I have been giving some thought to the dice mechanism in FU lately. A lot of thought. In fact, I have been chasing myself in circles thinking about how actions are resolved, what is important and why. I decided I would write this post so that;

- I could clarify for myself, just as much as anyone else, what some of the current options are, and
- Get some feedback from you all on what you like and how you feel about some of these ideas.

**This is a long article, where I spend as much time trying to clarify my own thoughts as I do trying to explain my ideas. Thank you to those of you who persevere! I want your feedback on dice rolling and action resolution – give me your opinion! Do you agree or disagree with something I have said? Do you feel strongly about something here? Comment below or over on Facebook!**

First, though, let’s back-up…

## System does matter

I am a strong believer in the idea that “system does matter”. This is an approach to game design that says, basically, “it does matter what system you use when playing your game”. If you like highly realistic, simulation type stories, for example, then you will want a set of rules and systems that support that style of play. You can read more about system does matter here. (In fact, you probably should, I will be making reference to the article below.)

#### So what type of play does FU best support?

So, what is FU about, and what should the rules be supporting? To me, FU is a *narrativist* game. When I wrote it I was primarily interested in helping players and narrators tell awesome stories about cool characters. Sure, there is a little *gamist* in there, too, as everyone likes to win, and it feels great to overcome the big bad at the end. But mostly, its narrativist.

Supporting this style of play are (in my mind) two key features:

**Tags**let players create the characters they want, and help everyone describe an interesting world with a minimum of intrusion from the rules themselves.**Yes/No/And/But**quickly tells the players the outcome of an action and helps to push the story into new and/or interesting places.

I am totally confident on this. For me, I know what the “outlook” of the game is. **What I am less sure about, however, is the resolution method. **

#### Dice? What dice?

In “system does matter” nomenclature, FU always has been *fortune* based – that is, you use a randomiser to determine the outcome. I like rolling dice, and so do a lot of other role players so that’s what I have gone with, and what I will probably go with in the future. It also helps to add that gamist edge to the game.

From personal experience, and getting to know you, the FU community, I have come to the following conclusions (please correct me if you disagree with any of this):

- Dice rolling should be quick and then get out of the way so we can just tell the story.
- Calculating positive and negative tags should be straightforward and clear.
- Any result should be possible, no matter how awesome or awful a character is at a task.
- Players like to know the chance of success before dice hit the table.

*FEEDBACK: Do you agree with this summation? Let me know in the comments.*

## Now what?

What follows is a brief summary of a variety of systems that have been used to resolve actions in games of (or based on) FU. This is probably not a complete list, but a start. I offer a brief overview of the system, my thoughts on the pros and cons, and some opportunities I see from using the resolution system (ways to modify it, or cool things it might do).

They are offered in no particular order.

## Classic

This is the original system. Roll a d6 to find one of six results. Add dice for bonuses or penalties and take either the highest or lowest result.

**Pros:** it’s quick to pick out the highest or lowest die from a group. You only need one type of dice and it doesn’t matter what colour they are!

**Cons:** adding more than one or two bonus or penalty dice very quickly adjusts the probabilities.

**Opportunities:** a larger die could be used. A d12 would mean each of the possible six results would be allocated two numbers (1-2: No and, 3-4: No, 5-6: No but, 7-8: Yes but, 9-10 Yes, 11-12 Yes and). It is just as quick to spot the highest or lowest die, though it may take a fraction longer to determine the actual result as you think through the result (“A 3? That’s a… ‘No’ result.”).

Other dice could also be used…

## Classic with bigger dice

This is similar to the classic system, but uses dice with more sides. A d8, d12 or d20 could be used by reducing the possible number of outcomes to four: Yes and, Yes but, No but or No and. You still roll the multiple dice and pick the highest or lowest (quick) and use the number to determine success;

**Pros:** still quick to pick out highest or lowest die. The broader range of numbers means that you can roll more than one or two dice without turning things too much one way or the other.

**Cons:** Lose the straight “yes” and “no” options. Takes a moment to align roll with actual result (“a roll of 4 is…”). This may get easier with time (and lots of rolls). Rolling multiple d20’s might actually make the results a little too random.

**Opportunities:** ummm…. gives you a use for all those d20’s….

## Vagrant Workshop’s variant

This method is very close to classic FU. It uses a d6, the six possible results, and the classic “beat the odds” where even numbers are good and odd numbers are bad. The base die will determine what happens while bonus or penalty dice that roll a 5 or 6 will modify that result. One or more 5+ on bonus dice will let you change an odd number on the base die to the next highest even number (so turn a 3 into a 4, or a 1 into a 2). Penalty dice that roll 5+ will roll it down (turn a 4 into a 3). If two bonus or penalty dice roll 5+, you also get to add an “and” statement to the result.

**Pros:** similar to the classic system. Using customised +/- dice makes it quick to see if you get a bonus or penalty. Reduces the impact of bonus and penalty dice.

**Cons:** without custom dice it does take slightly longer to sort through options/results. Requires at least two different coloured/sized dice.

**Opportunities:** you could play with probabilities further by using a different sized die for bonus/penalties. The d6 with a bonus on 5 or 6 means you will get that advantage 33% of the time. A d8 with success on 7 or 8 is a 25% chance of a bonus. A d10 will easily let you give a bonus on anything between 10% to 90% in easy 10% increments. Of course, there is also the opportunity for funky custom bonus and penalty dice.

Instead of having bonus and penalty dice cancelling each other out you could roll them all together – just collect all the dice as you discuss/describe the situation and roll. + and – results cancel each other out and any remaining results modify the base die.

If the success ladder was “low numbers bad, high numbers good” (as opposed to traditional beat the odds), each + or – could modify the result up or down, rather than the current roll up an odd number to an even number.

## Buckets of dice

The buckets of dice variant uses pools of d10 to resolve actions. The player rolls 3d10 versus an opposition pool of 3d10. Bonuses add dice to the player’s pool while penalties add dice to the opposition pool. Roll and see how many of the player’s dice are higher than the opposition’s highest roll:

**Pros:** minimises the impact of bonus and penalty dice. Is a more traditional “versus” dice rolling situation. Only requires d6.

**Cons:** is a more traditional “versus” dice rolling situation (narrators have to roll dice). There is potentially buckets of dice clattering the table. Need to compare two pools does take longer.

**Opportunities:** it doesn’t have to use d10s, so other dice could work (though smaller sided dice will draw results towards the middle more).

You could have asynchronous dice pools, where a player rolls one type of die and the narrator a different type (d10’s vs d12’s, for example) depending on the difficulty. In fact, rolling dice up or down to the next larger or smaller might be an alternative to adding dice to the pool. You can use polyhedral dice as bonuses or modifiers to take into account power scale or other abilities.

“Left over” dice (beyond the 3+) could be used for other effects such as damage, special rules, etc.

## FUBAR variant

Roll five dice and count up how many roll an even number. Positive and negative descriptors add bonus or penalty dice to the base dice pool and you read the best (bonus) or worst (penalty) 5 results.

**Pros:** impact of bonuses and penalties is reduced. Get to roll a handful of dice. Only requires d6.

**Cons:** sorting through a lot of dice, which can take time.

**Opportunities:** rather than look for even numbers, look for 4+ as it is slightly quicker to look for the highest dice. Doing this could let you use polyhedral as modifiers/bonus/penalty dice to account for power scale.

Removing the straight “yes” and “no” results means the same system can be achieved with a base pool of 3 dice. 0 successes = No and, 1 = No but, 2 = Yes but, 3 = Yes and.

Extra dice in the pool that roll 6 (bonus dice) or 1 (penalty dice) could be used for additional “and” phrases or other effects.

**A variation on the above:** Use the 3-dice resolution above but do not have a pre-set pool. Just roll one die, add more dice for bonuses and remove dice for penalties (to a minimum of one die). Rolls of 6 let you roll an extra d6 (meaning even when starting with just one die you might get a “yes and…”).

## Dice Pool

This is the system described here. Begin with a dice pool of one positive and one negative die (dice of different colours). Add positive and negative dice based on descriptors etc as normal (they don’t cancel each other out). Roll the pool. Match pairs of positive and negative dice that rolled the same number and discard them. Look at the remaining dice – if the highest die is positive the result is a “yes”, while if it is a negative die the result is a “no”. If the next highest die is the same type (positive or negative), add “and”. If the next highest die is the opposite type add “but”.

**Pros:** relatively quick to assemble pools as you just gather dice as discussing. Quite random.

**Cons:** quite random. Possibility to cancel out all dice leaving an outcome not covered in the traditional rules. Straight “yes” and “no” results can only occur when an odd number of dice are rolled. Requires at least two different coloured/sized dice. Difficult to determine probabilities.

**Opportunities:** the ability to succeed with low dice rolls means the results could be used to determine effects such as damage, power drain etc.

You can use polyhedral dice as bonuses or modifiers to take into account power scale or other abilities.

## Wastburg variant

This is the system used in Cedric Ferrand’s *Wastburg* RPG. The basic 6 results are possible, but the number of dice the player rolls depends on the difficulty of the action, determined by the narrator. There are 5 difficulties for a closed question: (- -), (-), (0), (+) and (++):

- (- -) Roll 3d6 and keep the lowest die
- (-) Roll 2d6 and keep the lowest die
- (0) Roll d6 and keep it
- (+) Roll 2d6 and keep the highest die
- (++) Roll 3 6 and keep the highest die

Descriptors can either add +1/-1 to whatever the result die rolled. (This is a slight variation on the actual *Wastburg* rules as descriptors and FU points work differently to classic FU.)

**Pros:** very close to the classic system. Only ever need 3 dice. Finding the lowest or highest die is quick.

**Cons:** two steps to resolution – determine difficulty and roll, then add modifiers. Even 3d6 will heavily swing results (as it does in classic FU) and the addition of a modifier on top may further impact this. Using a difficulty means there is no need to calculate penalty descriptors for opposition/scenery etc.

**Opportunities:** have the narrator indicate difficulty, then modify that scale by your character’s descriptors (The task is difficulty “- -“, but I have a Magic Sword, improving it to just “-“, so I roll 2d6 and keep the lowest result…).

## Roll + modifiers

Roll a die (probably a d8, d12 or d20), add modifiers for bonuses or penalties, and compare final result to outcome chart. A roll of “1” is always a “No, and…” (even if modified to a larger number). A roll of the highest number on the die (8, 12 or 20) is always a “Yes, and…”.

**Pros:** very familiar process to role players. Relatively quick. Only requires a single die which makes it quick (no searching for results). Easy to determine odds as each bonus adds a clear % increase to the result (e.g. +1 when rolling a d20 increases your chances by 5%)

**Cons:** requires some math before the roll (adding and subtracting modifiers). Requires referencing a chart until players are familiar with outcomes.

**Opportunities:** special effects on a natural 1 or maximum result. The larger the die the more modifiers that can be added.

## Almost done…

Ron Edwards suggests in the article mentioned earlier that game system should consider search time and handling time. I personally think that handling time is pretty quick in FU, no matter what system is used – as soon as you have your yes/no/and/but answer the story moves on (perhaps with a small amount of book keeping).

Search time, which is the process of selecting dice, calculating modifiers and counting up the result is where attention needs to go. I am beginning to think that it almost doesn’t matter how you get to your yes/no/and/but result, so long as it does not take very long or require too much brain power. How much brain power? How long? They are the million dollar questions!

### What are your thoughts on dice and resolution systems in FU? How many of the above have you tried? Do you feel strongly about any? I have been attempting to try as many as I possibly can, but my current playtest opportunities are limited. If you get a chance to try any out, let me know about it, either in the comments or via an email. As always, your feedback, ideas and opinions are welcomed and appreciated!

Have you contemplated on expanding the Answer Oracles? I’m planning on using that to test drive the damage systems you suggested in another post. I like the Answer Oracles because they use a simple to understand percentage system. Except the first one or two incarnations, which used d6s. Plus, it’s just nice to just have a general concept of the character (like ninja, accountant, or just “everyan”) plus a goal for the character, and play right away.

I had not thought about expanding the answer oracle. Have you got an idea what that might look like?

I agree, one good thing about FU is how quickly you can create character and get into play!

I will have a suggestion or two by the weekend

I have Answer Oracle version 8.1 in front of me as I write this. I believe this is the last version of the Answer Oracle. The one thing that stands out the most is the column “Skill Expertise” as I believe it should be in reverse order from what it is. At the top of the column is the designation is “Paramount.” If it’s the character that’s supposed to have that level of expertise, then it’s wrong, as the character will fail no matter what the character rolls. If it’s a skill level the character is up against in an opponent, then it would be pointless to have the character go against that opponent, since the character will always lose.

I have a question about another column. The “Difficulty (Absolute) column has number going from 10 to 1 in descending order. I was wondering if that was supposed to be some tie-in to the FATE ladder? It would be interesting if someone made a FATE ladder oracle. (Someone made a FATE dice oracle called FATE Solo. But it’s not as clear as it needed to be for instructions how to use it.)

Other than those two things, the Answer Oracle 8.1 is quite good.

We’ve only used the “classic” dice rolling, and it is fairly swingy with that first or second bonus. If you are 1d down, it’s usually not worth rolling. If you are 1d up, you are usually pretty safe.

We may experiment with your Dice Pools system, although in CONS you mention that it’s very random. Didn’t someone do the math calculations to show that the curve is “better” than classic?

Gut-feel is that using 1d6, with Fate Dice +/- is the best compromise for me.

PS: Happy that you are blogging again. I was wondering when we can expect some more FU goodness, or (no pressure) FU2.

Unfortunately I have come to the conclusion that the classic system, as much as I love it, just doesn’t work for anything but the most light of games. It could work as part of an “almost” diceless system where you primarily just compare bonuses and penalties.

The Vagrant Workshop method is very good and still feels a lot like classic FU. I have found, though, that custom + and – dice really speed up the process.

I am also glad to be writing again. There is more on the way, but locking down some dice mechanics for future articles/projects has become important.

I like the Roll + modifier – my more traditional players would probably dig it, too, and I have a handful of pretty d20 that I’d love to use more. You could even make it pretty easy to “uplevel” your descriptors/modifiers over time that way.

Not sure yet how to implement negative modifiers from the environment or opponents, that could lead to a lot of calculation. The classic FU system works so well because you can put the actual dice on the table and group them this way or that for canceling out etc., which you can’t visualize as easily with just one D20 and more abstract modifiers. Maybe some kind of token…

I also like the Wastburg system; it’s actually pretty similar to what I’ve been discussing with one of my players – he suggested instead of counting up all the negative descriptors, it would be quicker and easier to just give a difficulty rating (like: “you get 2 negative dice for this; it’s pretty hard to scale that wall”).

The roll+modifier does have some distinct advantages. I find it very fast, though I am not sure if that is because there is only one die to refer to, or because I am used to that style of calculation.

It is interesting you mention counting up tokens to track modifiers, as a previous draft of that article actually mentioned that as an option! Great minds! 😀

There is always the entierly custom dice route as used by the Fantasy Flight starwars games. The essentailly run two sets of symbols. Succes / Failure, and advantage / disadvantage

OK actualy three sets as they also have the Triump / Disaster pair which represents cirticals, but apear at most once per die.

I like the Fantasy Flight approach, but it takes some getting used to. The other problem that many players flag is that it’s difficult to determine the probability of success or failure. I’ve only played a couple of games, but did find it hard to work out how to effectively challenge players. Activities that I thought would be straight forward sometimes were very tough for players! (This might be my lack of experience with the system though.)

That was a very, very insightful post! I also believe systems matter, so much so that I have spent a good amount of time tweaking the system to my liking.

My main concern was to find a way to add different bonuses or penalties depending on how applicable a descriptor was to the context. So I changed the size of the dice, but I think the canceling process is a little clunky as it is.

My second concern was with the probability curve. So I like the idea of rolling a single d6 as base, and having a d10 as a single bonus/penalty die with 10% increments. One applicable descriptor: move up one step on the ladder if you roll a 10, or something like that (I haven’t done the math yet).

I might experiment with the modifiers. They do seem simpler, and I certainly don’t think adding and subtracting 1 multiple times is such a big deal.

Do you mean the descriptor will adjust what you need to roll on a d10? Or is it always a flat 10% chance to get a “bonus”, but if you do it will adjust the base die by how many relevant descriptors are in play.

Both are interesting alternatives. The first option means you always now what you are looking for on the d10 (a 10+), but if you are lucky enough to roll that, it could adjust the base die significantly (depending on how many bonus descriptors you have). The second option requires a bit of math up front to find your d10 target number, but will also mean that the base die can only ever be adjusted by 1. This will significantly reduce the impact on final outcomes, even if you get a bonus on a roll of 2+.

It would be something like:

– Base die is always a d6

– For every applicable descriptor, you have +10% chance of increasing your d6 roll by +1

So, if you have one applicable descriptor (or detail/condition), you have to roll a 10 on a d10 to increase your d6 roll. If you have 2 descriptors, you have to roll a 9 or a 10. If you have 3, roll 8, 9 or 10 and so on.

That does dramatically decrease the influence of the modifiers (descriptors, conditions, details). So an alternative would be granting 20% (roll 9 or 10) for a fully applicable descriptor and 10% (roll a 10) for a somewhat applicable descriptor.

The problem is: no matter how many applicable descriptors you have, you’ll never increase your d6 further than +1.

Anyways, I’m pretty much thinking out loud at the moment.

Thinking out loud is mostly what I do here, too! 😀

That explanation is what I thought you meant.

I’ve been messing with this and I came up with something I am starting to like. It’s super easy to use but somehow it seems hard to explain. I’ll try:

– You only ever roll two dice: a d6 (base die) and a d12 (modifier die);

– Depending on what you roll on the d12, you get a +1 or a -1 on your base die;

– Bonuses and penalties determine what you have to roll on the d12 to get +1/-1;

– Bonuses are translated into “top numbers” of the d12. So a highly applicable descriptor might grant you a “top 3”, meaning if you roll 10, 11 or 12, you get +1;

– Conversely, penalties translate into “bottom numbers”. So a somewhat applicable flaw might mean “bottom 2”, meaning a 1 or 2 on the d12 would result on -1 on the base die;

– Bonuses and penalties stack up and don’t cancel each other out. So, after considering all the variables, the GM might say something like “I give you Top 5 Bottom 3 for this roll” (or T5B3 for short). So if the player rolls 8-12, that’s a +1; if she rolls 1-3, that’d be a -1;

– In case Top Numbers and Bottom Numbers overlap (because of the number of applicable variables), those overlapping numbers become neutral.

Is it clear? I swear it’s simpler than it looks, hahaha.

Instead if choosing highest dice how about allowing any bonus dice that is higher than the base to increase base by +1. Similarly, penalty dices lower than the base would lower the base by 1 level. This should lower the swinginess a bit without changing the system. For example, rolling base of 3 and bonus dices 3,5 would rise base from 3(No, but) to 4(Yes but).

That’s a great suggestion. I’ll give it a try.

I was looking at the probabilities of the FU, Vagrant and the proposed method, which indeed lowers the swinginess when compared to the FU system. Interestingly, as you add first bonus dice, the left most result (No, And) drops the most while the probability of the remaining results increase uniformly. Adding additional bonus dices decreases probabilities of the left most results one by one while increasing uniformly probabilities of the right results. This contrasts with the behavior of the Vagrant system where probabilities of the left results (No And, No, No, But) decrease uniformly and with the same amount as the right most terms increase. Hope this makes sense. I guess it is a matter of design choice. If you want your character to still have a reasonable chance of the extreme results No, And for bonus dices and Yes, And for penalty dices, Vagrant method is better. If you desire more safe-proof characters, who while still capable of failure, fail thru No and No, But results, the proposed method may be chosen. I am trying to link a probability spreadsheet.

https://drive.google.com/open?id=0B-2aYyet6KdXU2VvSXh2M3A5WGM

Very helpful. I can’t read the spreadsheet very well on my iPad but will check it out when I get a chance.

Are you saying with the die pool suggestion that a +1 bonus increases the chance of “yes and…” by X, but decreases the chance of a “no and…” by a much bigger number? While the Vagrant method the chance of a “yes and” increases by X while at the same time decreasing the chance of “no and” also by X?

Is the Vagrant method essentially the same as a “traditional” roll a die and add/subtract a modifier from the result? D20 +1 increases the chance of a “Yes and” by 5% and decreases the chance of a “no and” by 5%????

Probability really isn’t my thing!

For one bonus dice, yes. For 2 bonus dices I had a slight error, I did not consider an extra AND statement coming from a pair of + results, which I have now corrected. The calculation of the probabilities for the Vagrant method is base of the following interpretation of the rules:

1 successful bonus dice upgrades any of the No result to closest Yes result and has no effect on any of the Yes results.

For 2 bonus dices, whenever a pair of + results is scored, besides the upgrade from No to Yes, an extra And statement is also added. This extra And statement is incorporated as follows:

Scoring 1: No And, is turned to 2: Yes, But by the first successful bonus dice and then the AND statement due to the second successful bonus dice turns 2: Yes, But to 4: Yes. I did this as there is no result of Yes, But And, and for comparison with other methods I still want to retain the same range of results. Also, while 1 bonus dice does not affect any of the Yes results, a pair of + results gives AND statement which upgrades Yes to a higher level. Finally a result of Yes, And And is categorized as Yes And.

As far as the modifiers are concerned, it is a bit difficult to make a direct comparison with a dice pool method. Adding +1 to a d20 roll changes the range from 1-20 to 2-21. If we classify 21 as 20 to retain original range of results then the probability of scoring a 1 drops to 0%, probability of scoring 20 becomes 10% and probabilities of scoring any of the remaining numbers remains same at 5%. So we are no longer able to roll 1, an extremum, whereas in dice pool method even if you add more than 1 bonus dices, you will still have, even if it is very low, a chance to score an extremum No And.

So I know I’m a bit late to the party but the variant that I have been toying with is roll 3D6 and then add a pluses (+) or minuses (-) in one (1) point increments based on tags with a max + or – of three (3). Then look at a chart for results.

4 or less = No, And

5 to 7 = No

8 to 10 = No, But

11 to 13 = Yes, But

14 to 16 = Yes

17 plus = Yes, And

I found this creates a nice bell curve which is lower on the ends with the base 3D6. However once you allot + or – the curve shifts to the appropriate end with the full +3 or -3 removing the opposite extreme (i.e., No, And or Yes, And). This might be a Con but I like the idea that the PCs can, with tags, remove the worst outcome, and in the same way obstacles can remove the best success. But neither make it impossible to succeed or fail.

The percentages in this have most of the rolls, at the base 3D6, hitting the (No, But or Yes, But) which is not bad to me because both allow for action and plot advancement. Again once you add + or – this changes.

Another thing to point out is FU points can be still be used to increase success chances by PCs allowing them to triumphantly overcome in places not normally allowed.

Hopefully all that was understandable. I will close by saying I really like the FU system and appreciate all the work that goes into it. FU has even replaced the Apocalypse World engine as my choice for my home grown game.

Great suggestion, Chris. I’ve been toying with d12 and d20 with + or – descriptors. The d12 can work with the full range of six “answers”:

1-2: No and…

3-4: No…

5-6: No but…

7-8: Yes but…

9-10: Yes

11-12: Yes and…

A bonus or penalty of 2 or more will remove any chance of getting an extreme result at one end.

Reducing down to only four answers widens the spaces and means you can have more bonuses or penalties before any option becomes impossible:

1-3: No and…

4-6: No but…

7-9: Yes but…

10-12: Yes and…

OR

1-5: No and…

6-10: No but…

11-15: Yes but…

16-20: Yes and…

Only rolling a single die does lack the bell curve that comes from multiple dice.