After much "aha! the fix is obvious! except... oh... but then.. dammit... but...", I've finally come up with a complete new mechanic for apocalypse girl. This mechanic is more complicated than the original, and the description below still needs radical rewriting to be user-friendly, publishable rules, but it is coherent enough for me to seek comment -- and for people to playtest it, if they really want to endear themselves to me.

The essence of the mechanic remains the same: story-elements (characters mostly, but also groups, ideas, and warring tendencies in a single person's personality) are represented in the game by "Engines" rated for "Power," i.e. how much they let you do, and "Identity" (formerly "loyalty"), i.e. how likely they are to stay under your control. But there are several crucial differences -- in brief:
1) Funny dice!
2) Loyalty is renamed "Identity" -- and gets more resonance than it had
3) "Charges" (i.e. Capes Inspirations) are gone entirely
4) Time becomes part of the currency of the game, in a way cribbed entirely from Tony Lower-Basch's Capes
5) "Fallout" occurs as a result of a very Vincent Baker Dogs in the Vineyard-like die-matching system

Engines
Each Engine is defined by two things: Identity and Power. Each has a description -- a single word, a phrase, or a short sentence -- and a numerical score (higher numbers being stronger). An Engine’s Power description should tell you what that Engine does to influence the story; its Identity description should tell you why that Engine is worth caring about in the story. These descriptions needn’t be exhaustive or exact; what you want is something suggestive that will help everyone come up with cool narrations.
[Identity renames the old "Loyalty" -- sort of: Because Power and Identity now have descriptions as well as numerical scores, they both convey the "meaning" of the Engine and have, hopefully, real thematic significance. They can also change much more easily in this version: In the original, the only way to rewrite the Meaning of another player's Engine was to take it from them; now, you can rewrite an opponent's Engine's Power or Identity any time you reduce its score -- and you can only rewrite your own Engine's Power or Identity by increasing their scores. This subtle change is, I think, a big deal in terms of supporting potential "female" story types about personal change and "what do I become" as well as "male" story types about "do I choose this loyalty over that one?"]

No narration can state something about an Engine that contradicts its current Identity and Power descriptions -- but whenever you or any other player successfully changes the numerical Identity or Power score, you can rewrite the appropriate description as you like.

A newly created Engine starts at Power 0, Identity 0. If an Engine’s Power is ever reduced below 0, it is eliminated from the game, and what it represents is eliminated from the story. An Engine’s Loyalty cannot be reduced below 0: Instead, if an Engine is at Loyalty 0, the first player to roll successfully to change its Loyalty back to 1 takes control of it -- regardless of who originally created the Engine.
[This makes it significantly harder to capture an Engine than the old rule of "exceed the Loyalty Score": Now you have to beat the Identity (Loyalty) score down to zero and remake it in your own image]

On Your Turn
On your turn, you can do one and only one of the following four things:
1) Defer your turn, saving it for later -- “foreshadowing.”
2) Put new dice into play on one and only one of your Engines --- “development.”
3) Transfer unused dice from one and only one of your Engines to another -- “support.”
4) Roll dice from one Engine onto another existing or newly created Engine to try to change it -- “conflict.”
A conflict scene is the only way to change the numbers and descriptions of any Engine in play; it is also the only way to introduce a new Engine into play.
You can never simply “pass”: Instead, defer and foreshadow (it will get you a turn to use later!).
On your first turn, you will have no dice, so your only options are to defer or to develop.
[These four options are crucial to the new currency. The original draft locked the dice:time exchange ratio so that you could only roll one die a turn; Unco Lober's and BJ's playtest groups immediately started playing with multiple dice on a turn, but even then they kept the idea that everyone refreshed their dice pools for all their Engines at the same time. Now, you're forced to choose every turn to do one thing with one Engine: get more dice, move dice around, or roll dice -- or defer the turn altogether. This means that while small Engines (e.g. the Power 1, Identity 1 "crawler") are still more efficient in terms of dice invested vs. dice returned, the larger Engines are now more efficient in terms of time, because one turn with one large Engine may give you a whole bunch of dice that you would need many turns to build up with a lot of small Engines. Making time a part of the currency this way is an insight I learned from Capes.]

Foreshadowing Scenes: Defer your turn

You can defer your turn, saving it for later.

Narrate a short, mysterious scene that hints at the possibility of something later -- foreshadowing. Make a note on a 3”x5” card. Then play proceeds to the next player.

At any later time, at the end of any player’s turn -- including your own -- you can tear up the 3”x5” card, narrate something that builds on your foreshadowing earlier, and take a normal turn. (If two players both want to do this at the same time, the one with the most foreshadowing cards on hand gets to go; the other does not get a turn at this point, but does not lose their card either).

You can defer as many times as you like, building up an unlimited number of foreshadowing cards; and you can spend as many of them as you like in a row to give yourself multiple turns in a row.

Development Scenes: Get new dice

Choose one and only one Engine you control.

Any unused dice still on that Engine from previous Development are discarded: “use them or lose them.”

Place new dice on that Engine whose total Power equals the Engine’s Power:
- a d2 (i.e. a coin tossed with “tails” being a 1 and “heads” being a 2) costs 1 Power
- a d4 costs 2 Power
- a d6 costs 3 Power
- a d8 costs 4 Power
- a d10 costs 5 Power
- a d12 costs 6 Power
No other types of dice are used in apocalypse girl.

On average, many small dice will produce higher totals than a few larger dice (because their Power costs are equal to the average value they roll, rounded down): Six d2s, on average, will roll a total of 9, whereas one d12 will on average roll just 6.5. However, small dice are more likely to cause you Fallout (see below). That means that many small dice are the high-risk, high-reward option and should be narrated as physically or emotionally risky actions in the story, while a few large dice are the low-risk, low-reward option and should be narrated as more cautious actions.

[Yes, there are funny dice. I worry about them myself. In fact, I scorned any nonstandard dice -- until I played Dogs in the Vineyard. Now I'm sold that different die types can be used for very cool emergent effects]

Narrate a short scene about the Engine you’re developing. You must reveal (or invent) at least one new thing about that Engine that has not come up in the story before. If the Engine is a person, personality trait, or group of people, you should also describe his/her/their emotional state in a way that makes sense for the kind of dice (many, small, risky, or few, large, cautious) you are adding.

You may only add dice to one Engine on your turn.

Supporting Scenes: Transfer dice

Choose any two Engines you control. Transfer one, some, or all of the unused dice on one of the Engines to the other: Those dice now can be used off the second Engine (and, conversely, are no longer available off the first).

Narrate a short scene about how the first Engine is somehow supporting the second: perhaps one character in the story is giving another information, weapons, or an inspiring word (or an infuriating one); perhaps one is secretly manipulating the situation behind the scenes to help the other without the beneficiary ever being aware of it.

The transfer of dice in a supporting scene does not create any kind of lasting relationship between whatever the two Engines represent in the story, and the two Engines are not considered Linked, which means the dice transferred cannot rebound from one to the other as Fallout (see below).

You can only transfer dice from one Engine in your turn, and you can only transfer those dice to one other Engine.

[The "supporting scene" allows you to play very cautiously: generate new dice on valuable, high power Engines, but then transfer them to some expendable Engine before you use them, so your high-power dice-producers aren't vulnerable to Fallout, because they're not in the conflict. This is a very expensive strategy in terms of time, however, because of all those turns spent on transfers instead of getting and rolling dice]

Conflict Scenes: Rolling dice

Choose one and only one of the Engines you control: For this turn, that is the “active Engine.”

Then choose one and only one other Engine to be the “target Engine.” This Engine can be
- any other Engine you control: You usually do this to build up your own Engines.
- any other Engine controlled by another player: You usually do this to attack and wear down your opponent’s Engines.
- a new Engine you just made up: Write “Power 0, Identity 0” on a blank card, write up appropriate descriptions, and put it into play. This is the only way to introduce a new Engine into play.
The active Engine and the target Engine cannot ever be the same Engine: No Engine can ever roll dice on itself! [The "no-bootstrapping" rule -- which also means that if you have a single, super-powerful Engine, and nothing else with any dice left on it, your uber-Engine can't defend itself!]

Take one, some, or all of the unused dice on your active Engine and roll them. You can only roll dice from one Engine on your turn. (If there are no unused dice left on the Engine you chose to be active, you’ve just wasted your turn).

Place those dice you just rolled on the table in between your active Engine and the target Engine: Those dice are now considered “staked” and, in the story, symbolize an intense relationship between the two Engines. (You may want to move the two Engine cards around on the table, run a string across the table between them, or use some other physical indicator to make this clear, especially).
[My stumbling block for a long time was that I wanted to incorporate relationship maps of some kind into the game, but I was thinking of them as restrictive: A can only act on B if a relationship exists. But, with the rocks in my head loosened by a few judicious whacks from Ron Edwards, I realized that the relationships on a r-map aren't restrictive at all: The little lines between A and B and C don't represent the only possible interactions, and in many cases two characters with no prior relationship can and should interact, but what the little lines do represent is unresolved dramatic intensity -- potential energy, if you like.]

Declare whether you intend to reduce the Power or Identity of the target Engine – in which case your dice are considered to be “staked against” that Engine – or whether you intend to increase the Power or Identity of the target Engine – in which case your dice are “staked for” that Engine. If you stake dice for an Engine, then all other dice staked for that Engine are considered “allied dice” and any dice staked against that Engine are considered “opposing dice”; if you stake dice against an Engine, any dice staked against that Engine are “allied dice” and any dice staked for for that Engine are considered “opposing dice.” [Here the terminology gets laborious. Suggestions for revision are very, very welcome].

Narrate your active Engine doing something in the story to influence the target Engine for good or ill: maybe one character is talking to another, or fighting, or singing on the radio and the other just happens to hear, or maybe the target Engine’s character is remembering something the active Engine did in the past, or maybe destiny is at work and the target Engine is about to do something that will cause the active Engine’s future.

Now you have three options:
1) End your turn now and let play pass to the next player. Leave the dice you have rolled and staked on the table, with the Engine cards and the dice all lined up to clearly indicate which Engine the dice are staked from and which Engine they are staked for or against. The dice will remain in that position until someone either cancels them (see below) or expends them (see below).
2) If there are any opposing dice already staked on that same Engine (i.e. dice staked for the Engine if you are staking against, or dice staked against if you are staking for), then you may try to Cancel them, at the risk of Fallout. As long as there are any opposing dice staked on the target Engine, you may not expend dice and attempt to change it.
3) If there are no opposing dice currently staked on your target Engine, there is nothing for you to Cancel, but you may expend dice and attempt to change the target Engine.

Conflict Scenes: Cancelling Dice and Creating Fallout

Choose one, some, or all of the dice staked on the target Engine on your side -- that is, dice you rolled and staked this turn, or dice you rolled and staked in an earlier turn, or even dice rolled and staked by another player that are “allied” to you (because you both are staking for the Engine, or both staking against it).

Set the dice you just choose slightly aside and add up their value: That is your “set.”

Now make another set by choosing one, some, or all of the opposing dice staked on the target Engine -- i.e. staked for the Engine if you are staking against, staked against if you are staking before -- whose total value adds up to less than the value of the dice you just picked. (If the two sets of dice are equal in value, or if the opposing set is higher in value, you need to start over).

Both sets of dice are now “cancelled” and no longer staked on the target Engine. One and only one of three things now happens:

1) If the two sets each consist of the same number of dice, then discard all the dice: There is no fallout. (For example, if your set consists of a 4 and a 8 from one of your Engines, plus an allied 3 staked on an earlier turn from another player’s Engine -- three dice totalling 15 -- and the opposing set consists of a 2 and a 10 from one Engine, plus a 4 staked from a different Engine – three dice totalling 14 – all six dice are discarded and none of the Engines involved suffers Fallout).
2) If your set consists of fewer dice than the opposing set, then the opposing side takes “fallout”:  Your dice are discarded, but each of the dice in the opposing set is now immediately staked against the Engine which it was staked from before. Move the dice around on the table to an appropriate new position. (So if you had that same set of 4, 8, and 3 -- three dice totalling 15 -- but the opposing set consisted of a 2 and 6 from one Engine plus a 1 and 5 from another Engine – a total of 14, but from four dice – then your 4, 8, and 3 are discarded, but the 2 and 6 are now staked against the Engine they came from and the 1 and 5 are now staked against the Engine they came from).
3) If your set consists of more dice than the opposing set, then you take fallout: Each of the dice in the opposing set is discarded, but each of the dice in your set is now staked against the Engine it came from. (So if you had that same 4, 8, and 3 -- totalling 15 on three dice -- but the opposing set consisted of only a 9 and a 5 -- totalling 14 on just two dice – then the 9 and 5 are discarded, but your 4 and 8 are now staked against the Engine they were rolled from, and the allied 3 is now staked against the Engine it was rolled from).

Note that since you get to choose what dice to put in each set, you get to choose whether you will take fallout or not -- this turn: It may make sense to take fallout now voluntarily to prevent your opponents imposing worse fallout on their turn, or to get opposing dice out of the way so you can change the target Engine.

[Belabored writing aside, this is straight out of Dogs in the Vineyard, except the escalation is sideways -- "do I bring in and risk yet another Engine?" -- rather than upwards. It's all "okay, I need to match and exceed a bunch of your dice with a bunch of my own, and I could use all these lovely little d4s (and in this game, d2s!) but they're going to hurt me." This replaces the original "cancel" mechanic, and the old "capture" dice-matching mechanic is just gone, because I was trying to reinvent the wheel instead of just borrowing the Ferrari I'd fallen in love with. If you're going to steal from Vincent, just steal from Vincent already]

You may keep making sets and cancelling dice as long as there are still dice staked on this given target Engine on both sides.

If you cancel out all of your dice and all of any allied dice staked on this Engine, your turn is over.

If you cancel out all of the opposing dice staked on this Engine, and you still have dice left, you may immediately expend your remaining dice and change the target Engine. (See below).

Conflict Scenes: Expending Dice and Changing Engines

If you have dice staked on an Engine and there are no opposing dice staked on that same target (whether because no one ever rolled any or because all those dice have been cancelled), you can try to change the Engine:

If you discard dice whose total value is greater than the Power score of the target Engine, you can increase or decrease its Identity by one and rewrite its description accordingly.

If you discard dice whose total value is greater than the Identity score of the target Engine, you can increase or decrease its Power by one and rewrite its description accordingly.

Once you’ve done this, if you still have enough dice of high enough value left, you can reduce or increase the target’s Power or Identity again.

P.S.:

The Gun, in this implementation, is very straightforward: It doesn't need any special "recoil" or "ricochet" rules at all. Instead, it's an Engine that no one controls or can change, which anyone can take dice from, which gains +1 Power and refreshes all its dice every time it's used -- and which always gives you nothing but d2s. Enjoy the fallout.

Endgame is unchanged; special powers for each role remain an idle thought -- but I think that these mechanics give enough room for different strategic combinations, each with different story implications, that can express each player's thematic stance very tangibly:
- emphasize Power, Loyalty, or a balance?
- lots of cheap, expendable, little Engines that take a lot of time to mobilize, or a few big, precious, costly ones that can all hit in a few turns?
- stock up on deferred turns or act every chance you get?
- use transfers to concentrate dice on a few expendable Engines, or put your best Engines directly at risk?

Prior threads, for reference:
Official Ronnies feedback - http://www.indie-rpgs.com/forum/index.php?topic=18165.0
First playtest reports - http://www.indie-rpgs.com/forum/index.php?topic=17702.0
Unofficial Ronnies feedback - http://www.indie-rpgs.com/forum/index.php?topic=17632.0
Original rules draft - http://www.1km1kt.net/rpg/Apocalypse_Girl.php

Sydney,

Unclear. As I understood it, if you have more dice staked that your opponents, then their dice are discarded. But your dice are moved back to your engines in a fallout, and are staked against it. Thus, how come it is useful to stake dice at all, and how come one has dice staked after the conflict, so that he could use them to modify engines, at all?
I seem to have misunderstood something completely here, and humbly asking for explanation.

It is also unclear to me, whether you have to overcap Engine's Power/Loyalty with the sheer _number_ of dice, or the results rolled. If the first, then how the numbers rolled are used and where?

Sorry for me being somewhat dumb {:{)= .

Playtest follows, probably even this evening, and "featuring" a non-roleplayer (though a mmorpg-er). Also, thanks for the great game, again, Sydney!

Aha. Exactly the kind of question I needed to be asked. Let me try to explain myself better:

The first thing that matters is not the number of dice rolled, but the total value of the results.

If my staked dice exceed the total value of your staked dice, then I can cancel out all your staked dice and still have dice of my own left over to change the target Engine. Conversely, if your staked dice exceed the total value of my staked dice, then all my dice get cancelled out and I can't do anything to the target.

And, again, with the dice I have left over, what matters first is their total value compared to the Power/Identity (formerly Loyalty) score of the target Engine: If the total value of my dice is greater than Power, I can change Identity; if the total value of my dice is greater than Identity, I can change Power.

The only place where the number of dice, as opposed to the results, comes into play is for determing Fallout:
If I cancel your dice with a smaller number of my own dice, you take Fallout; but if I cancel your dice with a larger number of my own dice, I take Fallout.

To try an example:

So, say I've staked a 2, 4, 6, 8, 8, which is five dice totalling 28; and you've staked 1, 1, 1, 3, 3, 5, which is six dice totalling 14. I've beat you, because my total is higher. I've got a couple of options now, but let's choose just one, because inventing these examples is brain-breaking....

Say I cancel your 1, 1, 1, 3, 3, 5  -- six dice totalling 14, again -- with my 2, 6, 8 -- three dice totalling 16.
My cancelling dice total a higher value than your dice to be cancelled, so it's a legal move. (If I'd offered my 2, 4, 6, for a total of 12, that would've been too little and you'd have looked at me and said, "nope, try again.")
And I cancelled your six dice with only three of my own, so all six* of your dice become Fallout: They are now staked against whatever Engines they came from in the first place.
Now what do I have left over to actually change the target Engine, which is why I did all this in the first place? Just my 4 and 8 -- the number of dice doesn't matter here, just the total, which is 12.
So if your target Engine is Power 7, Identity 6, then I can spend my 8: That beats either score, but I'll choose to beat Power, so Loyalty goes down to 5.
And now, darn it, I'm stuck: I have my 4 left, but that's less than your Identity of 5, let alone your Power of 7, so I can't do anything else this turn.

* And maybe that's too harsh; maybe it should just be your excess three dice that become Fallout.

Sydney, why do I ever stake dice and not immediately expend them?  Seems to me the only reason is to lie in wait for the other players so I can inflict fallout on them.

Another great question. The answer in my head was "I stake dice and don't expend them because I know I can't achieve the effects I want now -- I'll just get all my dice cancelled out, or take bad Fallout, or not have enough on the left-over dice to change the target Engine -- but maybe if I keep these here until my next turn, I can stake a bunch more and do the cancelling-out process on more favorable terms. Unless the other players respond by staking even more dice on the same Engine and hammer me."

Maybe you get to this point because you thought you had it and then rolled badly, or maybe you went into the roll knowing you were going to get to this point but feeling you had no other option but defer the cancelling-out process until you've gotten more turns and more of your Engines involved. Either way, it's a tactical (and thematic) risk.

As a practical matter, I'm not sure the added dimension of gameplay/storytelling is worth the added complication in the rules. I'm tempted to fall back onto making the cancelling-out process automatic: "are there dice staked on both sides? Then cancel 'em all out until only one side has dice left."

....

And I realize that you said "expend," not "cancel." D'oh. Well, having explained the thing I thought needed explaining, to answer your question:

a) You might stake dice and not immediately expend them because you can't: The other guy has too many dice staked on the other side for you to cancel them all out and still have enough points to overcome the target Engine's Power or Identity.

b) You might stake dice and not expend them because you want to wait until you can do something better: Maybe, after all the cancelling-out is done, you have a 5, for example, and the target's Power is 5 and its Identity is 6 -- which means you can reduce Identity, but not Power -- and maybe you really, really want to save that die until you can stake more dice and whack Power. But that frankly seems like a stupid move to me.

c) Or you might be trolling for Fallout, as you said, and not really interested in affecting the target Engine itself at all, only in luring the other player(s) into throwing more dice against you so you can give them worse Fallout (basically, you'd be pretending to do (b), the dumb "I'll wait" choice). But that's awfully risky: It assumes the other players will stake more dice on their turn, but not enough to cancel out your dice and inflict Fallout on you. Again, I suspect it's a "bunk choice," tactically foolish.

So (a) is the only sensible answer to your question, I suspect.

The thing I'm getting at, Sydney, is that the game starts out with no dice staked anywhere, right?  So the first time anybody initiates conflict, they'll be doing it unopposed.  If they roll high enough, and all the numbers start out low enough that this shouldn't be a problem, there's no reason why they don't expend the dice they just rolled to get their effect.  So as written, I see a lot of people initiating conflict and immediately expending their dice -- I don't see the dice piling up, except in the few cases where somebody gets a crap roll.  Now, the numbers may force this to work in a different way and I'm just not seeing it.

Personally, I was thinking that on your turn you could stake dice, but you couldn't expend them until a later turn.  In terms of the narrative, this would lead to lots and lots of cut scenes, though, and I don't know if that's what you're after.

I keep forgetting to mention that I really really like this new version over the old set of rules.  Much more straightforward!

Ok, I see, thanks.

I keep forgetting to mention that I really really like this new version over the old set of rules.  Much more straightforward!

Joshua wrote: the game starts out with no dice staked anywhere, right?  So the first time anybody initiates conflict, they'll be doing it unopposed....I see a lot of people initiating conflict and immediately expending their dice -- I don't see the dice piling up....

I was thinking that on your turn you could stake dice, but you couldn't expend them until a later turn.  In terms of the narrative, this would lead to lots and lots of cut scenes, though...

Sydney wrote: But my little 2 hangs around, and if I can stake even a single 1 beside it next turn, you're going down.

Another good question. The physical layout of the table should make it clear that the leftover 2 came from such-and-such an Engine, so you can narrate how the lingering, unresolved effects of what that Engine did on the last scene (or ten scenes ago, or whatever) set the target up for good or bad stuff on this scene.

I'm toying with the idea of making up not only blank Engine cards (i.e. "Power [blank]," "Identity [blank]," and "stack dice here," with ideally a different little picture of something evocative on each blank card in the set) but also big "Put Your Staked Dice Here!" arrows that you can cut out and put on the table with the blunt end at the Engine that's staking the dice and the pointy end at the Engine being staked upon. Trying to explain such a physical procedure on top of everything else was too much for me in that draft, though. Do people think such a visual aid would be helpful?

Can't remember if you read my blog, Sydney, but when I was poking at an index-card-based game, I did ponder how to set up relationships between the cards.  (I really like your set up with the relationships as active interactions!)  The thing of it is, any simple system that you set up will hit actual gameplay and complicate like mad.  I wouldn't be at all surprised to see attack cards "mobbing" a target card, leaving trailing dice left over, and then getting involved in other mob-attacks.  The more these pile up, the harder it is to make the web of cards "connect" in any physical way.  So triangle-shaped cards to indicate staked dice would, I think, just add to the confusion on the table.  (However, you might want to playtest with and discard when unweildy, rather than playtest without and wonder if that might have fixed things.)

The cards with Power/Identity/Dice slots would be neat.  If you laminated them, people could write on them with dry-erase markers.

Lamination -- heh -- that's Tony's thing. I was thinking of a bank of blanks you could photocopy and erase. At some point I'll put out a call for art for them.

Back to the mechanics:

One potential problem is that in this system, Fallout is (a) purely negative and (b) only affects the loser (albeit you can lose one exchange of dice and win another). This is very different from my model here, Fallout in Dogs in the Vineyard, in which Fallout is often incurred by both sides of a conflict (albeit never both sides of a given dice-matching exchange) and, more importantly, can have positive effects (albeit mainly at low levels). What's more, the mechanic in the alpha draft of apocalypse girl that "Fallout" replaces, the "Charges" (themselves a direct lift from Capes Inspirations) were also divided between winner and loser, and were purely positive.

Why does this matter? Because if the only side-effect of conflict is harm, then the currency flow discourages conflict, whereas both Dogs and Capes strongly encourage conflict by holding out a very strong possibility of mechanical reward (positive Fallout, small Inspirations) even if you lose. Heck, even D&D gives you XP for killing only some of the monsters before you have to run screaming. I worry without this kind of mechanic that "pays you to fight," apocalypse girl will encourage cautious, defensive play -- which is not what I want at all.

I wrote it this way for simplicity (ha! yes, these rules are pretty baroque right now, I know), but I'm now pondering how to allow for positive Fallout in this system -- which would mean having some cancelled dice, under some conditions, return to their source Engine and be staked for it (i.e. to increase Power/Identity) rather than against it (i.e. to decrease Power/Identity). How to make that work I'm still uncertain, and I'm very open to suggestions.

When a die becomes Fallout, is it then staked from the original target Engine, as well as against its original source?  If so, you could have dice bouncing between Engines several times as Fallout before they got cancelled.

Regarding positive Fallout, what would happen if whenever two sets of different sizes are opposed, the smaller number becomes staked for their sources, as well as the larger number becoming staked against their sources?  If I'm not mistaken it would turn staking a die into a long-term connection between the Engines, since there's fewer ways to discard dice.

Also, could this mechanic work for the old "Schrodinger's War" project, with Engines as historical events, and die-staking as causal connections between them?

Good questions; I don't have good answers yet.

When a die becomes Fallout, is it then staked from the original target Engine, as well as against its original source?

Regarding positive Fallout, what would happen if whenever two sets of different sizes are opposed, the smaller number becomes staked for their sources, as well as the larger number becoming staked against their sources?

Sydney,

On the fallout thing. It proved fun in the previous version of AG when conflict dice went: odd to the winner, even to the loser.  This seems to probably change things a bit, though, in terms of balance.

There's nothing like trying to explain something to another person to make that the idea solidify in your own head -- so my thanks to Anna B. for provoking me into (and being patient with me after) refining [ulr]=http://www.indie-rpgs.com/forum/index.php?topic=18539.msg195614#msg195614]some ideas about the feedback cycles in different kinds of reward systems (http://www.indie-rpgs.com/forum/index.php?topic=18539.msg195614#msg195614).

Applying those ideas to apocalypse girl, I'd say I'm very satisfied with how the current mechanic uses the reward of "introduce new resources into the economy": you only get more dice by a combination of (a) creating new Engines, which means bringing more elements you care about into play, and then (b) building up existing Engines and (c) recharging existing Engines, both of which require you to narrate some kind of character-development bit. Details of the cycle no doubt need fine-tuning, but it should create a nice story arc naturally: Start with just your Core, add more and more elements to the story, accumulate enough dice to win "destructive" conflicts where you start taking your opponents' Engines out of the story, endgame.

What I'm still struggling with is how to use my Dogs-inspired "Fallout" mechanic to encourage players to start conflicts, rather than having Fallout make conflict so dangerous that people don't want to start them. Adapting the structure I laid out for Anna, there are actually two areas, each with its own set of options:

I. What happens to dice "spent" to reduce/increase the Power/Identity of an Engine?

1. They're discarded (the current option). This helps drain dice out of the game's economy, which is almost certainly necessary given the number of ways to introduce new dice. But it's not an incentive to do anything, one way or another.

2. They return to the Engines they were rolled from as "good Fallout," i.e. potentially increasing the Power/Identity of the winner's Engines. Besides not draining dice out of the game, this also creates a really strong positive feedback loop: Every die I spend to take down your Engines or boost my own Engines comes back to me to boost my Engines more. Probably too strong.

3. They return to the Engines they were rolled from as "bad Fallout," i.e. potentially decreasing the Power/Identity of the winner's Engines. This is too weird a negative feedback loop, I think: Every time I win a conflict and accomplish something, it's guaranteed to hurt me, and the more I win the more it hurts.

4. They roll onto the loser's Engines as "bad Fallout." Again, this is giving way too strong a positive feedback to the winner.

II. What happens to the dice that are matched up and cancelled out?

The tricky thing here is that it's not necessarily the winner or the loser of the conflict that's affected by this kind of Dogs-derived mechanic: Whether I win -- i.e. I can eliminate all of your dice and still have dice of my own left over to affect the target Engine -- has very little to do with whether I cancelled your dice with a larger or smaller pool of my own dice. The winner's one advantage is that s/he's the one matching up the dice, so presumably s/he has a slightly better chance to making the Fallout go as desired -- but that chance may be very, very slight.

So, options:

1. The smaller pool of dice is discarded; the larger pool of dice become "bad Fallout" for the Engines they were rolled from. This is the current mechanic. It helps control inflation in the short term because it's a mild dice sink (mild, because the smaller of two groups of dice is discarded, not the larger). In the long term, because Bad Fallout can only make Engines weaker, not stronger, it reduces the number of dice in play.
But is it positive or negative feedback? Well, this kind of Fallout potentially weakens Engines involved in the conflict (which could be either the winner's or the loser's), which is "negative" to the extent that people look at a potential conflict and go, "damn! If this goes badly, even if I win, I could get hurt." But it also gives you another chance to smack the other player, i.e. by making sure the bad Fallout goes away from you and towards him/her -- "dang! If I put in just a few more dice, I could not only win this, I could clobber that other Engine of yours too."
So this strikes me as a system that would make people reluctant to start conflicts, because of the risks, but also encourage them to escalate once they're committed, because of the potential double-punishment for being the loser and the potential double-reward for being the winner.

2. The smaller pool of dice becomes "good Fallout"; the larger pool is discarded. The inverse of the current mechanic. I think it makes it psychologically easier to start a conflict, and also easier to give up instead of escalating: "hey, even if I lose, I might get some dice back to boost my other Engines."
Of course, the fact that the other guy could get more dice to boost his/her Engines should be a disincentive, but psychologically it's just not the same as you yourself getting smacked -- and in fact it's not the same in game terms either, since this is a three player game and not zero-sum (at least in the short run).
Now, is this is a dice sink? It looks like one in the short term, but since Good Fallout actually leads to Engines getting bigger and bigger, it probably makes inflation worse in the long run.

3. The smaller pool becomes "good Fallout"; the larger pool becomes "bad Fallout." There's no dice sink in this version either short-term or long-term.
This is also a tripled feedback loop: Potentially, if everything goes my way, I can (1) change the target Engine, (2) smack your Engines with Bad Fallout, and (3) boost my Engines with Good Fallout. I'd think this would encourage people to be pretty cautious about getting involved in conflicts -- you don't want to give your opponent that 1-2-3 "hat trick" if you can help it -- but very tenacious about sticking with conflicts and escalating rather than giving in.
Plus the massive fallout effects could overshadow the actual target Engine in contention. I suspect this option would create really, really unstable games, often leading to one big blowout conflict (with someone reaching for the Gun out of desperation, probably) and then, BAM!, one side's so crippled and the other so pumped up that the game's basically over.

Those are the three options I can come up with, assuming the current system of "fallout means the dice in question are now staked back on the Engine they came from, either against that Engine to potentially decrease its Power/Identity (bad fallout) or for that Engine to potentially increase its Power/Identity (good fallout)." If dice can become fallout on different Engines than they came from, that opens up a huge mess of options that I can't get my brain around at the moment.

Looking at these two things in conjunction, I wonder if my current combination of options is too strong a disincentive to start conflicts: I-1 plus II-1 basically means you get no extra reward when you win the conflict and change the target Engine, and the fallout along the way can only hurt you.
There may be much more potential in either
   I-2 plus II-1: Winning the conflict gives you (a little) Good Fallout for your participating Engines, the Bad Fallout along the way can hurt either you or your opponent. The potential extra reward for winning encourages people to start conflicts; the potential harm for losing control of the conflict (Bad Fallout) encourages people to stick with conflicts and escalate.
   I-1 plus II-2: Winning the conflict doesn't give you anything extra, but along the way you can get nice Good Fallout for your participating Engines. This feels too nice, though: I'd rather keep the threat of Bad Fallout in the system, because the key idea of Fallout in the first place is to give players the option to put more and more things at risk in order to win a particular goal.
   The other options all seem too prone to flying wildly out of control through excess positive feedback.

Do people think I'm analyzing the incentive structures right here? Does I-2 plus II-1 strike everyone else as the most potentially interesting variant?