge.  Basically, it's anything that you can know.  What sort of creatures live in Giant Spider Forest?  That will be under Knowledge somewhere.  Which lever launches the catapult?  That's in knowledge.  What is the capital of Assyria?  Knowledge.  What is the wind-speed velocity of an unladen swallow?  Ok, now I'm stretching it.  You'd probably have to pass a Memory test to recall the Knowledge you need, unless it is some how "common knowledge", at least to your character (for example, a non-engineer might not know right away what a screwdriver was, but an engineer isn't going to "forget" something so basic), and after a Memory test you'd probably do some other Mind-type test to see if you can apply your knowledge to the situation at hand.  It kind of seems that any true "skill" would be a subset of Knowledge, though, as a person's abilities are a combination of innate talent (e.g. attributes) and learned skill (e.g. Knowledge).  However, I think "physical knowledge" i.e. how to drive, or walk a tight-rope, or block a punch, would be different from "academic knowledge", which would be more like mental knowledge like facts and trivia.  When walking a tight-rope you aren't thinking about how to distribute you weight so your center of mass lies directly over the rope, and how to shift your weight to provide an exact opposite torque if you start to fall off etc.  You aren't thinking about physics, you aren't "thinking" about anything, it's something learned at the physical level, not the mental level.

Anyway, I'm starting to ramble again.

Hmm, V posted while I was typing this up (no surprise, it happens to me a lot).  That sounds a lot like what I'm thinking I could really do with this.  Expand an aspect if it's important for your genre, collapse it if it's not.  If I every design a complete system with this, I think I'll name it Telescope.  Kind of odd the ideas that we associate with telescopes, considering the etymology (tele = far, skopein = to see, both from Greek), but it's that idea of collapsing and expanding that I want to capture.

[Vordark]

Egonblaidd:  You could choose one of the following for the relationship between attributes and skills...

1.  A higher attribute allows you to learn skills more quickly.  That is, it reduces the cost of the skills under it at character creation time and reduces the cost to improve them later.

2.  The number and level of skills you possess under a particular attribute increases the level of that attribute.  To oversimplify in order to create an example, if you have three Strength skills, each at level two, you have a six Strength.  This sort of arrangement was suggested by a friend of mine for a potential system a while ago.

The benefit for a high attribute in the first scenario is obvious, more skills at a higher level.  The benefit for a high attribute in the second scenario is dependent on when the characters use the attribute, and only the attribute (forcing a door open?  adding to damage?).

Yet more food for thought?

[TheDeadlyPlatypus]

I have no doubt that a system like that could be highly adaptable, but micromanaging 27 different sub-attributes would mean much longer character creation, and causes more micromanaging, which ultimately leads to min-maxing. Having so many different stats encourages munchkins, which is a real killer for a game. Honestly, it could be definitely more streamlined as well. I'd even go so far as to say that you could simply have the base three stats, each of the three sub-stats, and get rid of the sub-sub-stats. In place of the sub-sub-stats, you could simply substitute whatever skillset your gamers or GM decides upon, set in the bracket for whichever sub-stat they would fall under.

[Egonblaidd]

Platypus, you're missing the point.  I don't expect anyone (well, almost anyone) to play with all the sub-sub-attributes.  It's like some kind of fold out foldy-thing.  If you want to play a game that focuses more on physical attributes, then you fold out Body.  If your game is more about social interaction, the fold out Social.  You would choose which attributes you wanted to use for your game, and then stick with those.  If you WANT to play a highly complex game and min/max it, then you can.  If you want to use only the barest rules and focus on story, you can.  The point is to provide a system where all aspects of it can be collapsed or expanded, so that you can collapse what you won't use and expand what you will.

Example:  For a hack 'n' slash, expand Body, expand Athletics, drop Senses, expand Mind, drop Cunning.

Attributes: Strength, Endurance, Speed, Dexterity, Intelligence, Will, Social, seven total attributes, similar to DnD.

This attribute set will support warriors, rogues, and wizards in a typical dungeon crawl adventure with an average complexity.  If you wanted it more or less complex, you could expand or collapse some of these attributes.

In short, the point is to make it so that the game can be as simple or complex as the players want it to be without writing a whole new system.  You pretty much just pick and choose what you need, and the mechanics easily account for simplification.  In a game with little social interaction, then you'd probably collapse Social all the way, so you only had the one attribute for social interactions.  All skills that then used some sub-attribute of Social would simply revert to just using Social, and you'd have cut out a large portion of the attributes without having to rewrite the skill system (you'd probably collapse many of the social-type skills, too).

As for the relationship between attributes and skills, that will take some thought.  It will depend on how resolutions are handled (e.g. the dice mechanic), how attributes and skills advance, and what sort of relationship I want to have between attributes and skills.  I'm not sure what sort of dice mechanic I want (or if I even want to use dice, but hey, they're pretty versatile), so I'd have to look around and compare different options.  In keeping with the telescopic design, I'd probably want to have a dice mechanic with a layered complexity, e.g. that you can easily make more or less complex without changing the system much.  As for advancement, I'm loathe to hand out XP for combat, since that skews gameplay to favor combat.  Handing out a fixed XP amount at the end of a session seems both much simpler and balanced (that, and it avoids munchkinism).  I also like the idea of advancement through practice, use a skill to level it up.  It's an interesting idea to make resolution entirely dependent on skills (except for maybe attribute tests) but make attributes determine the XP cost to increase a skill.  In any case, I do want there to be a significant difference between skills and attributes, and that difference will lie mostly on how the two interact with one another.  Higher attribute = cheaper skill?  That's a significant effect.  Test difficulty (or facility) = attribute + skill?  That's a trivial difference.  Test difficulty (or facility) = attribute * skill?  That's less trivial, but requires something more (like fixed attributes) to make it entirely significant.

Anyway, finals are coming up and I have a lot of work to do (probably why I haven't been able to find playtesters for my heartbreaker), but soon I shall have some breathing room to work on RPGs.

[TheDeadlyPlatypus]

That makes sense. It just seems that without my somewhat limited grasp of RPGs that not many rulesets allow you to completely ignore complete attributes. The way I had interpreted what you were saying was a system akin to the AD&D optional character generation rules, where you roll for your stats and split them between two sub-stats each, just with a bunch more sub-stats on your system.

Keeping that in mind, if you were to completely forgo a certain branch of attributes, what would the consequences on gameplay be? Or are you suggesting that for whatever group, they can determine how complex it is by deciding which attribute tier they're going to use? If that's the case, it makes complete sense, but from a practical point of view, I can hardly see a group wanting to do that much accounting for a character to justify 27 different attributes.

But a mix and match approach would be interesting, although to expand certain areas and neglect others would likely make a system seem a bit incongruous. I'm not trying to shit on you idea or anything, but to make an aspect a lot more complicated is a bit counter-intuitive to focus. If you want a story to focus on combat, then just run something combat-oriented. To add in a bunch of extra stats in an attempt to flesh it out most often takes away from the gameplay instead of enhancing it.

However, in fairness, that's really a decision for whoever is running a game, but as a design for a catch-all system, it's good. My suggestion would be to beta it with different games for each different focus and see if it works out the way you expect it might. But then, that's probably obvious anyway.

[Egonblaidd]

Some background on me: I'm far more familiar with, say, Final Fantasy than I am with DnD.  So don't expect things to compare with DnD too well (as one of the first real RPGs, there will obviously be certain similarities).

To better understand what I'm trying to say, imagine that you have a skill, say, Threaten, that relied on your Influence, which is a sub-attribute of Charisma, which is a sub-attribute of Social.  Let's say you want to play a game more combat oriented, so you might want to use the Threaten skill, but there's no reason to keep track of a bunch of Social skills.  So you collapse Charisma.  Since Influence is a sub-attribute of Charisma, you will now use Charisma instead of Influence when you use Threaten.  If Social is collapsed, then you would use Social instead of Influence or Charisma for Threaten.  Most other Social skills will probably not be used, so the fact that you are using one attribute instead of nine to govern all of them isn't a big deal.

So it's more of a mix and match approach.  If similar skills and attributes aren't going to be used a whole lot, then you can condense them into a smaller group of skills and attributes.  Some people will want to have everything, but many people would prefer to focus on a couple aspects of gameplay, like combat or magic.  You CAN focus on everything, but part of the point is to allow the players to narrow it down to the specific elements they want, ANY element.  So far, all I have are attributes, but you should be able to see how even just the attributes can be narrowed down.  Let's say you want to design three modules for my system, in one the characters are players on a sports team, in another they are scientists at a lab, and in the third they are politicians.  In a game about sports, the attributes under Body become very important, while Mind and Social are less useful.  Similarly, scientists would be more focused on the Mind attributes.  And of course the politicians would need to make extensive use of the Social attributes.  In each of these three modules, you could collapse the two unused attributes and expand the used attributes, perhaps even all the way if you want it a little more complicated.  In a game about fighting evil, no one is going to care if you have eight ranks in Decorative Cake Frosting, therefore there's no point in having it as a separate skill, you can collapse it into a more general skill.  I want to make everything like that.  If it's not important, then you can group it with a similar set of things and treat them as one thing.

If I actually get around to making it and write up a rulebook, it will probably make more sense.

[TheDeadlyPlatypus]

The only big problem with the system of foldout attributes is if the players want to play a different style than the attributes that have been allocated would allow. It's still a very good idea, but I'm just trying to give you some food for thought so that you might be able to think of a way around it. And in all fairness though, just because it's a game based on combat, doesn't mean that you couldn't use a seemingly useless skill like cake decorating in a useful manner.

[Selene Tan]

The collapsing/expanding attributes-skill idea is pretty neat. But if you're using point-buy or some other fixed allocation for attributes, you'll need to be careful about the costs of broader attributes versus narrower ones. I've seen and played a few games where the designer wanted to emphasize one thing and de-emphasize something else by changing the number of skills. In one game, stealth was broken into several skills -- hide in shadows, move silently, pick pockets -- while combat skills were reduced to one-handed weapons and two-handed weapons. All characters started with the same number of points, but thieves had to buy up 3 separate skills to be effective, while warriors only needed one high skill. (It also didn't help that we wound up in a lot more combat situations than sneaky ones...)

One thing to keep in mind when picking the sub-attributes and sub-sub-attributes is that the main purpose of skills is to be able to differentiate between characters. The hide / move silently split in D&D is bad in that respect because it's rare to find a situation where you need to be silent but don't care about being seen (or vice versa), so the skill checks are always paired. Rogues have to spend twice as many skill points as they would otherwise, and you'll never find a Rogue who has maximized one but not the other. (Admittedly, they might have intentionally made Rogues spend twice as many skill points on sneaking since Rogues get more skill points than any other class.)

In your current set, I think Confidence and Self Esteem are pretty close; I'm having trouble picturing a character who is very confident but has low self-esteem, or vice versa. Perhaps swap one of them out for Courage?

[Egonblaidd]

I'll admit, I had trouble coming up with Social type skills, so they'll probably need reworking.  I'd think Courage would be more of a Will type attribute, though, probably a combination of Self Control and Determination.  What I was thinking when I chose Confidence and Self Esteem was that Confidence is basically making other people think that at least you think you know what you're talking about.  If you're trying to explain to an NPC why what they're trying to do won't work (even if in reality it would), then a higher Confidence will make it easier to convince them that you know what you're talking about.  Of course, if the other person is knowledgeable enough, then it would just make you look like an idiot, but that should already be obvious.  Self Esteem, on the other hand, might be a type of emotional Endurance, which would affect how much emotional damage you could take before bad things started to happen.  This might contrast with Composure by being different types of damage; a loss of Composure makes you angry (probably a short-term type of emotional damage), while a loss of Self Esteem makes you depressed (probably a long-term type of emotional damage).