ID:104498
 
Keywords: design
I been playing with a hand-me-down iPad lately and, perhaps it's just the Apple creativity rubbing off on me, but I'm feeling somewhat productive again.

The 8K Cartridge Classic mostly has me thinking how cool it will be when I'm making a game not so restricted. Perhaps that's one of the subtler goals of the contest, to make you realize how cool it is you can do even bigger things, on top of the more obvious goals of getting people to familiarize themselves with libraries and code efficiently.

Another thing I've been thinking about lately are the fundamental elements of game design. It's one thing to say, "I'm going to make a [genre] game" but it's quite another to look at the games of those genre and wonder to yourself why they work and more importantly can they be done better.

A major example that I've worked through quite a bit with the idea of roleplaying game statistics. We've largely inherited the idea of hitpoints and core attribute trackers such as strength and dexterity from Dungeons and Dragons. I've been thinking of ways to get rid of those statistics entirely.

After all, roleplaying games that are solely about accumulation (e.g. making stat points go higher) tend to get boring the moment the playe realizes making numbers go up without any real context is pointless. The greater thing that a roleplaying game needs to be able to do is tell a story, and each statistic the player keeps track of should be elements of that story.

What I'm thinking is that the character will accumulate a series of states that determine some basic, constantly in flux physical attributes. Mental attributes (such as "intelligence") are replaced wholly with skills, and the only physical attribute influence on that comes down to a character's sanity and/or focus. The physical attributes are multipliers on the effectiveness of the character's application of their skills in all the skill tests in the game.

The concept of an "attack" is rethought. There's no hitpoints, instead there's just a basic overall health measure of the character which is adjusted by the various states currently affecting the character. Characters die from being of too poor of health due to having endured too many negative states to overcome. This would only be one kind of attack, the kind that depletes health, other negative states may go after a character's coordination, focus/sanity, or strength.

This is just one way we could rebuild the concept of a role-playing game from the ground up, and is interesting food for thought.

Unfortunately, the level of sophistication involved in this system is a bit much for the Cartridge Challenge II, as developing all these states that the characters can be inflicted with would undoubtedly eat up a ton of code space. So I'll just do something simple for that, maybe a Xybot-like game will work if I restrict the projectiles to 90 degree angles.

I wonder if I can bridge the gap between a Roguelike, MUD, and space-empire development game like X3? I'd probably completely annihilate BYOND's obj/mob limit in short order trying, but it'd be a fun prototype.
Characters die from being of too poor of health due to having endured too many negative states to overcome.

So much work to remove the numbers, but how do you represent this system without numbers internally? In making an RPG you just have to design the most tempting skinner box--think EVIL.
You know, like WoW.
I suppose it's possible I could represent the system without numbers if I were to have the game be 100% state-based and have no derived statistics involved. Basically, it would be like a card game, with various cards in play representing the states.

However, I'm not really looking to remove the numbers internally. When it comes down to actually performing a skill-check, quantification is necessary. I need to measure the overall capability of the character compared it to the overall difficulty of the task presented.

What's more, I'd like the player to have some sense of what's possible and what's not, so I don't want to hide the numbers, per se. If anything, this state-based system may make the character seem considerably more lifelike.

It's a bit of an antithesis to the skinner's box approach in that it's not about variable rewards, it's about compelling story generation.
Instead of increasing the player's stats perhaps you could change the attributes of their abilities? I'll use the fireball spell as an example.

Lv 1: Shoots a generic fireball to where the target is standing.
lv 2: Shoots a homing fireball at the target.
lv 3: Homing fireball explodes doing area of effect damage when hitting target.
lv 4: Homing AoE fireball also catches the target on fire and does damage over time.
I know what you're on about, and the idea you have is nice, but you're going to have a metagame if you involve numbers at all.


I think the reason this hasn't been done yet is that it would take several orders of magnitude more effort make it work. Numbers make sense to people; but how are you going to make a system with dozens if not hundreds of states easy to grasp? How are you going to balance it?
Somewhat reminds me of what Lord Raven has got going with Lost Realms. Worth checking out if you haven't done so already.



Don't be scared to do something complicated with the Cartridge Classic II. Just find ways to do complicated things with small amounts of code (RECURSION. LOTS OF IT. BE WARY OF THE TINY CALL STACK). Also, try to pick one or two complicated things for the Cartridge Classic II and focus on them, rather than going with a whole myriad of interdependent complex systems.

I've given up on Litamora qualifying for the Cartridge Classic II, because that game is has numerous interdependent and admittedly overly-complex systems. Even with all of the object definitions offloaded to external files (over 70kb of offloaded data atm), the code is over 40kb. No way I'm ever going to be able to crunch that down to 8kb.


I have started on a new entry for the contest though.
It involves this.
Toadfish wrote:
I think the reason this hasn't been done yet is that it would take several orders of magnitude more effort make it work. Numbers make sense to people; but how are you going to make a system with dozens if not hundreds of states easy to grasp? How are you going to balance it?

It's trickier than a simpler system, no doubt about it, but I wager it'd be not all that trickier than a system that incorporates a plentiful number of buffs/debuffs such as the typical MMORPG. The main difference would be that these states reflect a whole lot more potent of an adjustment. One of the goals would be to avoid having states stack up until they lose all meaning.

D4RK3 54B3R wrote:
I have started on a new entry for the contest though.
It involves this.

That is a pretty impressive bit of terrain to generate for a game the operates within an 8k limit!
Status effects in MMORPGs are easy to predict because they affect numbers.
Toadfish wrote:
Status effects in MMORPGs are easy to predict because they affect numbers.

If you can come up with something programmed on these over-glorified adding machines we call computers that cannot be found to be affecting some number somewhere, that would be quite a feat. ;)

That said, it's not so much that I'm trying to avoid affecting numbers so much as I am thinking there's probably more novel ways to do RPGs than hitpoints and attribute scores.
It's true that everything eventually comes down to numbers, but there is a difference between treating something as a number and treating it as a "state". Somewhere along the line, the computer thinks of all strings as a bunch of numbers. But the developer who is using strings to say, output system messages, doesn't think of them like that. The same logic applies to states, which are not directly treated like numbers by the developer, and "regular" statistics, which are treated as such. Wasn't this sort of the point of your post? That character sheets can be represented to the player as more than homework in maths (to clarify: I agree with this, but I'm not sure the system you described is necessarily the solution)?

there's probably more novel ways to do RPGs than hitpoints and attribute scores.

It certainly is novel! But can you make it work?

I feel you came up with something in the right direction, but also that this system needs more refinement (or simplification, maybe) to work. Or maybe a model that is easier to understand and balance. In the end, the purpose of a character sheet is to give players an accurate representation of their character. I think the developers of the first roleplaying games initially thought of something like your system, but had a difficulty simplifying it enough that keeping track of states won't be a chore for the players. Numbers are a nice and abstract way to avoid complexities like this.
I'm one of the greater naysayers of my own ideas, but oddly enough, it doesn't seem so very impossible for me.

I think it's because I'm regarding states in the same light as I do content. I could add as many or as few states as I feel the game needs. In the end, they're just elements that may come into play.

In a way, states are monsters. The player encounters them in the game as obstacles to overcome. You could have a gigantic menagerie of monsters/states, and if you decide to give them unique behaviors, then that's just a bunch of attention to detail that you've opted to code.
Well, the burden of proof is on you. I hope to be playing that Rogue-MUD-Empire prototype! :)