Classic versus learning: a computerized RPG system

[George the Flea]

This may not be an appropriate post to the Forge, since it seems to be mainly focused on pencil and paper RPG design.  However, people here obviously have well considered opinions and long experience with RPG's, and I need any input I can get.

A little background to see where I'm coming from: I'm an avid player of RPG computer games, but have never played a pencil and paper game (intriguing as they are).  I have limited experience with the D&D system, but only through games such as Baldur's Gate and Neverwinter Nights.  As such, I don't really have much breadth in my knowledge of RPG's.

I am currently involved in designing a computer game RPG, and am in the initial planning stages.  My dilemma is that I am unsure what type of character development system I should use, and I have been unable to find any good resources.  As such, any links or suggestions you can make will be much appreciated.

As for specifics: I am debating whether to use a classic experience/leveling system to drive the game (gain X experience for an act, eventually get to the next level, and purchase better abilities/feats/skills/whatever as a result of leveling), or if I should attempt to build a system based on learning algorithms (get better at a skill by practicing it).  I'm leaning toward the latter system, but because it is such a total break with the vast majority of computer RPG systems I have been unable to find any discussion or information that could help me figure out where I'm headed.

Are there any systems that you know of (or discussions about theory/implementation of such a system) like what I've described?

Do you have any thoughts on such a system?

Obviously the kind of system that I'm thinking of is really only practical when a computer can handle all the calculations necessary to control character development, but I really would appreciate any help I can get.

Apologies in advance for this fairly broad, ill defined topic.  I'm definitely floundering around trying to figure out if this would work, probably without even knowing the language necessarily to converse about it intelligently.  Let me know if you have questions or need clarification for what I'm trying to figure out.

[Mark D. Eddy]

I don't know about computer RPG's but the old RuneQuest system (later used in their BRP System, I think) had a process where sucessful use of a skill called for a check to see if that skill improved at the end of the gaming session. I would guess that you could set up an algorithm where if you have a successful use of a check, a tally is made, and after some number of tallies (determined by you), either the skill goes up automatically or there is a check that determines (based on the current level of the skill) whether or not improvement has occured.

In fact, with a computer system, you could even have a "you are eligible to level up [skill x]. Do you wish to try to level up, or keep practicing?" and if they keep practicing, their chance once they do try to level up improves.

I don't know if any of this helps...

[Simon Kamber]

If you go with a practice system, remember to keep something up to the players. They used something like that in Dungeon Siege, and I personally found that since there was little control of the character, the player actually didn't have much to do. Most of the fights involved sitting there and watching the characters fight automatically, and clicking a potion once in a while to keep them from dying. And character advancement happened automatically, so that's a dead end too.

So, if you go with a practice system, just make sure you have something for the players to do when they're in combat.

[George the Flea]

So, if you go with a practice system, just make sure you have something for the players to do when they're in combat.

This shouldn't be a problem; the game engine that we're planning to build this game on is actually a first person shooter engine, so combat will be almost exclusively player driven.  Although I haven't really played it much, I believe Morrowind is the most similiar to what our gameplay will likely end up being.

Thanks for the info Mark!  I'll have to track down the rules for RuneQuest and see if I can find any inspiration in it.

Anyway, here's a bit more of what I'm thinking about; I realized that I didn't get into it very much in my initial post.  Any commentary would be great.

At the beginning of the game the player sets an initial group of attributes for their character (things like strength, dexterity, intelligence).  I haven't decided yet if setting these stats will be through a point system, a roll system, or even perhaps if the engine will assign specific values based on dialog responses from the character ("I'm very strong, and moderately intelligent" -- engine assigns a random stat to strength in a certain high range, a random stat to intelligence in a medium range).

Anyway, over the course of the actual game the basic attributes change very little, and can only be changed by practicing other skills or doing activities specifically aimed at improving base stats (say, weightlifting).

There are then disciplines or groups of skills (say, melee combat).  Your initial ability in skills in those disciplines is set by calculations based on base attributes, and how well you can learn the skills in those disciplines is also based on the attributes.  So if my character is very strong, then I'll have a sight advantage with any type of melee weapon over someone with less strength, even if I haven't trained at a specific melee skill (like swordplay).

Some skills can be learned right off the bat by anyone who practices them, and then you can purchase or otherwise "learn" new skills by talking to NPC's, completing certain quests, or whatever.  To get better at a skill, you practice it (so fight with a sword if you want to be better at swordplay).

Obviously players control what their characters learn by controlling what they do, but I'm considering also allowing the player to choose a focus/talent/etc., perhaps in a certain discipline.  Then the skills in the discipline that recieve the focus are learned a little quicker, and the focus could of course be changed if say the character maxed out the skills in the currently focused discipline.

Thoughts?  Could this work, and if it did would it be preferable to the classic leveling system (or some variation thereof)?  This system would take a lot of control out the player's hands, but by making it unneccessary to perform calculations to try and maximize player effectiveness I'm hoping to make the whole experience more immersive.

Thanks for your help!

[Silmenume]

Hey George,

Welcome to the Forge!

I wish you the best of luck in your endeavor!

I don't have anything to offer about what would be better or why.  However, I do remember a computer game that was first person and used the same basic use and improve method of advancement that you are wondering about.  That game was Elder Scrolls II: Dagger Fall, was developed by Bethesda and published by Virgin Interactive.

I am only familiar with Dagger Fall, but apparently the first game in the series was Elder Scrolls: Arena and the follow up to Dagger Fall was Elder Scrolls III: Morrowind.  Apparently an upcoming  addition to the series is called Elder Scrolls IV: Oblivian.

I don't have anything else to offer, but I hope that by giving you these game names you might find some leads or information that reflects in some way what you are looking to do.

[Callan S.]

Hi George, welcome to the forge,

I see your sort of floundering about between a leveling or a practising system, unsure which to choose. I find when I'm not sure what to choose, it's time to sit down and write out what the hell each does, rather than intuitively guessing which is a bit more 'right' than the other.

One great idea I got from the here is to write a quick transcript. Write out how you imagine some play would happen. Never mind how you'd get it to work, just write out your dream of how it would go.

But basically I've come to believe that skill improvements are three things: A reward for playing, an incentive to play more, and a way to insinuate the players choices into all elements of the game.

I'll talk about the last one. For example, say you program makes a spot check for the player. Now, it's just a dice roll...the players skill doesn't come into that at all. It's nothing to do with him. Until he's able to influence the bonus to that roll, by his choices (distributing points or choosing what to practice). In this way hundreds of rolls can be influenced by the players choice, without him making a choice directly before or after that roll. I think it's why computer games which have spreadable points call themselves RPG's...they've associated that name with that feeling of player choice having an influence this way.

[Simon Kamber]

This shouldn't be a problem; the game engine that we're planning to build this game on is actually a first person shooter engine, so combat will be almost exclusively player driven.  Although I haven't really played it much, I believe Morrowind is the most similiar to what our gameplay will likely end up being.

Sadly, I never got around to playing Morrowind :( But the idea isn't bad.

At the beginning of the game the player sets an initial group of attributes for their character (things like strength, dexterity, intelligence).  I haven't decided yet if setting these stats will be through a point system, a roll system, or even perhaps if the engine will assign specific values based on dialog responses from the character ("I'm very strong, and moderately intelligent" -- engine assigns a random stat to strength in a certain high range, a random stat to intelligence in a medium range).

I don't like rolling in computer games. Not when the outcome of one roll has such wide consequences for the rest of the game. In Baldurs Gate, I quickly learned to keep pressing the roll button until I had a high set of stats. It's simply not fun that your defeat stems from a bad output from a randomly generated number, rather than something you can actually effect. So, personally I'd say go for the simple system without too many numbers to confuse a beginner (Very strong and moderately intellegent sort of system, but with every "very strong and moderately intellegent" being the same as the previous one, technically).

Anyway, over the course of the actual game the basic attributes change very little, and can only be changed by practicing other skills or doing activities specifically aimed at improving base stats (say, weightlifting).

Not bad.

There are then disciplines or groups of skills (say, melee combat).  Your initial ability in skills in those disciplines is set by calculations based on base attributes, and how well you can learn the skills in those disciplines is also based on the attributes.  So if my character is very strong, then I'll have a sight advantage with any type of melee weapon over someone with less strength, even if I haven't trained at a specific melee skill (like swordplay).

Unless the characters for some reason start off totally incompetent because of some story element, let them have a say. It gives the feeling of control right away. Making them able to influence their character makes it more like THEIR character.

Obviously players control what their characters learn by controlling what they do, but I'm considering also allowing the player to choose a focus/talent/etc., perhaps in a certain discipline.  Then the skills in the discipline that recieve the focus are learned a little quicker, and the focus could of course be changed if say the character maxed out the skills in the currently focused discipline.

Sounds like a nice idea :)

Thoughts?  Could this work, and if it did would it be preferable to the classic leveling system (or some variation thereof)?  This system would take a lot of control out the player's hands, but by making it unneccessary to perform calculations to try and maximize player effectiveness I'm hoping to make the whole experience more immersive.

The basic idea isn't bad. But I'm a little wary that creating the character ends up becoming a consequence of playing the game, and not a part of playing the game.

[xenopulse]

If you want to create a skill-based system, there's no way around taking a look at Morrowind. It's an incredibly popular (and excellent) game, and you need to know your competition and what's already been done to make something better/different.

In short, in Morrowind, you train skills by using them (or by paying for training). You initially pick 5 major and minor skills each. Gain ten ranks in any combination of these by using them, and you can level up, which lets you increase three of your attributes. Your attributes can rise more or less depending on which skills you used. So if you bashed things a lot and used other strength-based skills, your strength can be raised higher (up to 5 points) than if you didn't use any strength-based skills (just 1 point).

This system is great in that it really makes you appreciate the time you put into whatever skill you're using, it allows for leveling without combat (if you're all into magic, alchemy, lockpicking, etc.), and it just makes sense.

[daMoose_Neo]

I don't like rolling in computer games. Not when the outcome of one roll has such wide consequences for the rest of the game. In Baldurs Gate, I quickly learned to keep pressing the roll button until I had a high set of stats. It's simply not fun that your defeat stems from a bad output from a randomly generated number, rather than something you can actually effect.

Agreed.
When you roll low on chargen at the table, you can work things out with the GM or restructure campaigns or adventures. PCs, you just can't do that. Theres the difficulty level, thats what it is, the goal is to do this, not that.

If you're going with a tabletop approach, go with some quasi table top methods. Work some kind of scale into the system by which the characters are measured against. Instead of saying the badguy has 15 HP when the character should have 10, work it so the badguy has 1.5 times the characters HP, so that there is a statagy challenge, not a numbers.

[George the Flea]

Thank you all for your responses!

I will definitely have to look more closely into the Elder Scrolls games.  I'd given them a cursory exploration a while back and realized Morrowind was similar to what I wanted to do (because it was first person perspective), but I hadn't really looked into the skill system.  I'll definitey give that a go.

Straight rolling for stats is definitely out.  I can remember all too well sitting in front of the computer hitting the reroll button for Baldur's Gate, trying to get at least decent numbers.  Pain, pain, pain.

I'll definitely keep in mind writing a system for scaling opponent difficulty based on player skill.  That's a really good idea.

Brainstorming out an ideal playing session is a good idea, too.  That would definitely give me a goal to work back from.

Thanks again for all the comments!

[M. J. Young]

Most of what I'd have said about the advancement process has been said; but I've a thought on the character generation process.

Rather than "generate" a character, why not just pick one? I've seen a few CRPGs in which the players pick from one of several characters at the beginning of play, and use those values. I'm thinking to expand that idea. Create a list of perhaps fifty to a hundred different characters, each with different initial attributes and skills. Then rank them according to how much of a handicap this particular character would be compared to the others. List them in order from the character which would be the easiest to play (for reasons of strength and of the intuitive nature of the strategies involved--e.g., it might be that a character with maximum strength and a character with maximum dexterity would both have the same chance of winning, but that you'd have to think more about a great number of options if you're using the dexterity).

That way players can pick a character that will be an easy starting point or one that will be more difficult. You avoid the Baldur's Gate problem of players rerolling dozens of times to get the strong character (as they can just pick the strong character), but at the same time you don't eliminate the option of starting with the weak character for those who really want the challenge. And by ranking them you give some guidance concerning which is the stronger or weaker character.

--M. J. Young

[George the Flea]

Rather than "generate" a character, why not just pick one?

That's actually a really interesting idea.  Thinking about that, it would actually solve a lot of problems: I don't really want to let the player see the numbers involved (reasons here for the curious), which is solved with a list of premade characters.  If I let the character engage in a dialogue selection like that of Morrowind, I run the risk of letting them create an uber-character, unless I factor in some artificially limiting algorithms.  With a list, though, I could control for discrepancies, and still allow the player to customize the name, gender, and other largely superficial characteristics.

I really like that.  Thanks for that idea!

[Callan S.]

If I let the character engage in a dialogue selection like that of Morrowind, I run the risk of letting them create an uber-character, unless I factor in some artificially limiting algorithms

Ah, what's your category with this game? You might not know about GNS, so I'll just list two here:

Simulationist - this is like exploring a game world. Personally I think it has the appeal that nature documentaries have and such like, in that you engage it to learn and sort of experience the contents.

Gamist - This is about player skill in just the same way chess or sports are. The game presents a challenge in the form of a world, and partly using the system/partly using their knowledge of that world (and a lot of the real one), they see if they have the skill to overcome the challenge.

A quote from the link you gave

The fact that a traditional RPG is essentially a numerical simulation has spawned a number of very annoying trends in player behavior.  Most of these types of behavior can be subsumed under the term "numbercrunching."  Also called "min/maxing," numbercrunching largely involves the study of the game's numerical systems and figuring out how to use it to the player's best advantage.

It's a long document so I've scanned through it. The author (yourself?) is very simulationist. Further, it's assumed that while deciding to use cover in game is an okay choice for a player to make, making choices outside game play like looking at stats, is just wrong and waiting for the numbers to stack up or some such.

It's not a broadminded document. Elements that support another play style (gamist style), it simply puts down as stupid mistakes (In "Hit points! PAH!" style). BOTH of those styles are valid, and once you get a handle on that you'll have a better focus on the one you want, rather than reflexively flinching away from miss-associated phobia of elements, like below:

It is my firm belief that the axiom "most players are self-centered bastards who will ruin other players' experience at the drop of a hat" is greatly exacerbated by this numerical obsession.  Why do players steal kills from other players?  Because doing so will help their numbers increase.  Why do players exploit bugs to kill monsters (or players) with relative ease?  Because doing so will increase their numbers.  Why do players use cheats and plugins that give them unfair advantages in the game world?  You get the picture

Numbers shape play, they don't just lead to everyone screwing everyone else. You don't flinch away from numbers, you make the numbers form the behaviour you want.

For example (a simulationist one, IMO): Imagine your character has been climbing a hill for hours, pushing through dense undergrowth. And when he get's to the top, he see's an amazing vista in the valley below! PING! He get's some points...but no, it's not XP that'll raise his fighting skill or strength. No, the message shows that his area knowledge has increased and these points mean he is now that much closer to learning where a certain special cave is where amazing crystals are (they are said to be alive or such), assuming he doesn't find it first by exploring really thoroughly.

That's shaping play, shaping it to support exploration.

Once in a while, he may receive a system message telling him that he has learned something new about pottery, but these messages should be unreliable and ambiguous.  He may even be able to compete for titles in various contests of skill, but this is only an indicator of prowess, not a measurable figure that you can watch increase as you fight your eight millionth orc.  Sure, there are players who will still go camp the goblins for "skill," but he can't really be sure it's doing him all that much good, and if the designer has been building his system holistically, it's not.

I mean, here it's basically saying the player shouldn't advance (from act's of his choice, like killing goblins) despite having a system for it. If you follow this it sounds like your going to go to the effort of designing an advancement system, then go to the effort of nuetering that system.

Just don't have this advancement to begin with. You don't have to have advancement of the traditional type. Shape play with other designs, like the example I gave before. That's advancement in the direction you do want to go.

I think I've written too much, inspired by some IMO wrong headedness.

[George the Flea]

Thanks for the response Noon!

I'm not sure how you would classify my game.  I think it would probably fall best under gamism (as I understand it from your post and a quick read of some of the stuff in the GNS forum), since the main point of the game is to accomplish quests in order to avert disaster.  (I know this sounds like a bazillion other games, but the disaster is of an unusual kind and the game takes place on a college campus.)  Anyway, the game will also hopefully have narrativistic aspects as well, since quests will have multiple ways to beat them, and depending on what quests you do (and how you do them) the ending will be different.  Simulationism will be evident to some extent (planning on leaving little Easter eggs around), but since the base map is fairly small, it will be largely irrelevant.

I didn't write that outside link, by the way.  Sorry if I presented it as such.

However, I am using quite a lot of the theory there to guide design.  Granted, the guy is extremely outspoken and generally goes for trashing anything that's been done rather than discussing any pros for the things that currently exist ("It's not a broad minded document" is a bit of an understatement), but on the whole a lot of his basic ideas make a lot of sense to me.  (Good to note that he's ranting about MMORPG's, too, so a good portion of his discussion is moot in my case.)

A lot of his issues with numbers stems from the fact that most if not all MMORPG's on the market have no level cap, and so end up catering to a bunch of uber characters and leaving the low level characters behind.  He argues that by focusing on the literal numbers that make up a character, a good number of players loose focus of the things that matter (role playing, exploring, having fun) and just power level until they can destroy entire cities on their own.  I think that he actually would want your climbing a hill idea to work, because that is an example of numbers not corrupting the system.  On the other hand, he might argue that if players can see a direct advantage for climbing a hill that you'll have hundreds of people just climbing the hill over and over again to increase their stats.  Not exactly role playing.

To explain why I agree with the third quote, I think I really have to describe the one MMORPG that I ever really got into:

Clan Lord is a Mac-only MMORPG with some very different rules and gameplay (and pretty old school graphics, but if you get into the game you don't even notice them).  In order to advance in Clan Lord you have to find an NPC trainer.  Trainers will tell you things like "I will help you improve your body" or other general things such as this.  Once you've agreed to train with them all of the experience that you gain from fighting goes into gaining ranks in the trainer's associated skill.  However, you don't have any way to see how many ranks you have.  When you gain one you usually get a message saying you feel a little stronger (if training for your body), and when you talk to the trainer again they'll tell you something like "You are one of my better students" or something else that gives you a general sense of where you are in advancement, but doesn't really pinpoint it.

Now it's true that some people count the ranks that they get.  But because the system doesn't allow you to just look at the numbers that make up your character, it is in many ways more immersive and fun.  You fight low level creatures not to power level, but because you want to get out of town and be able to explore more of the game world.  Not seeing stats can be a little weird, but it definitely shapes game play in different ways than are seen in many RPG's.

(I feel it incumbent on me to note that despite my rather glowing description of Clan Lord's advancement system, the game has its share of problems, which after a few months overwhelmed its positive aspects for me.  And I'm not designing a Clan Lord-esque system, I just thought it was a good example of no-stats gameplay that worked, or at least mostly worked.)

Anyway, I am hoping to shape gameplay by awarding people for what they do.  So if someone wants to solve problems with a sword all the time, then they'll get very good at sword skill, whereas someone else might try to persuade people to their point of view a lot and become really good at persuasion.  I'm planning to have indicators that skills are increasing (likely visual, such as fading from one color which means not-so-good to another which means pretty good, to a third which means butt-whupping-good).  Sort of a practice makes better system.

Another reason I would like numbers hidden, or at least represented visually or textually, is because I've experienced numbers drawing me out of the game myself.  I can't count the times that I've been playing Baldur's Gate or a similar RPG, and then level up and completely get broken out of the game play because I suddenly have to start looking up what will happen if I add a point here versus there.

Okay, I've argued Mu's case long enough.  Let me know if you have any questions or critiques of the system (or theory the system is based on).  Thanks again for your response!

[Callan S.]

Thanks for the response Noon!

I'm not sure how you would classify my game.  I think it would probably fall best under gamism (as I understand it from your post and a quick read of some of the stuff in the GNS forum), since the main point of the game is to accomplish quests in order to avert disaster.  (I know this sounds like a bazillion other games, but the disaster is of an unusual kind and the game takes place on a college campus.)  Anyway, the game will also hopefully have narrativistic aspects as well, since quests will have multiple ways to beat them, and depending on what quests you do (and how you do them) the ending will be different.  Simulationism will be evident to some extent (planning on leaving little Easter eggs around), but since the base map is fairly small, it will be largely irrelevant.

The quests to avert disaster thing doesn't really indicate anything in terms of GNS. It could be part of any of the three.

On multiple ways to beat a quest, that isn't narrativism as I understand it. It could be gamism or sim. For example, Say I can sneak in to take a locket or walk up an con the owner out of it.

Now, if there's absolutely nothing I as a player can do to contribute to either, tactically (or even by reflexes), it's promoting sim, ie, you should be happy just watching how things go.

If there are lots of choices, even if it's just previously deciding to put points in skills like sneak or con and then evaluating the risks involved with each (which your Clan lords example sort of neuters by it's lack of info for the player), then it's gamism, as my tactics as a player matter.

And then there is a terrible middle ground (though it can be good), where it's supposed to be sim, but there are things you can do while sneaking or conning, but the sim people don't like it because they want to enjoy rather than strategize, and the gamists don't see it as enough input for all the 'waiting around' as the con/sneak happens.

But sometimes, it hits a perfect blend. I think this can only happen effectively in table top groups, where they customise play. Given human diversity, I don't think (IMO) you can get the right blend for a significant demographic.

Simulationism isn't (just) stuff like easter eggs. A focus on graphics is one type of focus on simulationist enjoyment (though not a primary one, as demonstrated from you Clan lords example), IMO.

On narrativism, for the player that isn't about beating the quest at all. It is for the PC, but not the player. The quest is just a platform for the player to express his PC. For example, the PC is given a choice between killing his betrayer cousin and letting a town starve. Which does he choose? This isn't like a quest where you need to make the town happy to win, it's basically about what your PC does when given a terrible choice. PS: I think if he can do both options to his satisfaction (feed the town, jail and not kill his cousin), you've let it slip into gamism. IMO.

I didn't write that outside link, by the way.  Sorry if I presented it as such.

However, I am using quite a lot of the theory there to guide design.  Granted, the guy is extremely outspoken and generally goes for trashing anything that's been done rather than discussing any pros for the things that currently exist ("It's not a broad minded document" is a bit of an understatement), but on the whole a lot of his basic ideas make a lot of sense to me.  (Good to note that he's ranting about MMORPG's, too, so a good portion of his discussion is moot in my case.)

I think there were some things to be had there as well. The problem is, every time he see's a gamist design with flaws in the gamism, he see's both the flaws and the good gamist design as flaws, and is trying to 'fix' them with simulationism. It's like 'fixing' chinese food with mayonaise...it's just yuck!

There's some good points/observations there, but trying to pick them out can be dangerous since you might end up adopting his stance in certain areas, from what you take away. Read with care! :)

A lot of his issues with numbers stems from the fact that most if not all MMORPG's on the market have no level cap, and so end up catering to a bunch of uber characters and leaving the low level characters behind.  He argues that by focusing on the literal numbers that make up a character, a good number of players loose focus of the things that matter (role playing, exploring, having fun) and just power level until they can destroy entire cities on their own.  I think that he actually would want your climbing a hill idea to work, because that is an example of numbers not corrupting the system.  On the other hand, he might argue that if players can see a direct advantage for climbing a hill that you'll have hundreds of people just climbing the hill over and over again to increase their stats.  Not exactly role playing.

Yes, the prob is he'd throw that idea away wholesale rather than put a binary flag on each hill so you can only go up to each once for a bonus. He chucks idea's like using numbers as if the whole idea is soiled. That doesn't help.

*snip* Now it's true that some people count the ranks that they get.  But because the system doesn't allow you to just look at the numbers that make up your character, it is in many ways more immersive and fun.  You fight low level creatures not to power level, but because you want to get out of town and be able to explore more of the game world.  Not seeing stats can be a little weird, but it definitely shapes game play in different ways than are seen in many RPG's.

Yeah, a simulationist way (a completely valid style, don't get me wrong). If you shuffle away the stats, it hides that type of reward and something like exploring becomes the more visible and thus dominant reward.

I'd make messages about exploring the world a lot more clear cut, to further emphasize this reward type.

(I feel it incumbent on me to note that despite my rather glowing description of Clan Lord's advancement system, the game has its share of problems, which after a few months overwhelmed its positive aspects for me.  And I'm not designing a Clan Lord-esque system, I just thought it was a good example of no-stats gameplay that worked, or at least mostly worked.)

What were the problems you experienced?

Anyway, I am hoping to shape gameplay by awarding people for what they do.  So if someone wants to solve problems with a sword all the time, then they'll get very good at sword skill, whereas someone else might try to persuade people to their point of view a lot and become really good at persuasion.  I'm planning to have indicators that skills are increasing (likely visual, such as fading from one color which means not-so-good to another which means pretty good, to a third which means butt-whupping-good).  Sort of a practice makes better system.

Another reason I would like numbers hidden, or at least represented visually or textually, is because I've experienced numbers drawing me out of the game myself.  I can't count the times that I've been playing Baldur's Gate or a similar RPG, and then level up and completely get broken out of the game play because I suddenly have to start looking up what will happen if I add a point here versus there.*snip*

It shaped your play into gamist, basically. I think your practice system is knife edged between sim and gamism. I think you want the experience of working with a sword...all those difficult times of practicing away, of hard battles with monsters since you just weren't that adept and now today you are an adept swordsman, with a rich past full of practice, before you got to this stage. Something to think about as you swing your sword with amazing accuracy.

While someone else might leave a book on their mouse button for an hour, so they can get that dragon and make some boots out of it.

I think that's the knife edge your courting with a practice system. I'd give some idea's on how to shape play so the experience is enjoyed, but first can you see the duality here?