How to make a numberless (but stat-heavy) system WORK

[obediah]

ENCUMBRANCE is a QUANTIFIABLE value, just like COST is.

EVERYTHING is QUANTIFIABLE.

Care to back that up? To save you time you don't have to quantify everything - just tell me how to quantify "value to IBM in position X" for any person on the planet, and I'll be too busu counting my billions to care whether or not you ever finish.

 

[Human Shield]

Turn on auto-leveling in Arcanum and never open the character screen.

All this tells me you still have no clue what we're talking about, yet you persist to complain Saint. ENCUMBRANCE is a QUANTIFIABLE value, just like COST is. All this time we've been discussing removing the numbers that have been artificially quantified. The GOAL is to distance the player from powergaming, and draw him into play a character role rather than a spreadsheet. OF COURSE there's going to be NUMBERS in the game for crying out loud! Where in the world did you come to the conclusion that we want to do away with NUMBERS for the sake of them being numbers?

Why not have measurements for the width of your arms and legs. Or body-fat %. Have clocked running time for a dash.

Why don't we simplify those numbers to a 1-100 scale?

 

[Rat Keeng]

I thought the same as Saint, that you'd do away with things like weight and encumbrance figures, just like to-hit numbers and all that. I mean, you can't tell exactly what an item weighs, how would you know that heavy battle axe rings in at exactly 27 pounds? See, i get the idea of eliminating powergaming and numbercrunching, but surely there's as much munchkinism involved in picking up the exact amount of items you can carry, measure everything up in weight vs. value, just so you can squeeze that extra piece of gold out of your current haul.

I actually like the idea of having a backpack that gets more and more "packed" as it fills up, and watching your character slow down and get penalties to appropriate stats (not shown of course ). It'd also be cool to be able to quickly drop your backpack on the ground before combat, or toss it over a fence if the extra weight wont allow you to climb it, but still retain access to items in your "quick pockets", or attached to belt etc.

Of all the numbers currently in RPGs, weight and encumbrance is surely the one i'd rather do without, as i've never been into item-hoarding and all that foolery.

 

[EvoG]

I thought the same as Saint, that you'd do away with things like weight and encumbrance figures, just like to-hit numbers and all that. I mean, you can't tell exactly what an item weighs, how would you know that heavy battle axe rings in at exactly 27 pounds? See, i get the idea of eliminating powergaming and numbercrunching, but surely there's as much munchkinism involved in picking up the exact amount of items you can carry, measure everything up in weight vs. value, just so you can squeeze that extra piece of gold out of your current haul.

Well no, you didn't think the same as Saint. See, he thinks that we WANT to do away with ALL numbers period...and he thinks thats stupid. You appear to like the idea of having no numbers period. :D

I actually like the idea of having a backpack that gets more and more "packed" as it fills up, and watching your character slow down and get penalties to appropriate stats (not shown of course ). It'd also be cool to be able to quickly drop your backpack on the ground before combat, or toss it over a fence if the extra weight wont allow you to climb it, but still retain access to items in your "quick pockets", or attached to belt etc.

Me too...and I loved the inventory model in Ultima 7(I think I'm in the minority no? ). As for dropping the backpack, thats another level of realism that I'd actually REALLY like, but you and I might be the "minority" on THAT, hehe. Anyway, would you believe I was(sorta am still ) toying with the idea that the backpack would be entirely physics based, with the items being physicalized "spheres" for each volume, so you could literally dump your items into the backpack and watch it fill up. The pack itself would be transparent depending on the facing(like how dungeon siege removes occluding objects), and you could grab and shift the items in the pack to take them out. Neat to play with, but would everyone like it?

Of all the numbers currently in RPGs, weight and encumbrance is surely the one i'd rather do without, as i've never been into item-hoarding and all that foolery.

I can agree with this. For what its worth, the basis of my argument with Saint is that really, things that are reasonably quantifiable like weight and cost would remain, as it wasn't "numbers" that we're trying to do away with, but numbers for things not easily or normally quantifiable. Of course there are those that argue this, but I think the point is clear seeing as how many of you understand what we're talking about here. I think they want to argue to argue.

 

[Section8]

Given that Encumbrance in the context of a game is rarely a variable that is actively tracked by a player, and that it only becomes an issue once the penalties kick in, complete with wacky sound effect if it's Arcanum, then it illustrates a good example of a hidden stat. The player doesn't need to specifically know how much they're carrying, they just need to know what is too much.

From there, an exact measure of what has the best selling price to weight ratio encourages power gaming. If the player can make reasonable estimates, then exact quantification is unnecessary.

As far as "bag packing" goes, I hated Ultima 7, because I'd played Ultima 8. They're both essentially the same, but Ultima 8 actually remembered the (x,y) coords of the items in your pack, so you could arrange it and not just have a jumbled mass of stacked items. You might argue that a backpack is realistically liekly to end up as a jumble like that, but an actual backpack is also likely to have pockets and other ways to effectively order it, so I think it's a reasonable abstraction.

But really, inventory management is a whole new discussion. Is it weight based, space based, both, etc.

 

[RGE]

But yea, the biggest problem I have with powergaming is that games that promote replayability and custom character creation seem to always boil down to "whats the most balanced character I can create and grow to see and do everything in the game". I'm guilty of this. If you have a completely combat oriented character, and then come across a door that is locked and very difficult to pick, and your lockpick skill is very low, knowing full well there is some prime loot on the other side, you'll just kill the hell out of everything in sight (because you're good at it in the first place!) and farm points to dump entirely into lockpick so you can gank the stuff. This sucks and is not roleplaying by any stretch. Yes it IS in roleplaying GAMES, but that doesn't make for truly playing a character.

I tend to do this too, but it's mostly because there is a finite amount of things to do in a game, and if the game is fun enough to play, it's almost certainly fun enough that I want to do more than there is to do. If there were more randomly generated content I'd probably focus on the things that suit my character rather than trying to build a character that suits the game. Also, replayability isn't very high in games where most things are already custom designed rather than randomized, because that would mean that the story is also set to a large degree, so I would usually not want to replay the game just to explore the story from a slightly different angle. Half the fun in an RPG, be it computer or PnP is the feeling of possibilities, and that feeling is seriously hurt by already knowing the story in advance. But this is where a CRPG without obvious stats would help, because with no visible stats it ought to increase the feeling that certain things are possible, even though the game might actually be ignoring those things. As long as the player doesn't find out, there's always a chance that there really is a secret cow level.

This is also a problem with the model of constant character growth. Growth perhaps should appear in spurts rather than linearly. An "ah ha" moment with ability improved due to a significant occurance. This follows well with our concept as players are then less focused on the next kill. Their "ah ha" moments will come when appropriate, so there's no need to powergame yourself to your next level.

Wasteland did it that way. When a task was difficult enough characters would often fail but also have a chance to raise their skill, and thus high skills came from difficult challenges. I like that method of increasing skills, even though many skills had very few usage opportunities.

For encumberance I'd like a system where the more you carry the slower you'll walk, until you stumble around more than you actually walk. How much you can carry would then become an issue of how slow you'd be wiling to go. That also seems like the perfect balancing for a platemail - it weighs a lot, so you'd walk and run faster with lighter armour, especially if you intend to carry loot.

 

[EvoG]

I tend to do this too, but it's mostly because there is a finite amount of things to do in a game, and if the game is fun enough to play, it's almost certainly fun enough that I want to do more than there is to do. If there were more randomly generated content I'd probably focus on the things that suit my character rather than trying to build a character that suits the game. Also, replayability isn't very high in games where most things are already custom designed rather than randomized, because that would mean that the story is also set to a large degree, so I would usually not want to replay the game just to explore the story from a slightly different angle. Half the fun in an RPG, be it computer or PnP is the feeling of possibilities, and that feeling is seriously hurt by already knowing the story in advance. But this is where a CRPG without obvious stats would help, because with no visible stats it ought to increase the feeling that certain things are possible, even though the game might actually be ignoring those things. As long as the player doesn't find out, there's always a chance that there really is a secret cow level.

I agree with all of this. Thats one problem I have with a most games...once I've squeezed dry the story, I do not want to play it again, even as a different character. Getting funny dialogue or picking all locks in the game did not amount to a full replay-experience for me. But the numberless system, you're right, would at least give the impression anything is possible within reason.

Wasteland did it that way. When a task was difficult enough characters would often fail but also have a chance to raise their skill, and thus high skills came from difficult challenges. I like that method of increasing skills, even though many skills had very few usage opportunities.

Funny you mentioned Wasteland! One of my favorite memories was losing a guy in battle, and using my female medic to try and raise him from citical to severe. He was so fucked up, and her doc/med (cant remember which or how that went) was not up to task, but had a "chance". It honestly must've taken 15 or 20 minutes of failing before I "got it!" and it was the greatest thing ever, as I didn't have to reload.

For encumberance I'd like a system where the more you carry the slower you'll walk, until you stumble around more than you actually walk. How much you can carry would then become an issue of how slow you'd be wiling to go. That also seems like the perfect balancing for a platemail - it weighs a lot, so you'd walk and run faster with lighter armour, especially if you intend to carry loot.

While I agree with this fundamentally, people are so used to carrying everything up to your max inventory, that they'd just complain that only half the grid spaces were worth anything because they didn't want to sacrifice their movement for 'loot'. On could abstract that the inventory you're given is what you could carry at its max and still maintain full movement, so yea, I dunno. :D


Cheers

 

[Mulciber]

New thought for a combat system:
Perhaps the physics geeks in the crowd could give a better answer to this, but why not implement a physics-based damage system for weapons? As has been said earlier, cmoputers are made for the number crunching, so the only limitation is on how involved you want the design to be.
I probably would design the system so that the damage done was measured in Newtons (or Joules, for heat damage). A weapon would have a wieght and an impact area (think blunt vs. slashing vs. piercing). The force delivered would be mass * velocity ( F=M*V). Velocity would be calculated based on how quickly the character could swing a weapon of weight X.

Impact areas determine how much damage can be done to the character. The larger the impact area, the greater the damage. This is also a method of modelling location-specific damage.
Armour could be described in terms of force dissapation and breach resistance. Stiff armor would be much better at dissapating force, but probably more prone to being breached by weapons that have a very small point-of-impact.
Character skills would influence many of the equations- reflexes for reaction speed for attacking and blocking, strength determining how high a velocity you can move your weapon, dexterity for reaction speed, anatomical knowledge for particularly interesting critical strikes ( any bloke can hit you in the head, but it takes a real whiz to know what angle you use to shiv a man through his ear-hole).

The final note about this is that if the designer has no intention of making this visible to the player (except, perhaps, as a matter of interest to the physics geeks) it can be as bewilderingly realistic as you care to make it. An ideal candidate for a numberless, yet stat-heavy system.

 

[Imbecile]

Has the potential for comedy as you pick up an object, then as your fatigue affects your strength stat, get crushed.

I like the idea, but rather like economies is this kind of thing better faked, rather than done "properly"?

 

[Section8]

I'm all for that sort of thing, the trick is, it's a lot easier for designers to crunch simpler systems in their head from a balance perspective. But hey, all you need is a bit of computational assistance on that front too.

 

[Balor]

Well, in fact, that would be a BEST of implementing it. I've thought about it too, btw.
Only 'numerical' values you'll see are weight (well, it's quite numerical in RL too, cause you can measure it accurately), and length.
Oh, btw, if you want a game that has is very good in implementing such a system (well, relatively) - try MUD Geas.
http://geas.de/

 

[Mulciber]

Well, with a system like this, it would elegantly solve all of the issues that people have with the arbitrary assignation of damage values. If you included the idea of vectors, you could also model deflections (soft parrys /redirections as well as shield deflections). It could also make the crafting skills in a game very interesting- no two swords would need be identical. A higher quality blade might be lighter, without sacrificing structural integrity (less likely to bend or shatter on that guy's armor, for example).

 

[Saint_Proverbius]

But yea, the biggest problem I have with powergaming is that games that promote replayability and custom character creation seem to always boil down to "whats the most balanced character I can create and grow to see and do everything in the game". I'm guilty of this. If you have a completely combat oriented character, and then come across a door that is locked and very difficult to pick, and your lockpick skill is very low, knowing full well there is some prime loot on the other side, you'll just kill the hell out of everything in sight (because you're good at it in the first place!) and farm points to dump entirely into lockpick so you can gank the stuff. This sucks and is not roleplaying by any stretch. Yes it IS in roleplaying GAMES, but that doesn't make for truly playing a character

You want to know a better way to promote replayability and not do something idiotic like hiding information from a player? How about give them MORE information so that choices aren't so easy. Make the models of things like weapons nice and complex to where most players won't even bother trying to powergame. Instead, they'll go with matches for what they already have or think they have.

Say for weapons, you have not only a damage range, but also speed factor for how fast a swing takes, a recover factor for how long it takes to prepare another swing, the weight of the weapon which affects damage bonuses and ability to block the weapon, and so forth. Furthermore, have these statistics based on varying attribute modifiers depending on the type of weapon. A dagger could have a recover factor based on dexterity to modify it whereas a claymore would rely on strength and a heavy flail might rely on constitution.

That way a character looks at his character sheet and says, "Hey, my character is pretty good at speed and agility, so what weapons can I use that accent his build type?" and "Do I use this rapier which gets a damage bonus from my dexterity or so I use the dagger which gets a speed modifier from my dexterity?"

Nothing is hidden, it's still complex, and the player is allowed to make intelligent choices.

ps. That ground texture looks like shit.

 

[Zomg]

How about a CAD-like subprogram in the game that lets you animate your own sword strikes, and then procedurally determines the force applied based on forced spring models of muscles and Newtonian physics? This could be combined with the weapon modeling, and you could also realistically model the effects of armor very finely (for example, wearing heavy bracers would actually increase the momentum of most swinging motions, controllability aside).

"I use backhand swings because biceps are my dump stat in my arms".

That game would probably have an audience of about twenty guys. But we'd be really, really supportive on the forums.

 

[Nutcracker]

Well, this is one of the key choices open to RPG developers - do you hire Nobel Prize winning physicists and 3 extra statistics men...or do you just go for Patrick Stewart and a couple of new l33t graphical artists?

 

[Balor]

Well, in a perfect world, we'd have both.
But since we don't seem to live in one... *sigh*

 

[Twinfalls]

What? That's not what's being discussed. Zomg is illustrating how removing levels of abstraction can make a game less interesting to the player. There's no need for nobel-prize winning anybody to make a good RPG, the choice you propose, Nutcracker, is a complete figment.

Integrity and common sense is what's needed.

 

[DarkSign]

Damn. I only just now saw this ...so Im going to have to wait until I get home, really read it...dissect it...then reply.

 

[Balor]

2 Zomg
Heh, I sense a kindred soul .
Try GEAS - it's indeed interesting, too bad I cannot play it for now due to internet probs and general lack of time.

 

[kingcomrade]

You're right Saint, I've always wanted to play a game so complex that I just have to strike out randomly in the darkness and hope I make a good decision. Balance is the way to resolve that problem, not by...what's the term for deliberate artificial restraints in games? I know there is one, I can't remember it.

 

[DarkSign]

Ok. I cant contain myself. At the risk of bringing up something already covered....

My problem with statless games is the user feedback. Either the feedback you get is crap...or it becomes the same as stats.

There's a post on the net called "So-and-so's incredibly long rant on MMO design" or something like that. It goes into why no numbers = l33t.

Let's take good old lockpicking.

Say the player is a flat novice and he comes up against the Electro-lock 3000.

1) The player can either be told that its a level 3000 lock and realize he shouldnt attempt it

2) The player can not find out that he's too low and struggle to open it until he frustratedly gives up

Neither of those are appealing to me. But lets go further.

When I said "be told that its beyond his level" there are multiple ways to tell him

a) some NPC tells gives him description information

b) the lock graphic tells him something before he attempts it (a sign that says "uber lock 3000")

Then there's the whole problem of using frustration as the sole indicator. Im playing a game to be entertained...definitely to be challenged...but not constantly frustrated. Numbers save me time.

Im the kind of player who, if Im failing at something, will think there was some step 2000 turns ago that I fugged up or some glitch in the software...and keep trying all 223432432 combinations of ways to do things. If I have to do that because I dont know that Im 2 levels too low for a lock, Im gonna quit the game.

Then there's the whole question of needing feedback for learning. I mean it might be a bit crude but after Ive swung my ax 10000 times and the liittle "Ax is now at 89% pops up, I feel like Ive been rewarded and am doing something right in molding my character.

We all want something new and intuitive, but lack of numbers wont be it imo until we get a lot better interactivity/physics/input hardware.

 

[Mulciber]

My problem with statless games is the user feedback. Either the feedback you get is crap...or it becomes the same as stats.

There's a post on the net called "So-and-so's incredibly long rant on MMO design" or something like that. It goes into why no numbers = l33t.

I think that most people in the thread would agree with you that user feedback is really the issue. Mostly what we've been discussing here is how to improve that feedback to the level where explicit stats become unnecessary. My stance is that it can be done, it just hasn't been done yet and that the reason it hasn't is that developers are at the point where they are locked into a legacy system when the technical issues that dictated that system have long since been solved.
It would be great if you could come up with some novel suggestions of your own. My personal favorite is the idea of keeping your character's representation within the game world- Quantification for things that can be quantified (weight, height, cost, nothing incongrous about that), personal opinion about the chances of success (perhaps along the idea of self-perception vs. actual skill), but most importantly, the chance to try (and possibly fail) at everything.

 

[DarkSign]

Ok. Thanks.

What are your thoughts on frustration, though? How much of feedback should be based on become pissed off that you've just failed too many times and decide to move on?

Yes, Im looking for a % :D - although it would really have to be play-tested to get it right.

With today's technology there are certain things you can do..others you cant. For example lets take picking up something heavy.

In real life you get two pieces of feedback..the pain you endure while picking something up and the speed with which you pick it up. Speed would be easy to code...and depending on the differences in speed you would be able to tell whether something was heavy or not. Pain wouldnt be that hard either...just would require watching your health meter any time you picked something up. In fact, critical failures could result in ripped muscles, torn ligaments and so on. (To really do it "realistically" youd have to simulate the speed with which you try to pick something up - e.g. "I tried to pick up the rock quickly and tore something" but that would either require a mini-game (tap the button to pick it up faster) or not be possible because of ergonomics)

I'll try to think up more as I sit here at work

 

[Mulciber]

Well, I've always liked the idea of some visual feedback of the character's frustration; An irritated punch to the door that you're trying to unlock, for example. But that's not what you're talking about, I think. Likely what you mean is the frustration that comes from consistently failing in a pass/fail system compunded with the insufficent feedback problem, right?
If frustration is caused by lack of progess, what about a system that allows progress when you marginally pass/fail a skill check? For example, if I was to try and pick a lock, I would apply the skill. Based on my skill and luck, i would make varying degrees of progress towards opening the lock. Complete success would open it, a marginal success (and you could have varying degrees of this) might represent my character exploring the lock's mechanism, not opening it, but making it more likely that I can open it on a successive try. Marginal and critical failures could be done in a similar fashion. The feedback could be given in a self-evaluation style where the character reports how they think they have done ("I think I made some progress here"), self-narrative seems the most straight-forward. I would distinguish between how the character percieves that they did and what their actual success rate was. Makes it more interesting to play characters who are arrogant or self-doubting.
What do you think?

 

[Claw]

I believe Fallout has special feedback for tasks far beyond the player's task.
It shouldn't be a problem to implement a self-assessment by the PC in this place.

I have a different take on Mulciber's marginal success concept.
I fancy autolearning as used by TES, but not the idea of learning by success. Instead, skills should progress depending on the margin of success or failure.
So repeatedly coming close to success in a task may result in significant skill progress until the PC succeeds.

The idea that the character's self-assessment depends on his personality is very interesting.