Up-front, I want to state this is a different type of thread. Anyone is free to contribute, of course, but I'm going to ask that only people who have actually published a complete game (that is, your game is available for free, in PDF, in stores, by direct sales, or however else, and you feel that it is complete) answer questions, and others ask questions in response.

So, my question is: what is the process by which you design a game? This sounds like an pretty open-ended question, but I'm specifically asking about the process. Do you:

- Evolve the game through play of it, finally having a complete rule-set after many sessions of inventing and testing new rules to cope with situations?
- Latch onto an interesting mechanic, and then design a situation (or setting) around it?
- Think of a situation (or setting), and start to design mechanics around it?
- Something else?

I'll answer, and admit upfront that my experiences are a bit muddled. That's why I'm asking the question, to be honest.

With Donjon, I played a game of D&D in which we used the major conceit of Donjon - whenever a successful roll was made, something happened. If I looked for secret doors, and was successful, there was one. Armed with that, I knew I wanted a game that did that. The original rules were a satirical look at D&D, with no 'successes' mechanic to judge how much narrative power one had. Through discussion with Zak Arntson, we came up with the idea of using successes in this manner. From there, the rest fell into place through discussion of my playtest version of the rules.

With Paladin, a different process although was followed. I knew exactly what the situation I wanted to emulate was - the struggle of being a holy warrior. From there, I tried different mechanics until I found one I liked, and wrote a simple version of the rules. Again, I developed it from discussion of the playtest version of the rules, although in this case, I only used input from actual play of the playtest version.

In both cases, I had to wonder - where did the mechanics actually come from? How did I find a mechanic I liked? I'm looking for input on what step-by-step process others follow, or if they follow one at all.

I haven't really published anything in a professional fashion, but several of my systems have been used by others in various gaming groups and distributed around informally, so I'll chime in on a technicality, because the topic interests me.

In general, I've always designed my mechanics with three goals in mind.

First, the mechanics of a game should be simple and elegant. I've never found it fun to be sitting around a table with a half dozen open books, trying to cross reference and figure out the results of what should be a simple action. My belief is that the core of a game, and the majority of the important rules, in their entirety, and the most common situational modifiers should fit on no more than two sheets of letter sized paper. So any time I find one of my designs spiraling up to where I've got a dozen pages of important things, I go back and start looking for places where I've got commonalities and can extract the important fundamental rules, leaving the differences as some sort of simple noted modifier.

Second, the system should never get in the way of play. It's fine for it to enhance play, but it should never ever slow it down or make it boring. As a result, if I find my mechanics are accumulating Search and Handling times, I tend to try and simplify them.

Lastly, the mechanics should fit the genre as much as possible. I always admire when someone manages to build a game system that really enhances the feel of a genre during play, such as Dust Devils' use of poker hands. The Amber Diceless system is another one where the system just fits hand-in-glove with the way the setting is written. Dying Earth is a good example, too.

As to the actual design of the systems, I'll admit I'm typically not very original in my creativity. Rather than come up with truly unique stuff, I'll take elements I like from several systems and attempt to synthesize something I like as a whole. Or sometimes, I'll take an entire existing system and try to reduce it to the simplest possible form.

To use a recent example, I've been working on a system I call Lux, which is going to be used for an epic fantasy game I'm starting for one of my online groups. We've played a lot of Amber, some World of Darkness games (adapted to be diceless), and a fair bit of BESM. But none of those were really suitable to what I wanted for the style of fantasy game I was looking at.

I'd been looking at S. John Ross' Risus, as several of the elements intrigued me, notably the idea of clichés as sets of skills and the fact that the whole system was very minimalist yet entirely functional. However, one of the things I've found is that diceless systems work much better online than ones where dice need to be rolled; not having to use a dice rolling program saves a lot of time. Several people had proposed variants on how to make Risus into a diceless game, but none of the results were very clean, in my opinion. However, I'd also been looking at Nobilis, as some comments on the mailing list had made me consider running Amber using the system from Nobilis, so the whole system was fresh in my mind.

Combining the idea from Nobilis and Risus gave me a very interesting start; Risus provided the fundamental structure for the game, while Nobilis' system of resource management made the diceless aspect work. A week or so later, I read a review of the Buffy the Vampire Slayer RPG, and a column on RPG.net about its Drama point system. I went down to my FLGS, picked up a copy, and got the last piece I needed to turn what was a seedling idea into a complete system that hung together well. The Drama Point system combined with the resource management in a surprisingly clean fashion. At that point the only thing I really needed was some tuning, which I hashed out with a friend who happens to be heavy into game theory as part of his economics PhD, and the end result was something I'm fairly satisfied with.

In general, most of the designs I've built have worked the same way. Having picked some particular basis from another system I admired, everything else just fell into place around it.

-DaR

(edit: I are good at speling and grammar. No, really.)

Clinton R. Nixon wrote: So, my question is: what is the process by which you design a game? This sounds like an pretty open-ended question, but I'm specifically asking about the process. Do you:

- Evolve the game through play of it, finally having a complete rule-set after many sessions of inventing and testing new rules to cope with situations?
- Latch onto an interesting mechanic, and then design a situation (or setting) around it?
- Think of a situation (or setting), and start to design mechanics around it?
- Something else?

I guess I qualify; Multiverser is in print, and I've got in the works three completed board games, two other RPG's, and two CCG's.

I'm co-author of Multiverser, and was brought on late in its development because of my peculiar approach to games: I'm a problem solver. That is, someone says, I want the system to do this, and I hammer out a way for it to do that. Thus for example, E. R. Jones had this rather nebulous idea about something called bias, which he knew would be a measure of character ability in different kinds of skills, and a measure of a world's support or blocking of such skills, and a defining factor of skills. But he didn't know how to make all that work. I devised the bias curve system, laid out the skills into levels and intensities, worked the bias factors into skill learning and performance, and otherwise made a functional system out of a vague idea.

Similarly, on another game, the designer had created a system in which he wanted it to be possible for characters to reach a point at which they could survive an attack that was powerful enough to destroy a star, but for ordinary mortals to be viable characters in the same setting. Tied into traditional thinking, he'd had characters with millions of hit points who really couldn't be killed by much of anything, against characters with hit points down into single digits. I looked at it and suggested revamping the entire damage system into a three-part concept. Keep damage points, but limit them to one hundred. Then add two additional defenses which could be accumulated as characters advanced. One would be what I termed "probable immunity". For any defined type of damage (e.g., kinetic, radiation, photonic, et cetera), the character would have a listed percent probability that he would take no damage from the attack. These could be raised incrementally to something around 95%, giving increasing odds that no damage would be taken but leaving open the possibility that the character might take full damage. The other tweak was what I called damage dividers; you could incrementally raise your scores to ten, again in categories for types of damage. If you were hit by an attack for which you had a damage divider, you cut the damage in half for each divider you had--but never took less than one point if you were hit. This enabled his high powered characters to throw around very dangerous weapons, but not to have the absolute certainty that they would survive the next attack; and it made it possible for an ordinary character to kill a powerful one, with the right attack and the right rolls.

So when I'm on my own (as I was with the board games, the first games I did) I generally pose the problems to myself, and then look for solutions.

There may be a "first step" behind that, and that is that I have to decide what I want the game to do. This could be called that moment of inspiration, when the light comes on and you see something that might make an interesting game. It doesn't always work--I think that for decades I've been trying to invent triads, the card game with triangular cards that is always played but never explained in sci-fi stories, and I always wind up with something unworkable. I think perhaps triangular cards are the wrong shape; but even with rectangular cards, it's been challenging--and perhaps largely because this isn't an inspiration of what do I want the game to do, but rather one of wanting to make a game that fits the name. If I start with a goal, then it's just a matter of solving the problems so I can reach it.

That's pretty much what I do when I answer game design problem posts on forums, too--I look for a simple way to solve the problem.

--M. J. Young

My experiences are somewhat similar to the rest of you.

I usually start with a setting idea or a piece of a story. There has to be some sort of emotional pool or "cool idea" reaction for me to even begin. Then, I tend to lay down notes, fleshing out the setting, what the characters do, and system bits. The system itself tends to evolve -- but I'm not really picky in regards to my system. In fact, I've been doing most of my guys with my own house system, because I like it (it fits my gaming style). The key for me is to get the why of the game -- and then I can play with the system later to support it.

Of course, I don't always do this the same. I've designed mechanics which support settings and settings that support mechanics... It's weird. I used to buy this "floating" idea that a game system has to emulate its genre or game to the "t", but I don't think I really believe that anymore. They need to fit -- but what does it mean to emulate? What is the difference between chunking a bunch of dice and rolling one die to resolve an action? In my book, none really -- probabilities perhaps -- but nothing that really changes the way my group games.

And really, I have nothing more to add....

Thanks,
Nathan Hill

Both of my "published" original games stemmed from a perceived unfulfilled need in my own play.

For Toast, I wanted to find a use for the bloody d12s I had accumulated. Thus was born the d10+ mechanic. I also wanted a detailed skill list that I could be happy with. Toast filled these needs and was a fine, workable system for a short campaign.

For Pollies, I was going for something simple and flexible that eliminated the need for the GM to select a finely focused target number, that still provided a sense of less granularity than Fudge.

Usually, my game design is a direct result of me having had an unpleasant or unsatisfying gaming experience that makes me think to myself, "I can do better than that..."

As to my actual process and steps taken for design, I usually start with a mechanic idea that fits the style of GMing and play that I want to support. Then I write it up and test it. Tweaks and additions are applied as necessary. I rarely start with a setting in mind, but usually I have an idea of what genre(s) (SF, Fantasy, Horror, etc.) I want the game to handle.

So, I guess I qualify; Torchbearer is complete, if not polished to a high-gloss finish.

My process with TB was sort of bidirectional:
The setting slowly evolved through over several years, through play and occsional writing. I had a good idea of the 'feel' of the TB world long before I began designing a game around it.
The system was, at first, just an attack at a cool mechanic, which I then tied to my mythic fantasy setting, and hammered at a few tsicking points until it fit. Now I'm standing outside of the setting again, and looking at ways I can expand the game, provide it customizability without losing the cleanness and power that I discovered in the TB system.
The appearance of Exalted galvanized my desire to create a game that modelled my heroes the way I wanted it to; it showed me that the mediocrity of D&D was not the only path to wealk.

Refreshing Rain is a game built on a triad of mechanics; the situation serves to tie them together. With Rain, my goal is for all the mechanics to meld smoothly together, blurring the borders between them.
Both this and The Calligrapher's Sword were just phrases that I liked before I made them into games; TCS is very much a game that tries to tack a bizarre phrase onto a slightly less weird initial conceit.

The Calligrapher's Sword is a game built on exploring a situation - a strange shared dreamworld that has elements of Wonderland, the Moorcockian dreamworld, and Lovecraft's Kadath. Its mechanic is obviously very influenced by lumpley's OtherKind; I wanted to take the deep decision-making of that game, and combine it with variability.

I always start with the characters. What kinds of characters are the players supposed to be in this game, and what do those characters do? From there I usually come to the question, what specifically about the characters is important enough to measure? -- the answer to that is most often a short list of traits/attributes/skills/whatever you like to call them. This can get very specific at times. I'm working on a set of mechanics now for running "Beyond the Mountains of Madness" -- not Call of Cthulhu in general, but just specifically that one adventure and nothing else.

After that, "How do I want to measure them?" is the next question. What sort of dice-rolling mechanic to use. To be honest, the answer, for me, is nearly always "It doesn't matter," -- I just need a way of generating success vs. failure that I'm comfortable with, like FUDGE. Then I can tweak it here and there to make certain kinds of actions more or less rewarding, depending on the genre tropes and how I want the game to feel.

Mostly it's intuition, and it gets revised often during play. I game with a tight group of friends who are very comfortable and honest with each other, so there's not much incentive to get every rule down with laser-perfect clarity. There's lots of fudging and "good enough" is generally, well, good enough.

In my experience, I start a game design with a seed, I have never set out with a goal for the game to become, rather I start with a core element that I intend to preserve through the design process. Often a more general element could split off into a variety of seeds.


For example:

About a year and a half ago I attempted to solve the problem of players who hate character creation. Since ultimatlely characters need to exist to some extent (otherwise it's a board game not an RPG), I developed a seed to start building a game from. The seed was "character creation occuring in play". This is a fairly simple idea, and ended up generating three seeds from it: CC (character creation) as discovered pastlives, CC as karmic manipulation, CC as matched heroic limitations / tragic flaws.

Currently the first of these seeds has been brought out to a game, aptly named Pastlife. The remaining two are still in the works.


The seed isn't necessarilly a mechanic, but from my perspective it needs to be one of two things:

1) A thought-provoking mechanic which strongly suggests some criteria for a setting.

2) A thought-provoking setting element that requires a fairly unique approach to it's mechanics.


For example:

As I'm writting this, thinking about mechanics which don't apply to my criteria, I considered mentioning tree climbing. Basing the mechanics of an RPG off of tree-climbing seems like an interesting idea (although the computer scientist in me wants these to be mathematical trees). The two key elements from a mechanic like this are the ideas of branching and of levels. That would be ideal for a game of diversive narratives, which eventually attempts to reconcile the streams. This sort of element fits a time-travel or parrallel world setting very well.


Althought my Speculative Physics column tries to formalize this approach, the basic idea in there seems to be how I do approach most game design, passing back and forth from designing a handful of setting elements and then a handful of system elements, each building off the other.

The core idea behind that, is each decision in game design should be carefully weighed and considered, all too often people include mechanics or setting elements for no apparent reason, other than this is how they've seen it done before.

On the other hand, I've also attempted to attack pre-existing settings, making games that simulate the external logic of the setting. However the same principle applies, plumbing the genre to find key elements which need mechanics, and testing the mechanics to refine their fit with the genre.


For example:

When I started designing Mythic Strains I was attempting to simulate the way comic book stories actually happen. Several key elements were necessary. First, characters could change drastically, but not necessarilly becoming more powerful in the process. Second, any character who fights is automatically very skilled in combat, and this level can easily change story to story. Main characters are always more potent than side characters, even though which characters are main characters may change each story.

In other words, I attempted to write a game which describes the evolution of the comic book characters, while accomodating the inconsistencies which are the nature of comic book settings. Not entirely sure I succeeded.


All in all, I've designed over a dozen playable RPG's, and have far too many on my plate right now. Ultimately I view the design process as an exploration of the possibilities. Roads not entered this time around, become places to be explored later. I simply cannot imagine designing a game with your destination in mind, especially with so many possiblities right out there.

-Mendel S.

Clinton R. Nixon wrote: So, my question is: what is the process by which you design a game? This sounds like an pretty open-ended question, but I'm specifically asking about the process. Do you:

- Evolve the game through play of it, finally having a complete rule-set after many sessions of inventing and testing new rules to cope with situations?
- Latch onto an interesting mechanic, and then design a situation (or setting) around it?
- Think of a situation (or setting), and start to design mechanics around it?
- Something else?
In both cases, I had to wonder - where did the mechanics actually come from? How did I find a mechanic I liked? I'm looking for input on what step-by-step process others follow, or if they follow one at all.

FVLMINATA started with a setting (Rome with guns) in need of a system. I looked into historical Roman gambling rules and I just played around with them … seeing how I could turn them into an RPG resolution mechanic. I came up with several different ways to read these strange Roman dice and we settled on one of them for Action rolls and another for Effect Rolls.

We did playtesting, and made small changes here and there. Changed the range of the Attribute scale … changed how Humors and Patron gods worked, that sort of thing.

The other main design problem was placing proper emphasis on social skills. Jason definitely did not want this to be a murder-and-plunder kind of game. Rome was a civilization, intrigue needed to be important. I designed the Influence system with one principle in mind: “I want social interaction to be just as certainly effective as combat.” I figured that most games encourage combat by making it the only way that the players can definitely affect the progress of the imaginary sequence of events. (mind you, this was all before I found the Forge) Due to time constraints, the Influence system went to press with far less playtesting than I would have liked, but it seems to work fine (see Jake Norwood’s post in Actual Play).

Well, Junk isn't an RPG, but it is a published game, so there you go.

Junk started out as a stupid conversation about Battletech. I don't remember any of the details except that we were talking about trash-can mechs and I insisted that such a game would require a weapon called the Porta Potty Cannon (to poke fun/pay homage to the Phase Particle Cannon of Battletech fame).

My friend started poking at the idea, and I ended up taking over design work. Something else that had wandered into the concept was the beer-powered engine. Originally it was just one option of power plants. I made it the only option. From the beginning it had been an unstable source of energy, and I codified that.

Once I got around to the idea of actually working on a mech game, I incorporated an energy allocation system, because I enjoy the strategy of such systems. This meshed well with the beer-powered engine concept. I also decided, somewhere along the line, that I wanted this game to be wild and crazy and random. Thus I tended to build in large amounts of unpredictable effects into the game.

So, to sum up, the game started as a seed idea, joined with a design feature or two that I liked from other games, which was fleshed out in keeping with certain design principles that were evolved from the seed idea.

As I look at this post, it does not seem to have been a very organized approach. And believe me, in practice it was even less obvious.

Seth Ben-Ezra
Great Wolf

Interesting topic

I suppose Wayfarer's Song qualifies me to answer.

Let's see... I think I personally went through several different phases in which I took different approaches to game design. I'll explain:

Intuitive
For some time I simply wrote up a game based on what felt right. I had a few basic ideas about what I peronally preferred in a game system, but these were little more than 'quick and easy mechanics are good', 'one page character sheets are good', 'game blance is good'. I did not really have a goal, so much as a general feel for how I wanted a game to work out. A lot of my early disasters occured during this phase. Also I tended to stick to the straight and narrow fantasy path (Hell, I still do, but now I want to rather than not realise there are other genres)

Rebelleous
A second phase I think I went through could basically be summed up by: Well D&D and every other game I know does it this way; so I will do it like this instead.
I wasn't so much trying to create a good game as a different game. The very early primordial ancestor of Wayfarer's Song came from this phase. I wanted to create a game system in which skills and difficulty were measured as descritive words rather than numbers only and purely because I hadn't seen it done before. I came up with a system and found that my players jotted numbers down on the character sheet, and just kept referring to the numbers. So I was wrathful. I changed the system not because I thought a change would make for a better game, but because I was damned intent on my original 'this is different' idea. So, in the end, it was almost pure chance that I ended up with a workable system.

Conceit
I'd like to think this is a more mature approach to game design and one that comes from talking more with other designers. I am still kind-of in a Conceit phase now, but am also quickly moving into Jaded (see below). In the Conceit phase I tend to have one or two ideas, maybe a bunch of ideas. These ideas may revolve around a setting, or what character's do in the setting, or how stories are told, or somethign else that I think will be fun or interesting to explore. I'll then begin to build mechanics around the conceit.
For instance a game I am still tinkering with 'Folk-in-the-Mist' came out of a few basic conciets:

* The default setting is 17-18th century Ireland
* Players each control an entire family living in a rural village
* Players can swap from one family member to another at will
* The characters have to contend with faeries, tax collectors, natural disasters, and English lords, to do well for the village over time
* The Storyteller lays out a natural landscape and the players collaborate on placing buildings and whatnot.

So, from a list of conciets I am already resitricted in the way the system can go. With a bunch of characters to look after (2-8 maybe), each player needs their characters to be pretty simple. There may need to be rules for swapping from one character to another in-game. Because there are guns, rules for firearms are needed. Rules for faerie magic. Rules for folk magic and superstition. Maybe all these rules can be simple one-line guidelines. Maybe more complex. I am not sure yet, it will depend on how it playtests. But, I can now begin devising rules that suit my conceits.

If my conceits were (to pick an idea out of the air):

* The setting is a surreal dreamland
* Players are mortal dreamers trying to save the dreamland from nightmares
* Characters and setting are based on Jungian archetypes

Then my system would be restricted and channelled in an entirely different dirrection.

EDIT: CHanged 'conciet' to conceit. Damned i before e rule.

Jaded
I am pretty much quickly slipping into this phase now-a-days. Maybe It is just a temporary thing, but I don't know. The Jaded Phase is the result of the following:

* I have seen a lot of mechanincs. In fact I feel like I have seen every possible mechanic that an RPG could use. I know this logically to be untrue, but I still feel this way.

* I can pick-and-choose from all the myriad mechanics I am familair with

* I know what I like

So basically I am in a frame of mind in which I feel like there is little or no point in working hard to be different, or innovative because I can create perfectly good, playable games, based only on what I already know, love and use.

So that is about it. I think that pretty much sums up the different approaches I have taken to game design over the years. Hope you find that useful/interesting Clinton

Chris

Interesting breakdown, Chris. Cheers.

Clinton R. Nixon wrote: Do you:
- Evolve the game through play of it, finally having a complete rule-set after many sessions of inventing and testing new rules to cope with situations?
- Latch onto an interesting mechanic, and then design a situation (or setting) around it?
- Think of a situation (or setting), and start to design mechanics around it?

Clinton R. Nixon wrote: Where did the mechanics actually come from?
How did I find a mechanic I liked?

Peregrine wrote: Jaded
I am pretty much quickly slipping into this phase now-a-days. Maybe It is just a temporary thing, but I don't know. The Jaded Phase is the result of the following:

* I have seen a lot of mechanincs. In fact I feel like I have seen every possible mechanic that an RPG could use. I know this logically to be untrue, but I still feel this way.

* I can pick-and-choose from all the myriad mechanics I am familair with

* I know what I like

So basically I am in a frame of mind in which I feel like there is little or no point in working hard to be different, or innovative because I can create perfectly good, playable games, based only on what I already know, love and use.

Universalis was a special case, in some ways. First it was a collaboration, which obviously affected the way it came out. But also, it was a game that developed from a single idea. That being the notion of having a world that was developed in play, mechanically. That's right, Universalis was supposed to have GM, PCs, and all the usual jazz, originally. Thing is, just as we were designing it, we were caught in the wave of people expeimenting with directorial mechanics, etc. It wasn't long before we just decided to expand the player's authority to develop the world to developing, well, anything.

IOW, we strted with a simple idea, and the game developed itself from there. Which is to say that I (and I believe Ralph as well) really had no idea that it would end up where it did. Still, the heart of the original idea is still in there, so I can say with confidence that it served as a guidepost for where we were going.

Which seems pretty straightforward, and corellates with much that's been said already in this thread. Basically, have some idea that's compelling to you. Then develop your other goals around the first. Then complete the game based on the goals you've chosen. You'll find that, if you do that, the design is just some work to finish it then. Not to make it sound easy, it's a lot of work to do the design right. But until you have a clear idea of where you're going, at least in one or two particulars, the only kind of product you can complete is a confused one.

That's my philosophy right now.

The other thing I suggest is to steal, steal, steal. Originality, as far as I can tell, is just taking lots of other existing ideas and mixing them up in new combinations. Oh, occasionally somebody will develop something completely new (seemingly, I suppose), but it's rare. And I would hazzard that it only regularly happens as an accidental offshoot of the normal development process. Which means that you have to be stealing, and compliling before you can change things to make them now. So start by stealing, and then try for originality along he way, if you feel the need.

People may note that Synthesis by myself and JB Bell, is nothing but mechanics cobbled together from several games, for example. Later in the design, I found some ways to make it all look a bit new, and I'm satisfied that it's unique enough. But it started very much as a game created from using what I needed to meet my goals taken from several other games. The cool thing is, if you steal from enough sources, the combination is new, anyhow. So, to reiterate, steal, and steal unabashedly. Let originality come later or as a product of mutiple thefts. :-)

Mike