next »

[gimymblert]

When talking about RPG people says the mechanics is about developing the character, But progressing in RPG generally involve killing as a primary mechanics, most lock are behind a kill gate (boss, items, xp, skills). But the ideal of rpg is ROLEPLAY, ie focusing on character advancement on a narrative and psychological level. I present below an analysis of the problem and then a solution that is just a little variation

Rpg have much more than killing of course:

1. You get story which gameplay is reduced to talking to NPC to trigger the next part, or deliver an item to trigger next part, that next part generally is a new fighting area or a new boss to kill.

2. Yay you are going to socially interact with npc a bit more than information dumping, the gameplay is for you to unlock a new companion for ease of kill, to get buff to kill better, or just unlock the good ending whatever. Moral system does not change much.

3. Farming? fishing? mining? exploring? collecting? crafting? fighting buff.

Basically anything that isn't fighting is just a nice diversion because fighting is the primary mean of challenge and progression, anything less is just an optimization loop for variety tied back to it. Just look at manual, secondary mechanics just need 1 page or just a paragraphe while the fighting get many page. Statistic is generally all about fighting, you may have up to 6 stat (defensive stat like wil, agi, def and offensive stat like int, dex, atk) Social skill can have up to 2 stat and generally are just unlock in terms of gameplay.

Character progression is therefore essentially physical (you get stronger) but most complain against RPG (mostly CRPG) is that the primary draw is immersion and story through rich interaction with a believable world. That mean people expect to build and interact complex character with emotional arc, the reality is more the genocide simulation of goblin and trolls (and cute rabbit for Jrpg). Most of actual character development happen through script scene or adhoc moral choice, not much in gameplay. Choice of character is generally choice of physical appearance and choice of physical prowess and interaction (class). Sometimes you may have some social aspect (race in elder scrolls) but it is another variation of unlocks.

Is there a simple way to change the mechanics while not necessarily create a new genre or even a new format or even a complete new system?

One problem is that the primary mechanics for progressing in RPG is level up. Basically you do stuff, get exp and level up, leveling up open new opportunities or gate. Most games have fighting as a primary source of xp and that is the main problem. The fact is combat is the most rewarding interaction, it is easy to understand, scale easily, is challenging with inherent risk/reward system, is flexible and have infinite variations. To counter act this aspect other games offer other sources of exp (questing, exploring, talking to npc, good dialog choice, any interaction, etc...) But most of them are not deep system with compelling anticipation and little readable risk/reward system.

I have a small solution that may advance the problem a little further but not shamble anything. The main idea is to change the perspective to reward actual character advancement and promote "ideal" role play rather than killing and that make sense on a narrative level. The idea is to promote behavior that is in "character" with a "class or traits" based reward system that change for each class. Basically a killer kill, a fighter fight and pacific avoid all these trouble.

For a roleplay system to work rewarding the pacifist when he kill or making inappropriate behaviour should not happen. Therefore there should be a "role exp" progression where behaviour are weighted by appropriateness to the role, there should be a "out of character meter" that fill with equally weighted behavior which should be a kind of punishment system, the punishment being a "character crisis" where the character is throw into a state where he should redefine his traits or class through behavior and allow the player to shift class, losing advantage but also constrain base on the previous class. It also allow the game to react to contradiction inside the character.

Basically it mean that the world became a web of different opportunities for different class of character. Thief get punished if they spot valuable things and do not steal them, or cruel fighter for giving mercy, and they are rewards the more they reinforce their character. It also give a natural progression to character arc and let the world react accordingly to player action.

Is this new? It looks like some game are already doing it! Well it's not entirely new, what I'm proposing is a streamlining of various mechanics into a single one for legibility. It used to be scatter through race, morals and exp sources and has such was under develop. My new system put focus much more on narrative affordance and actual legibility of "in character" progression in the world. So while not entirely new it's still a novel way to do the same things more efficiently. It also put a nice risk reward system on top of mechanics that have clumsy one and restrain the efficiency of a given mechanics to break character.

However the simple exp system has plenty years of refining and is an efficient, if constrain, progression mechanics. One major advantage is that it nicely map short terms (exp) mid terms (skill tree and level up) and long terms rewards. It is also pretty clear, kill stuff - get exp - get better at killing stuff - get better exp >> kill easy boss, kill hard boss, kill last boss.

My problem is that I'm not so sure how to reward player for progress and how to provide a progression loop. And what about character crisis, should this swipe clean the character back to a new class?

[s0]

sounds solid but i think i like the fable system more (your stats influence how npcs react to you and are the sum of all your actions).

the problem i'm seeing with your system is that if being a "perfect thief" or a "perfect warrior" or w/e is the best way to play the game (aka min-maxing) then you're essentially taking out the entire creative aspect and reducing roleplaying to a bunch of boring stereotypes. this is just my personal philosophy, but i don't think roleplaying should necessarily be a challenge.

you could counter the stereotype a bit by not using traditional character classes but letting the player freely choose a bunch of personality traits they would have to act according to during character creation, kinda like a cross between the sims 3 and choosing skills in the elder scrolls.

or

My problem is that I'm not so sure how to reward player for progress and how to provide a progression loop. And what about character crisis, should this swipe clean the character back to a new class?

how about you start the game without a class but are eventually locked into one through your actions? like if you steal and sneak enough you become a thief, if you kill enough monsters you become a fighter etc. the "character crisis" would reset you to the classless state.

[gimymblert]

Yeah it's supposed to be traits based, as each class has a bunch of traits, of course you could go for open class too, but the main concern is legibility first to promote the idea before diluting it. Also instead of traits you could piece together backstory facts.

On Multiple stats, it's a streamlining of that too, generally you have to check each stat to anticipate reaction, a single class system allow to get all data instantly with a single data, it's more a "framework" that a new mechanics, it is much more focused. It's also important for teaching player to break habit by focusing his mind on the right frame, ie "in character role play"

On locking action, that's the purpose of role play actually, however the idea of starting blank to evolve to a class might help solve the progression system, each "level" is a refinement of the nuance of the character. By the way the player is never truly "locked" since there is the "character crisis" mechanics", which might be an appropriate quest that deal with the kind of out of character interaction you did. Also the game can create events tailored to the level of crisis or to build tension by throwing pertinent moral dilemma to the character class.

Of course we can break free from stereotype, I used stock class reference to communicate the idea more clearly, it can work for simple stereotype game too, but the true strength is that it free RPG from fighting by giving another clear focus for the gameplay, the flexibility of the system is the main advantage, it create new opportunities and refresh older.

Actually that idea is a sub idea from a larger thought on narrative game, I was thinking about how the game need to profile the player first since he is the undefined character in the story, basically like a reverse "exposition" where the player tell his story to the game, not the other way around. Instead of building the complex system I had in mind (a kind of true conversation with the game by passing value and fact to the game world through behaviour, kinda like the game would be able to understand your "argument", "change is mind" and react), I found that this simple system would do the jobs while avoiding "the big dream syndrome".

EDIT:
It also allow player to be closer to DnD character sheet where the player is able to define his relation to the world more precisely.. Ie how people reacts, what is character arc, his character backstory and character destiny, all with their constrain/obligation influence: for example imagine the player having perks for letting his character "die bravely in a battle" and set a "father's kid revenge" arc. There is huge opportunities for letting player's expressions.

Notice that affordance also have neutral action that does yield xp, so it's not just punishment and reward ie constrain, only in and out character action are keys. Actually the player might have more rewards the harder he puts constrain. We can even have a system to take into account contradiction in traits that will set up an inevitable crisis (the pacifist killer is not going to end well), or just let it happen as the system resolve the contradiction itself through the crisis mechanics, it enforce a character arc and force to choose a new state for the character >>> instant character development!

[antoniodamala]

That's some very pertinent topic.
 
Well, the whole idea of the character crisis + attitude class mechanics seems very functional, but also too imposing. I don't like the idea of list-based progression, where you must look to it every five minutes because you don't remember how you were supposed to act. For example in Oblivion, you choose like 7 skills to be your main, so you will only level up if you use these skills. This leads to very weird situations, like if you are a warrior, you can freely maximize all magic skills and that's will not interfere in what you are. Which clearly doesn't make sense.

And roleplaying is not that black & white, it's mostly grey areas.

When it comes to actions, most of the time you just go ahead and do, but even when you do something that's not actually you, you don't stop being what you are. And mainly, what you think you are, it's not necessary what other people think you are.

So what i propose is this:
What if there were reactions based stats. For example, you may act nice with most people, but you once stole something from somebody, so no matter what the whole city thinks of you, to her/him you are a thief (In fact, how you are nice to everyone, this person thinks you are an even greater asshole). So, the stats you have collected in that city will stick with you forever (like : "funny guy +10", "robbing +2", "lying +7") and it will unlock new dialogues, and dungeons in other cities. 

[Muz]

I think the best solution to this so far is what Magna Mundi (EU3 mod) does. Magna Mundi detects your nation's 'personality'. By taking similar things, you'd get a bonus… e.g. a lot of good things happen if you have low freedom and have a highly centralized government. If you take any naval point, your navy is half price… take two or more naval points and it goes down to a quarter, in order to reflect naval culture.

The more you play things 'out of character', the more bad things happen. If you do anything, by default it will annoy someone. Strengthen merchants and your aristocracy get mad at the reduced power.

IMO one of the bigger aspects of the game were government skill sets, which works similarly to RPG skills. In a similar manner, you can let the player take any skills he wants. Some will probably give a synergy bonus.

The game was completely transparent about what affected what. Players were even encouraged to powergame the system. They knew the 'right' answer, they just had to decide if it was what they wanted.


It's hard to describe how to apply the system into a RPG in just a few words if you haven't played MM. I'd say instead of giving skills that improve game mechanic based stats, all the skills should denote character progression. Do away with skills like "Extra HP", "More criticals", "Sword and Shield".

Some better skills I could think up:
Charismatic - Very minor in game effects. May unlock a few good events. May unlock sex scenes. Also unlocks a few jealousy effects from your own party members and the general public.

Hardened spirit - Has to go through traumatizing events to unlock this. Gives maybe 20% extra HP, more damage resistance, but also lowers the chance of things like insanity.

Warrior of Faith - Gives blessed equipment, demon/undead slaying bonuses. Requires strict following of the faithful rules, or they'd get bad events like divine punishment. Also unlocks conflicts with agnostics and those of other faith, and will give you points towards 'religious intolerance' personality events.

Business sense - Bonuses for selling/buying. Unlocks some quests and conversations that allow you to start your own business in game, and it unlocks some options to settle conflicts via trade (or even do things like embargo a villain).

Lore master - Bonus to identifying mysterious potions and casting spells from books. But also lets you complete some major quest to save the world from plagues. Combined with the "Academic training" skill, it allows you to develop your own spells.

If they collect a set of certain skills, then they could unlock a specialized class skill. Or the player could choose to make someone with no special skills and make them a jack of all trades.

The core to this system would be to have events and bonuses based on the synergy of their traits. You don't give a "hard" statistical punishment for character/personality crises; you just make the player experience the effects of it. The player creates the character type they want and the game generates the experience from it.

[Shine Klevit]

You know, in the past year or so I played Arcanum, and I was thinking that the one thing I hated about it was the fact that it concentrated so many interesting things to do in one area to the point where the main mission became cumbersome, and the battle seemed entirely unnecessary. (It's still a damn awesome game)

If you ask me, what you're saying is right, and it's what the mainstream game industry is trying to do, but it fails in it's own demands to build bigger bigger bigger. You get games like Fallout 3 which are amazingly fun but get old so fast because they just beat you over the head with how much they can cram in.

I think personal character explosion could be mastered if they remove all the excess and focus on giving the player one area/situation but exploring every possibility of it. I mean, imagine if a game like Oblivion cut everything down to just one town. Focused on making a more logical 'law' system, and instead of putting 10000 faceless NPCs that have one or two problems you need to sort out, fill it with a few intriguing characters that have logical sensible interaction, a sensible economy, and sensible reaction from your decisions.

[Muz]

Uh, if you wanted a game that cut things down to detailing a situation, you should probably avoid all Bethesda games, because that's the exact opposite of their approach. Games like Fallout 1 and 2 did it right. Though it's more a result of how the Fallout 1 or 2 developers wanted the players to win the game no matter what characters they created. E.g. someone who created a high science/speech character could still win, and it created a lot of depth for people who didn't want to just play another hack & slasher.

[antoniodamala]

You know, in the past year or so I played Arcanum, and I was thinking that the one thing I hated about it was the fact that it concentrated so many interesting things to do in one area to the point where the main mission became cumbersome, and the battle seemed entirely unnecessary.

I think the problem here is not in the vast game world, but in how missions are imposed to players. Pretty much like on the Elder Scrolls, where it seems more free in general, but overwhelms you with side quests, items, dungeons, and then the game says: "this is not the main quest". Which is not just contradictional but stupid to say the least. While in the other direction of that concept lies Minecraft, that actually lets you do anything, and don't overwhelms you in the process (because you don't start with nothing).

I think personal character explosion could be mastered if they remove all the excess and focus on giving the player one area/situation but exploring every possibility of it. I mean, imagine if a game like Oblivion cut everything down to just one town. Focused on making a more logical 'law' system, and instead of putting 10000 faceless NPCs that have one or two problems you need to sort out, fill it with a few intriguing characters that have logical sensible interaction, a sensible economy, and sensible reaction from your decisions.

I see your point here, and agree completely with it. Quality over quantity all the way.

The player creates the character type they want and the game generates the experience from it.

How beautiful would it be if that actually happened in a story driven rpg?

[gimymblert]

First I want to point there is a distinction between "self insertion" (you are the character) and "role play" (acting as someone else), actually when role playing there is an never ending tension between the two (projection vs mimicry) which make it playful. This system is designed to actually support this tension with the "character crisis" mechanics.

However I did have trouble fully projecting myself in term of gameplay design and see how it all evolve, mostly what's the reward of "character crisis"?

However there is some good suggestion here
1. First having a system of "wants" like CA sinclair suggested based on affordance in the universe is a great feedback system.
2. Muz have talk about a "traits skills" system, I haven't play his game but its sound very similar to crusader kings 2 systems (which focus on character too). It simplify things because I was still stuck with a vision based on classic skills, which I thought didn't make sense. Therefore a system of "adjectives skills" that compliment the " attitude class" as unlocks, and which depends on how YOU play the class, is a great system totally tailored to the idea. You might be remorseless killer, or a vengeful killer based on your action and it shifts wants and affordance.
It also solved the character crisis problem because it became just a new progression towards specific skills, for example you were a killer and you have shift to peacemaker which would unlock the skill redemption, the effect maybe different social interaction available. It may also make a social crowd gameplay like "assassin's creed"'s one having more depth. The important things is that "attitude skills" does not just unlock new interaction like typical system, they also unlock new events aka the world is actually reacting to player's actions and their coherence. There is three anchor to make event trigger, in character progression, out of character progression and the metric difference between the two (is it positive, negative, big or small?)
3. It makes more clear what is the main gameplay, while your choice is mostly restrict by the class there is still plenty way to satisfy a particular wants, if you are a killer, there is many target that satisfy a kill. The implication is that the player must anticipate reaction to not corner itself into a dilemma (the correct option that reinforce character have undesirable consequence, not doing it increase out of character) which increase as both bar fills up. Therefore if a player want a particular destiny for its character, he must game the world strategically. He also must be careful to build the character he want to do, maybe he start as a killer but want to be an assassin (kill for money) therefore should avoid actions that divert from this nuance. Which mean that it reduce need for gauge like HP, MP and others, a state base system would be better (killed is not just hp = 0 but a distinct state).

It makes easier to make a story "event drive"n rather than "script driven", because "attitude actions" make sense on the narrative level, not just on the physical level, they have actual meaning, especially psychological and social meaning. Like MUZ say "in character" would bring good events, "out of character" would make bad things happen, mixed with the option lock and you have organic moral dilemma, character growth build into gameplay and player choice, mostly reactive story with increasing tension (attitude skill unlocks new class of event that happens).

I think the description of the system is complete enough to warrant a prototype

[moi]

I agrre, and I just want to add
the best way to make you're dum gam intresting: make it a zombie invasion.
put down hards of zombies with a uzi or ak47 : fun.
ride a horse with elfs: gayyyyyyyy

[Muz]

On further thought, why be limited to classical level/skill gain?

Instead, I'd say any "character developing experience" should give them a perk. If you kill a tribe of goblins, you get a "Goblin Killer" perk, giving you penalty to any communication with goblins, but bonuses to fighting multitudes of smaller creatures and goblins. Kill several small tribes - you don't gain any skills, because it's not a new experience (but it becomes easier because you know how to kill them).

The only way to get better at killing goblins is to try to unlock more extreme perks, like taking on major goblin hives and citadels, which will give you higher levels of "Goblin Slayer", with stacking bonuses and penalties.

Someone could still go from no "Goblin Slayer" experience to "Goblin Slayer IV" by sieging a grand goblin citadel, but their chances of survival will increase drastically if they go slowly. Also simulates the mechanics of "catching up" to skilled heroes.

At higher levels, goblin rights activists and goblin tribes will start teaming up against you, so this system applies equally well for linear and non-linear gameplay. It will also remove the problem of grinding by doing irrelevant things like killing bunnies. Don't have to worry about using exponential XP gain to balance against the decreasing challenge of weaker enemies.

You can still simulate grinding.. e.g. going through a melee training course might give an "Educated Warrior" kind of perk (+1 damage, +1 defense, +5 HP).

[Charlie Sheen]

if you want to bound a player to a role, use a class for that. of course, if you want to offer many distinct types of roles you *will* have to change a lot about the game.

but the bigger problem i see is this: what kind of a game a "pacifist" would play lol? how is that going to be ANYTHING interesting?

in any case, its easy to think of a class that will discourage killing. something like: "pacifists still live within good and evil and thus they feel bad when they kill other people. if they go too far, they fall into depression or something". thats from top of my head, but you get my point.

[Muz]

Pacifist class could work. I see it something like in the SWAT games, where they lose points for lethal methods. Or like a druid who uses tangle/charm spells.

[gimymblert]

I keep "levels" as a metaphor to keep a progress metric for the player, to keep him feeling is advancing in the game. One reason RPG lend themselves so well to open world is that the feeling of progression is not tied to geopraphique and topographic progression and has generic currency (exp).

However what is really a level we can actually change it, maybe a level is the number of current perk the player has "obtain" in total (not what is currently activated) or maybe the sum of current skills progression. The player would have a global feedback and may check in the menu details for more granular view.

The way the system might work is that each skills is responsible for checking "attitude": They manipulate wants by a weighting system (by prioritizing them according to current environment affordance) and they each check interaction for in/out character rewards.

We may implement an "attitude vision" which highlight which interaction are possible with degree of green for in character interaction, degree of red for out of character and blue for conflicting interaction.

Actually not displaying the in/out character progression may not be necessary, we may only need to show the "crisis balance metrics" which show the balance of in/out character action (with color intensity showing which metrics is about to level up) as it is the main concern of role playing character, aka how far from crisis I'am. It will let the player make moment to moment decision on how to grow his character (to or away from crisis).

ON PACIFIST:
Well it turn the game into harvest moon/animal crossing basically but with more progression feeling, however pacifist would be just a "traits".
Also we can lower emphasis on random combat encounter ...

OUT OF TOPIC RAMBLING:
It's fun to think that mario 64 has a xp system with the stars, once you have enough exp you literary "level up", you can also unlocks "skills" (can use new blocks to get new hats powers).

[baconman]

The covenant system in Dark Souls might be just what you're looking for.

The covenants themselves kind of resemble Dorks 'n' Dice alignments, and they each correspond with a gameplay mechanic that players are then rewarded for their focus on. The Sunlight Brotherhood for example, are all about co-op phantom summoning, the Forest Hunters are about PvP, and the Chaos Servants are all about acquiring Humanity items. As you complete more goalposts in your covenant's mechanic of choice, you're given a progressive reward system.

So, design a system of naturally conflictive gameplay mechanics, and after an "introduction to the world" phase, allow the player to align with a faction that encourages the type of gameplay - and character development - that the player is inclined to pursue. Treasure Hunting? Boss Bashing? Strategic Defense? Racing, even? You'll think of some things that'll tie together.

EDIT: In fact, just think of the different types of games people play; and how the flow of those games can tie into yours.

Then it's just a matter of encouraging that compulsion/feedback loop within reason, and perhaps incurring some kind of limited/reasonable penalty for switching gears. Not so severe as to discourage ever doing it, but enough so that players aren't changing sides faster than an in-play d20.

[Charlie Sheen]

I always wondered why people stick to obsolete "RPG" conventions. Part of the reason I often ask myself this question is that I never really liked RPG's to begin with (apart from SRPG's, but the "RPG" conventions have little to do with that subgenre anyways (and I never played tabletop RPG's..)).

I always thought that the next logical step for computer RPG's is to become proper simulation games. Basically, a game where you have many roles to choose from and thus many actions to support all these diverse roles. You'd also have a decently huge world to explore and you could have actions that could let you do the regular kind of stuff such as move around, interact with objects, steal, kill, talk to people, etc and perhaps even take it further by, say, complexifying conversations (thus allowing for mental violence of some sorts eh?). It would be also cool if you gave up on this whole level-progression shit and made it properly open-ended so that everything in the game felt alive, so that everything in the game moved, so that the game wasn't waiting for you to play it correctly before it moved forward. Of course, this means a lot of complexity, but you don't even have to bother with visuals because I'd play it even if it's text-based lol (and in that case, it would also have to be turn-based, since I can't think of an action text-based game lol text, after all, is slower to process for a human mind than real-time imagery).

And I think RPG's aren't the first (or the only) games to have non-geographical navigation. What about Civilization? What about technology trees? Isn't tech-tree, after all, a technology progression, a sort of progression that is non-physical? And besides that, in most RPG's, character progression is so retarded that I'd like to distance myself from the trope as much as I can lol.

[Muz]

People stick to it because it's easier than thinking new conventions. I've been doing trying to do RPGs for 10 years, even did a degree in simulation, and this thread is probably the closest to figuring out a better way to do it. Most designs are iterated off an existing good one and RPGs have way too many components to make significant changes.

Simulations are quite literally one of the hardest things to create. Many of the early kings of world simulation - Will Wright, Chris Crawford, Tarn Adams - have a very technical background and it shows.

Trying to simulate plot/character development is even harder than world simulation. You'd probably need a solid background in linguistics, psychology, or writing to get there.

Having gone into the hard, deep end of simulation, I'd say that you still have to take a lot of steps back and abstract things out. Because there's a hell lot of things that people just don't want to experience in a simulation. There's no "right" way to do it yet.

JRPGs abstract out most non-plot stuff, eventually branching off into visual novels. WRPGs are often too focused on the gritty action details and abstract everything else out… and branch off into tactical RPGs like Wesnoth, D&D4, Age of Wonders, Warcraft. Bethesda games focus on world development, and abstract everything else out, including the player characters.

Personally, I'd like to see a RPG system focused on character development - trials, personal vendettas, romances, conflicts, etc. You'd need to abstract out the things that don't involve character development - "random encounters", minor combat, tactical details, maybe even the world itself.

It will likely involve permadeath. With save/load, there is no real fear of death or failure, which is a vital component to character development. But death itself would be part of the development. Dwarf Fortress does a great job of this - a character last moments is simulated poetically, lying bleeding against a corner with a broken leg… the last memory being the minotaur's horns cracking through the ribs and puncturing the heart. Or a dragon's breath melting the fat off the character's face.

Character development mechanics need to "zoom into" these aspects. Getting married shouldn't just increment the "wives" counter. It should reflect the impact on their adventuring lives. Does the wife nag him when he goes adventuring? Does she provide battle support? Do they work as a team? Do they argue when splitting the loot? Will she go into a manical rage when a nymph casts charm on her husband? If the wife is higher level and 'invincible', would the protagonist be the subject of assassinations and kidnapping plots?

These kind of mechanics will shift the focus from immersion-breaking strategies like "which wife has the highest DPS" into more roleplaying thoughts like "what happens if my guy marries that girl?" Of course, this kind of focus will lose detail on other things - note that I didn't mention the setting of the world, or the combat/economic system, or even if this scenario has a plot at all.

[s0]

pretty sure crpgs are older than civ (both the videogame AND the boardgame it was inspired by which is the first game with a tech tree afaik).

It would be also cool if you gave up on this whole level-progression shit and made it properly open-ended so that everything in the game felt alive, so that everything in the game moved, so that the game wasn't waiting for you to play it correctly before it moved forward.

mount&blade has a dynamic world but that's about the only one i can think of right now. its always a shame to see rpg developers pull punches on sim mechanics though (see oblivion and the "radiant ai" debacle).

[thunderhead.hierophant]

I made a post in the "procedurally generated stories" thread in the writing forum that might be relevant to this topic.

I'd cut & paste, but I'm surfing with a PS3 and I'm not sure how.

[kiddRaddical]

Is there a simple way to change the mechanics while not necessarily create a new genre or even a new format or even a complete new system?

I do believe that this is a contradiction on multiple planes, good sir.

1) If it were simple/easy, I think somebody would've already done it.
2) Changing something so essential to RPGs as they exist would constitute a new genre in my book, but according to games' eff'd sense of genre labels, meh...

IMHO, people resort to fighting for skill because it's mentally lazy to implement (and I say that with the utmost respect). Games need conflict, and the simplest form of conflict to comprehend is violence. As developers, we tend to spend more time focused on implementation then conceptualization, so we never get down to the heart of "conflict" (or his sister, "relationships") as concept(s), and so never implement them on a more abstract basis.

The aforementioned argument also applies to progression, imo. Again, progressive fighting ability is easy to conceptualize:
1) I start off not knowing how to fight, so I get my butt kicked.
2) I fight more, so I learn new "techniques", and I get a little better.
3) Find an item that enhances my ability to fight, so I get better.
4) Repeat 2-4 until you are "The Guy".

This is where most designers stop, I think. Can't you just imagine what the code will look like? 

IRL, progression sucks. A LOT. Think about your long journey to become an ADEQUATE programmer, let alone the keyboard ninja most of you guys are. Think about your experience through school; learning how to socially interact with other people (first date = biggest xp gain of your life  ).
Leveling-up is hard IRL, and requires a lot of sometimes painful growth, which no one has portrayed accurately in a game yet. "Nobody wants to play that"  .