This is an idea for a core dice mechanic that I want to consider in separation from the game at large. I came up with it in the context of working on Illumination, but I might just as well use it for something different. Consequently, I'm leaving some blanks for game-specific components.

What I'm Interested In

*I want this mechanic to: facilitate inter-player cooperation on a mechanical level (regardless of possible in-fiction struggle) and facilitate following certain, game-specific, patterns of action or generating certain cycles of tone and color. Also, I want it to give the players means to strategize a bit, but mostly in terms of choosing options that should give each specific instance of resolution the desired impact on the session as a whole rather than in terms of success and failure. Does it seem it should be effective in pushing the players in these directions?

*Does it seem too fiddly or too complex? Maybe there seems to be too much stuff to do after each check?

*Does it seem like it might be broken in itself, in some particular way?

*Maybe I'm reinventing the wheel?

The Overall Dice Economy

There is a public pool of dice, of various sizes. The number of dice depends on some game-specific variables and is established anew at the start of each session. It's in the range of tens, but only a few dice are rolled in a single instance of resolution, so there's no real need for having buckets of dice (the pool can be tracked on paper or with tokens, or whatever).

The dice in the public pool are opportunities for gaining certain rewards. Each type of dice is attached to a game-specific reward.

For example, in the current version of Illumination d4s can be gained for proving one's beliefs to oneself, d6s for stretching them or dealing with doubt, d8s for proving them to the world at large, d10s for facing the force of opposing beliefs and d12s for expanding one's views. It could just as well be: d4s for adding exotic color, d6s for investigating, d8s for massive property damage, d10s for experiencing serious physical or emotional strain and d12s for proving something through violence. Or, d4s for acrobatics, d6s for creative incorporation of scenery, d8s for sparkles, d10s for being dramatically overwhelmed and d12s for, uh, visual poetry or something. Or whatever the game is supposed to promote.

I'm assuming that the rewards are arranged according to patterns in which the rewarded things are supposed to be most likely to appear in play.

In every scene, the player can gain dice of different types from the public pool, depending on the results of resolution and player-dependent awards. At various times, there are variable quantities of different dice available in the public pool, so there's always more to be gained by doing certain things than others. Sometimes, some rewards just won't be available.

At the end of the scene the dice spent in resolution go back to the public pool - however, they improve by one size. d12s turn into d4s (or, if there was a need for a death spiral in the game, they could be removed from play).

This is where the patterns are supposed to hit. Once you prove your beliefs to the world, you gain some d8s. This means that later someone may have to face the force generated by those beliefs. When you spend your d8s, they will turn into d10s that will be available for everyone. Or, with a different setup, exploring exotic locations creates good opportunities for investigation, or whatever. Depending on what rewards the dice are attached to, there is a certain pattern of content escalation as more rewards of a given type become available.

This is, also, where I expect the cooperative element to be present. At different times, different types of dice might be more useful than others. The players "breed" the dice in the public pool as a group. There's a limited amount of resoures available, but it pays to give them to others, so they could spend them and change them into different resources that will be up for grabs for anyone.

The Resolution

Whenever there's a need to roll the dice or it is called by someone, the player states his intentions, picks a number of dice from his private pool and rolls them. It's fine to mix different sizes.

No dice = no means to influence the fiction and get what one wants when it's contested. Unless the player decides to roll the dice, there can be no mechanical consequences for him, however.

There is a cap of dice that can be rolled in a single check, dependent on some character-specific variables. There's also a limit of checks per scene, as each of those character-specific caps can be used only once per scene. Using the same trick over and over again is not possible, and after some checks it's time to wrap the scene up and move forward.

The results are read as follows:
*Every odd result is a Success, every even result is a Failure.
*Every "1" opens new options - the player can add a single die of his choice to the public pool. I'm not sure if it has some additional in-fiction effect attached.
*The highest showing Success die other than "1" is Good Stuff.
*The highest showing Failure die is Bad Stuff.

(So, rolling different types of dice results in different mechanical impact.)

There is no difficulty or anything like that. If you roll more Successes, you get whatever you wanted and narrate. If you roll more Failures, you don't get exactly what you wanted and someone else (e.g. the GM) narrates. That's base 50% for success regardless of the number of dice, unless some game-specific effects like re-rolls, bonus Successes or whatever are invoked. What matters, however, is the specific impact of success or failure.

Depending on the result, Successes or Failures translate to the dice gained (of a type dependent on context, provided there's enough available in the public pool), and can have some additional game-specific effect. In the current, still untested, version of Illumination, they allow for re-arranging the symbolic structure responsible for scene framing options available, and it affects dice caps in turn. However, this could be something else - e.g. changing relations between characters, establishing facts about the backstory or whatever.

Also, after each check the player gets some Good and Bad Stuff (or gives it to someone else, depending on the check's context). Whatever these are is game-specific, but generally once enough Good Stuff accumulates something beneficial happens and once enough Bad Stuff accumulates there are some negative consequences.

The player has an option of dropping both the highest Success and the highest Failure.

If one happens to act against another player, and only in such a case, both players roll. Whoever gains more Successes wins and reads his dice as normal - the loser's check is effectively voided (this is entirely provisional solution for now).

Hi Filip

I'd like to be more helpful, but all I can tell you from what I read is that I grok the idea of a pacing/structuring mechanism tied to resolution and symbolic link between dice and nature of conflict. I'll check out further developments and playtests!

If evens = failures and 1's are a good thing, then aren't higher dice potentially worse than lower ones, or was adding in an element of risk something you wanted to have happen when characters used higher dice. It really works to have high dice come with a greater degree of threat and potential reward, but it also makes the desire to collect those big dice less strong.

I have to agree with masqueradeball, higher dice are a hazard rather than a boon. I understood that you don't want to increase the chances anyway but there must be an incentive for players to acquire higher dice and an advantage for rolling them. Otherwise the narration could be purposedly tamed in order to keep the pools with mostly d4 and d6.

How about any result higher than 3 is a success (d4 and d12 have 25% and 75% to succeed respectively)?

The dice ladder in the common pool is a really good idea. It could (perhaps) invite players to anticipate the narration a bit more than the usual, to think of their action in a longer term. Keep it on.

Well, you seem to assume that Bad Stuff is something to avoid. However, it doesn't have to be the case. It's a balancing issue, and I can see two ways to handle it:

a). I can make long-term impact of Good Stuff greater than the negative impact of Bad Stuff. If these don't affect gameplay equally, in the +1/-1 manner, a bit less Good Stuff might be worth risking getting high Bad Stuff.

b). Accumulating Bad Stuff will lead to in-fiction consequences, but these don't have to be negative for the player. In fact, I could make Bad Stuff as useful a resource as Good Stuff, if the player is willing to accept the in-fiction impact of possible consequences.

(For the record, in Illumination these accumulate towards expanding and changing the character's belief system in various ways. It has little mechanical impact - mostly, it affects scene framing options and colors possible approaches to in-fiction problems. In the other game I could use this mechanic in, I can see Good Stuff accumulating towards a progress meter similar to InSpecter's Job Dice, leading to definite resolutions of long-term issues, and Bad Stuff as means of pumping up the immediate drama.)

Now, this is the tricky part. But my goal is having all types of dice equally useful in general, only not equally useful in specific contexts. This is why I'm looking for a trade-off between low dice (useful for building up the public pool at the beginning, not so useful once the pool is big enough) and high dice (stronger impact on fiction, useful in bringing things to a close). Also, I don't expect that the amount of small dice could be increasing fast enough to be a serious problem - after all, the spent d4s and d6s dice grow in size, and there's no need to generate more dice past some point.

However, I'm pondering ways of reducing the complexity of handling the results that would allow for retaining the trade-off.

As for the idea of successes at 4+ on a die, I've been considering such a mechanic recently - but it's not exactly what I need here, I think. I want the checks to result in achieving (or not) the intentions roughly 50% of the time, so that effectiveness wasn't as much of an issue as long-term impact of the roll. At the same time I want only a small range of Successes in a single check, as these translate to regained dice and/or moves on a grid of a limited size.

There's a problem with that, though:

Soon after posting the idea I discovered that the probabilities are broken, though. Odd numbers of dice produce 50/50 win/lose chance, but with even numbers there's a possibility of a tie. Now, with two dice the player has 50% chances for a tie and only 25% chances for win/lose result, respectively. With subsequent even numbers of dice the chance for a tie dwindles and chances for win/lose gradualy increase towards 50%, never reaching this value. This means that regardless what the tie means, a strategy that affects effectiveness is possible, and this goes against one of my basic goals.

So, I can either rework the mechanic drastically, or accept that there will be a strategic meaning to a choice between rolling odd or even pool of dice. For example, I could introduce a choice here: maybe in case of a tie the player can decide to narrate the failure of his intentions himself or give the narration of their success to another player. This would create an additional trade-off between rolling low or large pools of dice, as with low, even pools the player would accept the possibility of a control-switching tie.

I'm warry about adding too much fiddlyness, though. Too often I come up with a robust concept and then wind up cutting out needless stuff.

I worked on the resolution a bit more. First, I introduced more choices to the process, but I hit the point when things got too fiddly while narration procedure became somewhat awkward (potentially in an effective disconnect with the mechanics, too). But in the end, I removed the issue of success/failure entirely and, following a friend's suggestion, invited the whole table to each instance of resolution, using a mechanic similar to Unsung's gifts.

It requires playtesting, obviously, but the current procedure is as follows:

1. When the player wants his character's actions to have a mechanical meaning, or another player asks him for a check, he says what he wants to do and how. There is a game-specific limit to the number of checks per scene, most probably approach-based.

2. The player picks some dice from his private pool and rolls them. There is a game-specific dice cap, probably based on the position of the thingy used for the check on the board.

3. Everyone else has an option to suggest a possible development:

a). Success of intention, but with some negative twist.

or

b). Failure of intention, but with some positive twist.

This is backed up by an offer of dice from the public pool, of the size fitting the suggestion. A number of dice up to the current cap can be offered.

4. The player counts his odd and even results and needs to choose:

a). Narrate a success of intention. Afterwards:

-For each odd result, the player can take one die from the public pool, of the size fitting the narration.
-For each even result, the player can make a single move with one of those thingies on the board. Various positions of the thingies on the board have different meanings (i.e. each thingy and each field symbolizes something, and these can be combined and interpreted). Rearranging the positions affects scene framing options (scenes are framed around the thingies, with aspects of its current position included), as well as future dice caps.

(Possibly, there could be a requirement to showcase the thingy's change of position in narration somehow, but I'm not sure if it wouldn't be too constraining. Also, I'm not yet sure whether it's better to move the thingies before, after or during the narration, or whether the order needs to be set in stone at all.)

b). Narrate a failure of intention. Afterwards:

-For each even result, the player takes a die.
-For each odd result, the player makes a move.

c). Take an offer of success along with the dice. The player who made the offer narrates the success and gets to make one move per even result.

d). Take an offer of failure along with the dice. The player who made the offer narrates the failure and gets to make one move per odd result.

5. Good and Bad Stuff (i.e. the highest showing odd and even result) are applied. 1s add dice to the public pool. Dice spent for the check go back to the public pool, growing in size. (I'm not yet sure whether it's better if they return immediately after the check or at the end of the scene; immediate might be easier to handle.)

All in all, I'm still looking for the right balance between Good and Bad Stuff, and it's possible I'll switch odds with evens in the procedure.

It's not as fiddly as it used to be, but there's still quite an amount of handling per check (take the dice, make the moves, apply Good and Bad stuff, return the dice).

I'm not sure how to deal with player vs player checks now, or whether there's a need to have them in the first place.