An RPG with an API

[Asmor]

One of the things I really liked about 3rd edition was the variety in the classes. Particularly towards the end of its lifecycle, things like The Book of Nine Swords and the Factotum stick out in my mind as being wildly different in playstyle from other things out there.

The d20 SRD lead to a lot of innovation, but when you came down to it, everything still had some pretty similar assumptions...

I'd like to see a system that takes it a step farther, and implements an "API." For those not familiar with APIs, it's a concept in programming where basically I make a program, but then I publish documentation which tells you how you can make your own programs interact with mine, e.g. if you issue command x I'll give you response y.

So I picture a system with a set of required "interfaces" and guidelines.

Here's an example of a very simple interface:

*All characters can make an Attack, and every Attack must be opposed by a single specified defense (reflex, fortitude, and will). If the value of the Attack exceeds the value of the defender's defense, the Attack hits and deals damage.
*All characters have reflex, fortitude, and will defenses.
*All characters can receive damage.
*All characters have a level attribute, which is a rough estimation of their combat prowess.
*All characters must be able to receive a "disabled" condition, which reduces their damage dealt by half, and doubles damage they take.

And some guidelines to go along with those:

*Defenses should be approximately equal to the character's level.
*A character who takes damage equal to approximately 10 times their level is defeated.
*A character who is capable of taking more damage before being defeated should have lower than average defenses, and vice versa.
*A character's attack value should on average be equal to its level.
*If a character hits, its attack should deal approximately 3 damage per level.
*A character with a lower chance to hit may deal more damage, and vice versa.
*Additionally, characters should have attacks which are limited either expressly (e.g. certain number of uses) or by difficulty of pulling off. These limited attacks may deal increased damage and/or have better chances to hit, proportional to how limited they are in actual play.

Ok, so nothing terrible groundbreaking yet... but now that we've got this API defined, we can start letting our imaginations run wild.

I'm running a game using this system. My players are Bob, Carol, and Dan.

Bob is playing a fairly traditional character, a fighter class that came with the main "book". He has a character sheet with numbers on it and rolls a d20 for his attacks and 2d6 for his damage.

Carol is playing a mage, but rather than the spell-point based mage that came with the book, she's using a D&D-inspired Vancian "Fire-and-forget" wizard published by a third party. She's got a selection of spells she can use, such as "Spirit Shackles", which she can use once per day, does 5 damage, and disables the target until Carol's next turn.

Dan is playing a huckster, a class he found on a message board that he thought sounded really fun. Thematically, hucksters are mages who use gambling implements like decks of cards to channel chaos magic. Dan doesn't even have a character sheet-- he has a deck of cards. Whenever he takes damage, he discards cards from the top of his deck, and when it runs out, he's defeated. He also maintains a hand and draws cards every turn. His basic attack requires him to discard a pair and does damage equal to the value of the pair (10 for face cards). He can also discard a full house for a devastating chaos attack that deals 20 damage.

It's the climactic game of the campaign, and they've finally tracked down the BBEG, the brilliant, cunning, and unfortunately-named Dr. Puppetmaster. I set a Go board down on the table and place a couple dozen black stones on it. "He gets to play one on each of his actions," I announce. "Whenever you deal damage you get to place that many white stones on the board. When you capture the last black stone, you win!"

 

[joelesko]

Based on your example, it sounds a lot like Mousemate. =)

[ame=http://www.youtube.com/watch?v=ycM844Bfzsk]YouTube - Mousemate[/ame]

_______________

Tabletop Quest - 100% Web-Based VTT

 

[Nagol]

This can be done with a bit (read a lot) of prep by a GM using the 'toolbox' style games like Champions, GURPS, or Mutants and Masterminds. I'm most familiar with Champions, so I'll use it in my discussion.

For the guidelines, the GM sets a basic starting power level. There are a few ways to accomplish this; the suggested way to to offer an idea of appropriate levels of points spent on attacks, defence, and other areas of the PC. I typically announce a limit on total Active Points spent on a beginning character and let the chips fall a it more widely in design, but with the characters more evenly 'balanced' across wide capability.

Bob is building a d20 attack, 2d6 damage character. The normal die to use for attacking is 3d6, wanting low. Bob takes a +0 advantage on his combat power that reads "Attack rolls are linear: Attack value is 21 - d20 rather than 3d6". This make Bob more effective on really hard to hit opponents, and less effective on more ponderous characters. The attack is a either a standard sword off the equipment list, or purchased with active points, depending on GM settings.

Carol wants to play a Vancian mage found in another book. She and the GM build guidelines for what the spells can accomplish, what advantage and limitations all the spells will share, and then what power framework will best represent that type of magic. Carol can then buy her initial framework and spells known. That's the preferred (best balanced, most objective and comparable to Bob) way. If you want to avoid that, you could use the book directly at the table, but it is likely that the book was not written to meet the interface specs correctly and some conversion will be necessary. To use the book directly at the table, decide on general conversion rules and assign an arbitrary cost to the character for intiial access and the experience cost to get better.

Dan wants a minimal character sheet (he'll still need one to track the info you require for your API to handle impromtu actions like trying to hit someone with bob's sword). His is the hardest to implement because you have to understand the probability of getting "attack" configuration with the cards, effective capacity to accept damage, etc. and reverse engineer the equivalent cost in game terms. Once the mechanic is understood and defined, it is assigned a point cost. The rest of Dan's sheet is pretty blank unless he wants to do something outside the defined character concept like say learn to fix a car. Then he'd spend points on abilties found in the basic game and use those resolution mechanics for those areas.

The Go board can be implemented directly without mechanical in-game representation though the GM needs to think it through -- it is easy enough to set up a pair of eyes and make his capture impossible assuming you are using normal Go rules.

 

[Asmor]

This can be done with a bit (read a lot) of prep by a GM using the 'toolbox' style games like Champions, GURPS, or Mutants and Masterminds. I'm most familiar with Champions, so I'll use it in my discussion.

I don't really have any criticism for that idea... I just like the idea of an explicit API, which many different publishers could produce for (not unlike the SRD). Doing it your way puts too much of a burden on the individual groups, since they'd pretty much have to come up with everything on their own, and in particular that would make the more esoteric ideas less feasible since it would require a greater amount of mathematical knowledge and playtesting to properly balance than most groups would be capable of.

 

[francisca]

Huh. When I got the 3e books, I pretty much said they read like an API for a programming language, complete with defined terms (stunned, etc...) and MAXINT (20d6), though I meant it in a bad way, as in, "this reads like an API, and is as about as boring".

 

[Nagol]

I don't really have any criticism for that idea... I just like the idea of an explicit API, which many different publishers could produce for (not unlike the SRD). Doing it your way puts too much of a burden on the individual groups, since they'd pretty much have to come up with everything on their own, and in particular that would make the more esoteric ideas less feasible since it would require a greater amount of mathematical knowledge and playtesting to properly balance than most groups would be capable of.

The problem I'd have is trust. Unlike a compiler, there is no explicit way to verify that the design conforms to the API spec save working out the probability details yourself before letting the addition into the game.

Take Dan's card system. Verifying his defenses and attack capability is non-trivial.

 

[Asmor]

The problem I'd have is trust. Unlike a compiler, there is no explicit way to verify that the design conforms to the API spec save working out the probability details yourself before letting the addition into the game.

Take Dan's card system. Verifying his defenses and attack capability is non-trivial.

Well, the same thing can be said about everything... Look at D&D 4e, for example. Despite having fairly tightly-defined guidelines, Wizards itself would frequently publish monsters that were too powerful or too weak for their level. Granted, this was more a problem in earlier products.

And similarly, with the d20 SRD, there were plenty of 3rd parties that put out ludicrously overpowered feats, prestige classes, etc.

You'd just have to read reviews, and more than likely certain publishers would gain a reputation for creating well-balanced, well-playtested products.

Balance in a system like this is difficult, but just because something is difficult doesn't mean you should avoid it.

 

[Nagol]

Well, the same thing can be said about everything... Look at D&D 4e, for example. Despite having fairly tightly-defined guidelines, Wizards itself would frequently publish monsters that were too powerful or too weak for their level. Granted, this was more a problem in earlier products.

And similarly, with the d20 SRD, there were plenty of 3rd parties that put out ludicrously overpowered feats, prestige classes, etc.

You'd just have to read reviews, and more than likely certain publishers would gain a reputation for creating well-balanced, well-playtested products.

Balance in a system like this is difficult, but just because something is difficult doesn't mean you should avoid it.

Actually, I think it's be a neat idea, but the catch phrase would have to be "Trust, but verify."

Part of the specification for release should be a math breakdown of probability / expectation for the sub-system. Most people don't grok the math well enough to understand the implications.