Fudge Designer's Notes

Or, as we say in the industry: Blather

These comments copyright 2000 by Steffan O'Sullivan
This page last updated January 5, 2001

This document is in three parts:

  1. The History of Fudge
  2. Design Goals
  3. What Next?

The History of Fudge

The role-playing game Fudge was born in the newsgroup rec.games.design, in November, 1992. My fourth book from Steve Jackson Games, GURPS Bunnies & Burrows, had just come out a few months earlier and I was "between projects."

I wasn't really consciously dissatisfied with GURPS, which had been my system of choice since Man to Man was published in 1985. However, I had previously had the contract to write GURPS Faerie, and eventually gave it up. Part of this was due to the limitation of GURPS: it doesn't scale well. When you try to make a race with an average Strength of 2, you find some real problems right away. For example, it only costs ten points to get your Strength at 1.5 times the racial norm - that's the equivalent of ST 15 for a human, which costs much more than ten points. And then it really didn't matter: even with a Strength one-and-a-half times the racial average, you're still not going to make very many ST rolls when you have to roll your ST or less on 3d6.

More importantly was the issue of Faerie itself: GURPS is simply too rigid a system to handle something as nebulous as Faerie.

So there was some dissatisfaction with the GURPS system, but not a lot. I was quite happy running human and bunny games with the system, never dreaming I'd ever want to change.

(And in fact, though I tend to pick on GURPS later in this document, that's merely because I'm most familiar with it. I would pick more on other game systems, but I don't like most of them well enough to have gotten to know them as well as I know GURPS. So please don't take the comments below as anti-GURPS - they're really not. They're just examples of what I found lacking in all existing games, which drove me to write Fudge.)


I don't remember exactly when the newsgroup rec.games.design was created. I'd been reading it since its beginning, though, and enjoyed it back in those early days before it was taken over by computer game designers. Twice in the past there had arisen threads of "let's make our own role-playing game - everybody can chip in!" I'd seen them both die due to the logical result of such a statement: no one can agree on anything, there is no unified direction, and the project eventually peters out through chaos and flame wars.

So the third time it came up (November, 1992), and people were happily debating about incredibly detailed and complex skill rules, I decided to try an experiment. As I said, I wasn't consciously dissatisfied with my current RPG, so I'm not sure what unconscious forces prompted the experiment. Nevertheless, I posted a statement to the effect that I was creating a splinter group: I wanted to create a rules-light, freeform type of game, and I wanted people's input in such a game to make it the best possible game we could. But in order to prevent it from dissolving into chaos, I would take control of the project and make all final decisions. In return for people's input, I promised the game would always be found free on the internet while I was alive. In order to do this, I would have to copyright the material, and if this bothered anyone, then they were warned not to contribute to the thread.

Earliest version of Fudge

So I posted my first skeletal proposal for Fudge, and the input started. (Fudge was actually known as SLUG at first, by the way - the name change came later.)

My first post on Fudge, after the introductory stuff mentioned above, started with:

-----begin quote from earliest Fudge post:-----

Here's what I think net.rpg.freeform should have:

  1. A catchy name.
  2. A loose character creation system that can be as brief or detailed as the player wants.
  3. A single, easy to remember, non-chart-bound game mechanic to handle all actions that need resolution, including combat.
  4. A way to incorporate supernormal abilities (magic, psi, cyber, etc.) without unbalancing the game.
  5. A smooth and logical way for the character to grow in experience.
  6. Open for suggestions . . .
-----end quote from earliest Fudge post:-----

I then proceeded to give suggestions of the first five of those things. The whole post was around 1600 words - a far cry from Fudge today, and, in fact, less than half the length of this document ...

One thing that hasn't changed except in name: the seven-level system. The level names changed a few dozen times, and literally hundreds of words were proposed. My first post included the suggestions:

-----begin quote from earliest Fudge post:-----

Terrible, Inferior, Poor, Average, Good, Superior, Excellent.
-----end quote from earliest Fudge post:-----

Only three of those words survived, but that's actually not bad, when I think about it! The hardest for me to let go of was "Average" - but more than one person was confused by it, since a character Good at something would return a Good result on the average ...

Contributors to Fudge

The early contributors to Fudge were great. We had some fun and even deep discussions about game design theory, and a lot of thought went into every part of Fudge, and not just my thought. Some of the key contributors are named in Fudge itself. Andy Skinner was especially active and helpful - the Scale system, which fixes the problems I found in GURPS, is largely his work, based on my crude beginnings. Andy was also most helpful for bouncing dice ideas off of.

Fudge Dice

Oh yes, dice. Fudge was originally proposed with a card-based action resolution system, and I tried to make that work for a few months before scrapping it. The deciding factor, oddly enough, was shuffling. I playtested so many variants of card-based mechanisms that I came to realize that frequent shuffling was a pain compared to rolling dice over and over, and so finally scrapped the idea of using cards.

I tried to use dominos for a while, but they're too limited. Fun to pull out of a bag, though - I like the feel of nice dominos better than the feel of dice, actually, and they're readily available. But any domino system I tried was either too cumbersome or returned a poor distribution of results, or both. So I finally decided to be mundane and use dice.

But then I really wanted Fudge to be able to use some easily found, normal dice. Six-sided dice, preferably, but ten-siders would do (d6 and d10 in the jargon). I honestly believe that Fudge has had more dice techniques proposed for it than any five other games put together. Andy and Reimer Behrends and I came up with over a hundred just between the three of us, for example, and there were other people making suggestions, too. In fact, they still do! Every now and then someone proposes another dice technique for Fudge and my eyes glaze over ...

For a long time, we used 2d6, one positive, one negative. The lower number rolled is your result - ties give a zero result, as does a result with either die showing a "6". This was actually published in the December, 1993, version of Fudge which can still be found somewhere on the net. I used it in home and convention games extensively for over a year before deciding I had to scrap it. It simply returned a 0 result too frequently. (Without the "6" clause it didn't return a 0 result often enough.) Since no other use of normal dice would do what I wanted, I reluctantly turned to designing my own dice.

Oh, there were many, many designs once I had decided I had to make new dice. And I played with many types of dice: d4s, d6s, d8, d10s, etc. I still have, lying around in drawers somewhere, various d10s, d12s, and d20s with little labels stuck on half the faces, or some faces dyed with permanant marker ink. I stumble on them now again and they always evoke weird feelings. I think I put too much energy into dice mechanics.

By this time, my friend Ann Dupuis had decided she wanted to found a game company and wanted to publish Fudge as its flagship product. I was willing, so long as it was still to be available for free on the net, and she agreed. So with someone willing to have dice manufactured, we began to talk to dice manufacturers.

This was an eye-opening experience, and is directly responsible for the current design of Fudge dice (dF). Most of my designs were simply too costly to consider - they required new dies [as in "tool and die," not as in dice] to be made, large minimum orders, and other expensive considerations. So when I saw Koplow Games' d6s with three plus signs and three minus signs, bells and whistles went off in my head. It would not require any new dies at all simply to remove one plus and one minus, and 4d3-8 was one of my four or five top choices for a distribution curve. It quickly rose to the top when it proved that using plus and minus signs required no numbers at all: the fewer mathematical calculations used to figure out a dice result, the more likely you are to stay in role-playing. So Fudge Dice were born, and I like them a lot. They're a joy to use and don't slow the game down at all, one of my early design goals.

(I'm not fond of dice systems with a flat distribution, by the way - I'm solidly in the bell-curve camp.)

Grey Ghost Games

Ann Dupuis deserves more than the brief mention above. It was Ann who sparked me into doing the major rewrite that is now the current version of Fudge: the June, 1995 version. It was a lot of work, but worth it from my point of view - I think it's much better than it was. I haven't edited it since, though I have amended it with the Author's Latest Thoughts file on my web page and with Five-point Fudge. At any rate, much of Fudge's popularity is due to Ann's promotion of it. She even hunted up as many of the people listed as contributors as she could find and sent them a free copy of the book version! Since her company wasn't even founded at the time those people were contributing, I consider that great generosity.

The fudge-l Mailing List

The good folk on the fudge-l mailing list also deserve some mention. Carl Cravens started it long ago - I don't remember the year, but it's been a while. The list folk are great - we've had some great discussions over the years, and lots of good variants of Fudge have come out of those discussions. It's really an excellent mailing list, since the traffic is just right: not too much to drown your mailbox, just enough to give you something interesting to read now and then. And the quality of the folk on the list is high, in my opinion.


It is only fair to list some games that influenced Fudge. Unfortunately, it's been so long since I wrote it that I can't really be sure anymore which games were a conscious influence, beyond the few listed here. There may well be others. Also, some parts of Fudge were suggested by other people, who may have gotten the ideas from other games unbeknownst to me.

GURPS was the primary influence on Fudge. It really is an elegant game, and most of what is wrong with it has nothing to do with the core rules, which are quite lovely. Obviously, the whole Gift, Fault, Skill thing is a direct descendant of GURPS, fully acknowledged.

Melanda from Wilmark Dynansties was another conscious influence. This game was first published in 1980, many years ahead of its time. Fudge's character creation and magic systems are strongly influenced by Melanda.

Tunnels & Trolls, Dinky Dungeons, and Prince Valiant influenced Fudge in their conscious decision to be simple games, a goal near to my heart. I can't think of any mechanics that came from these games, but there are strong "spiritual" influences. Prince Valiant's use of binary dice made a deep impression on me, though, and fueled the search for Fudge dice.

And, of course, Bunnies & Burrows. Had it not been for this game, the first RPG I ever played, I wouldn't have written any RPG materials. It's the game that brought out my love for RPGs, and still my favorite genre to play and GM ... human PCs are so boring after playing Bunnies ...

Design Goals

"Freeform" is a word I used back in 1992 to describe what I wanted Fudge to be. It has since been used by other folk for other meanings - such is life. By freeform I meant a game that didn't use a rigid structure, that left a lot of decisions up to the GM, and that nothing ever had to be looked up in a book during a game.

Looking Things Up

This last point is one thing that bothered me about GURPS. In fact, I once had a great role-player who was sort of twisted by that aspect of GURPS. I mean, she used to just play the role, which is what it's all about, I think. But as she learned more and more GURPS rules she fell into the mindset that playing by the rules was more important: she became a rule-player as opposed to a role-player.

For example, her character was in a tense situation in one GURPS game, and I, as GM, said, "What do you do?" She replied, "I pull out my gun and shoot him repeatedly!" I asked her to roll the dice, but first she wanted to know the snapshot and recoil penalties. I said, "Don't worry about it, just roll the dice, quick, he's aiming at you!" But she had to pull the rulebook out, look up the various penalties (which I was just going to waive) and only then roll the dice. I found it sad. I still do.

So a big design goal was never to have to look anything up during play. This was actually not a hard step to take, once you analyze what goes on in a role-playing game, anyway.

So what does go on? Well, you're trying to simulate either the real world or some non-real, fictional world. Wouldn't it be helpful to have rules for every contingency? Well, yes and no:

  • Yes, it would be helpful to have easily-remembered meta-rules that covered every contingency.
  • No, it wouldn't be helpful at all to try to have specific rules to cover every contingency because the undertaking is impossible. Any list you could make would be incomplete and would fail at some point. In addition, you'd constantly be interrupting your game looking things up, and the GM would never feel comfortable just winging it to keep the game flow going, because it might turn out that the rules contradicted her.

So, meta-rules: these are actually easy, and many games have them. GURPS has excellent meta-rules, for example: roll 3d6 and compare it to your modified ability. If it's equal to or lower than your ability, you succeed. If not, you fail. That's beautiful, and what attracted me to GURPS in the first place. So Fudge has similarly simple meta-rules that are easily remembered and put into practice. They're different from GURPS, but of the same level of ease of use.

But why doesn't Fudge have rules for common things, such as falling damage, or how much damage it takes to knock a door down, or time to drown, and so on? Don't these things come up often enough to make it worth your while to include them?

My answer is simply: no, they don't. If they came up often enough, you'd never have to look them up. If they only occur now and then, you'd have to stop the game to look up the rule, then calculate the damage, etc. In the meantime, the role-playing mood is broken as index-searching, rule-reading, and technical calculations are made.

In short, that's not role-playing, that's simulation. And here's where Fudge fails the "good game" test for many players: it's aimed at role-playing rather than "realistic simulation."

Many GURPS players don't like Fudge for that reason, and that's fine with me. Because, simply put, Fudge is made for role-playing, not to simulate exactly how much damage you'd take if you fell off the roof of a two-story house. And if this makes a game unenjoyable for you, then you won't enjoy Fudge - be warned.

Of course, since all that information is already found in GURPS, you could just use your GURPS book to look things up like that, if you wanted ... But you may as well just play GURPS in that case, to be honest.

Ease in Role-playing

So a major design goal of Fudge is to be a vehicle for good role-playing. This includes:

  • not looking things up,
  • not having to make calculations when using dice,
  • the ability to return realistic results so as not to break the willing suspension of disbelief,
  • a character sheet you can understand without having to translate everything into English, and
  • minimizing out-of-character actions and statements.

I've already said you don't have to look anything up while playing Fudge, and Fudge Dice require no mental gymnastics to use.

What about returning realistic results? Can it do that? I've just said above that it doesn't do "realistic simulation" - but can it return realistic enough results so as to be believable? Yes, if the GM is any good.

So here we have another limitation of Fudge: you need a good GM to enjoy it. Since I frankly don't enjoy any game with a poor GM, I don't find that a very heavy burden, myself, but your taste may differ - be warned.

At any rate, if the GM is good, it's easy to return realistic results. Since you don't have to worry the book will contradict any ruling you make, you're actually relaxed enough to think logically when called on to make a ruling.

"Hmmm - how much damage would I take if I fell off a two-story roof? Would I be Hurt, Very Hurt, Incapacitated? Or would it be just a Scratch? Or, more importantly, how much damage would an active, physically fit, acrobatic hero in a cinematic story take? Well, I guess it would depend on how he landed, so I'll ask for an Agility roll. But he was pushed off, so I'll make the Agility roll at -1. If he gets a Good or better result, it's just a Scratch. Mediocre or Fair result, he's Hurt. Anything worse and he's Very Hurt."
All that actually takes longer to read than to think, and the GM simply announces, "Make an Agility roll at -1." When the player says the result, the GM instantly replies with a believable answer, and the game continues without interruption. Or so has been my experience.

Personal note: A carpenter friend of mine once fell off a two-story roof. He was Hurt (sprained one ankle), but actually got rid of his torticollis by grabbing the gutter on his way down and yanking his spine straight, so he figured it was a net gain. As an adult, I once fell out of an upper story barn window. I had just spent three days practicing Aikido falls, though, and walked away with minor bruises - "just a Scratch," in Fudge terms. However, I have read of people dying from falling off roofs and out of barn windows, so you never know. It's why I like some randomizing element in my games instead of diceless ...
The next things I list for supporting an atmosphere conducive to role-playing are the character sheet you can understand and minimizing out-of-character actions and statements. Can Fudge handle those? Yes, Fudge does these very well through the use of Adjectives instead of numbers.

It always bothered me in GURPS that players would ask each other, "What's your Stealth skill? Oh, 13? Mine's 14 - I'll go sneak up." In certain other games you'd hear, "I'm fifth level, let me go." It always broke the mood for me - yet it happened a lot. In a Fudge game, however, you hear, "What's your Stealth skill? Fair? Let me go, then - I'm Great at Stealth." It's still not great role-playing, but it's easier to stay in the mood with that type of conversation going on. Likewise with the wound track, discussed above: in Fudge nobody ever says, "Help! I've only got one hit point left!" Instead you'll hear, "Help! I'm Very Hurt!"

I once had the perfect experimental subject as a player in a Fudge convention game. A young woman was not only at her first game convention, she was also playing the first role-playing games of her life. She played in my Fudge game on Sunday afternoon, after having played three other role-playing games that weekend, using three different systems. She started laughing one minute after reading her character sheet. Well, it was a humorous game, so I wasn't totally surprised, but I asked her if everything was okay, just to be safe. She said she was laughing because she understood who her character was, what she was capable of, what her strengths and weaknesses were, and what motivated her - and it was the first time she'd been able to understand any of that all weekend. She told me the character sheets in the other games she'd played were just gibberish to her, even by the end of the game. The only thing she wasn't certain of on her Fudge character sheet was what I meant by the IRA. When I said, "Irish Republican Army. Did you see the movie Patriot Games? Yes? The bad guys in that movie," she was satisfied. And played the role brilliantly, I might add!

You can check out some sample character sheets to see what I mean. And those are mostly crude ASCII sheets - it's even clearer when they're slicked up a bit, such as the Adobe Acrobat Bunny characters (except that you might not understand all the terms if you don't know the genre - but that would be true in any game system).

So I personally think that everything about Fudge contributes, in some way or another, towards ease in role-playing.


Another major design goal was the game needed to be customizable to an individual GM's tastes. I felt this one was obvious, and am still amazed to find people complain about it. Why on earth would you complain that I think you should play a game to your tastes? But some people, it turns out, want to be spoon-fed. They want it all set up for them in advance, and they want it to their tastes.

To these people, I can only say, "Good luck!" You need to find a designer who is not only a good designer, but also has identical taste to yours. Personally, I've never met such a person, but you may. I wish you the best of luck in finding them.

At any rate, even its detractors admit Fudge is very customizable, so I don't have to go into much detail here. I'll just give two examples you might not have thought of. These address two complaints I've heard often enough to warrant some discussion:

  1. Some people want numbers, not adjectives, for their character's traits. Fine with me - that's taste, and it's not up to me to dictate your tastes. Me, I prefer adjectives, but if you would rather have numbers, go for it. Instead of Terrible ... Fair ... Superb, you could use 1-7, for example. Or even ...
  2. Some people really do think Fudge is too grainy (see the next section, Real Math). If you're willing to use numbers, you can simply expand the trait possibilities: 3-18 (or 1-20), for example, with 10-11 being average. This doesn't hurt Fudge at all - it's resilient enough to withstand changes like that. (However, if you do expand the range for traits, I recommend going to numbers rather than adjectives - it's hard to remember more than seven adjectives, and very hard to be able to make a clear progession using more than seven.)
However, one possible downside of customizable is that it actually requires some customization. I say "possible" downside because many people find such customization to be enjoyable. Others don't, however, and are put off by it. I'm sorry, there's not too much I can do about it. I did write Five-Point Fudge to help you make a character easily, but that's about as much as I can do. Others have customized Fudge in ways that suits them - you might want to check some of them out and see if any have tastes similar to yours.

[Note, Summer, 2000: Ann Dupuis has now written and published an already customized version of Fudge suitable for the fantasy genre - no GM-tweaking required (though still possible, of course). This is called Fantasy Fudge and can be found in the Fudge Expanded Edition.]

And by the way, you folks who like the objective systems in the book owe a lot to Ann Dupuis. Between the December, 1993, edition and the June, 1995, edition, I was tempted to rip out most of the objective systems since I don't use them, and in some ways they foster attitudes that go against what Fudge stands for. But Ann convinced me that doing that would not be serving my design goal of Customizable, so I left them in. Gotta provide for as many tastes as possible, after all ...

Real Math, or, Where Did You Get Those Weird Scale Numbers?

Is there any real math in the game? Not during play, but quite a bit in the design process, yes. Aside from a few dozen spreadsheets on various dice and card probabilities, there's some real math in the Scale section.

I'm sometimes asked why I chose such an awkward set of numbers for the Strength and Speed scales. Why not some more common logarithm? The answer is: reality test. I did a lot of research into human strength and speed, and found that the strongest human is about five times as strong as the average human, and the fastest human is about twice as fast as the average human. This is based on lots of data gathered from lots of countries over the last fifty years - well, if you work in a library, as I did when writing Fudge, you have access to lots of data and usually know how to find it ...

So if the strongest and fastest are Legendary, and the average is Fair, that means Legendary Strength should be five times stronger than Fair Strength, and Legendary Speed twice as fast as Fair Speed. Once I set those two numbers, the math followed easily.

Obviously, those are approximations, though, so if you feel like changing the Scale numbers to something more easily remembered, please do so. Just make sure Legendary isn't too far from the figures in the book.

Oh, there were other areas where I dealt with a lot of math - damage, of course, though I declined to reality test that, thank you. In that case, I just kept playing with the numbers until I was happy with game results. Again, if this bothers you, use your own numbers. Just don't reality test them on me, please ...

What about that "graininess" - doesn't that bother me? Nope. I think it's an illusion.

Let me first explain what is meant by grainy. Fudge has only seven levels to describe your ability with a skill or attribute. Some people find this too few - it feels "grainy," like granular sand as opposed to fine sand, which they somehow feel their other games are. But this is an illusion because most games have de facto graininess pretty close to Fudge's.

By this I mean that a 3-18 scale system, for example, rarely uses most of those numbers. Most stats will be in the 12-18 range. This is no wider a spread than Fudge has. True, true - many Fudge players don't use the bottom two levels, either, so that can be a problem. However, Joe Gill ("snowgen" on the fudge-l mailing list) came up with the following after a post Carl Cravens had made to the fudge-l list. Carl was having problems with his players not wanting skills below Fair, but he finally managed to define "Mediocre" positively enough to convince his players that Mediocre was an acceptable skill level for PCs. Joe had a different and more amusing solution to this problem: renaming the levels. So the levels would look something like:

  • Superb
  • Great
  • Good
  • Pretty Good
  • Fair
  • Mediocre
  • Terrible
This works because of the psychology of roleplayers: most of them don't want to have low-sounding skills, so they shy away from them. By shifting the names down, you expand the levels people feel comfortable listing on a character sheet, reducing the graininess. You could do this too, if the supposed graininess bothers you. In fact, why not go further, if you really want to?

  • Superb
  • Great
  • Very Good
  • Good
  • Pretty Good
  • Fair
  • Poor
Or whatever ... but we've strayed from Real Math into Psychology - sorry.

[Note added August 2011: I've actually now shifted to playing something similar to this. See VG Fudge.

So How Is It?

Well, I can't say - I'm too close to it. I like it; I know that much. I translate every setting I have any interest in into Fudge and run it that way. (I estimate that this year [2000] I have passed the 400 mark: over 400 sessions GMed (or played in) using Fudge. And I still love it - I consider that a successful design.)

So far, I prefer it to any other game out there. But that's really not surprising, since I wrote it to my tastes.

Oh, I steal from other games quite a bit. But that's the norm in this industry, you know: everybody's influenced by the games that came before. In my case, though, I'm able to be influenced by the games that come after. This is because Fudge is specifically designed to incorporate new ideas - that comes under the "customizable" umbrella, above.

So when something new comes out and it has a great idea, it'll probably show up in a Fudge game of mine at some time. For example, I once ran a Fudge Groo game at a convention using the Groo Trading Cards as action resolution tools as explained in Everway (which came out after Fudge). I found Fudge was able to handle it just fine.

So though I'm not impartial enough to give you a final verdict on Fudge , I will say this: it would be very hard for me to go back to a rigidly structured game at this point. I just couldn't do it. The freedom I find in Fudge is addictive.

And by the way, I have never made any money off of Fudge.

What Next?

Darned if I know. To be honest, at this point other people are doing better things with Fudge than I think I could. I may not write any more Fudge things at all. I've got a hankering to get into board game design, and I may just pursue that rather than continue to publish Fudge books. In a way, this is good. Fudge will grow and develop more if more people have input into it. This has always been true of Fudge from the beginning, and it's still true now.

So instead of answering, "What next?", I'll ask it of you, instead: "What next!?"

Back to The Fudge Page
Back to SOS' Gameviews
Back to Steffan O'Sullivan's Home Page