Resolution mechanic where players "spend" their rolls to purchase outcomes

[Greywander]

Some time ago I posted this thread (https://forums.giantitp.com/showthread.php?651790-Revisiting-this-core-mechanic-normal-2d10-cautious-3d6-reckless-1d20) taking about a resolution mechanic I'd been considering using as the basis for a system. To quickly recap, it's very similar to D&D, but offers the player several options for how to roll. A normal roll is 2d10, but players can choose to roll "cautiously" with 3d6, or "recklessly" with 1d20. Having crunched the math, a cautious roll is optimal with a target number of 9 or less, a normal roll is optimal with a target number of 10, 11, or 12, and a reckless roll is optimal with a target number of 13 or higher. Someone in that thread suggested a system like this would be more interesting with degrees of success rather than a binary pass/fail system.

I was thinking about this, and the example in my head was of trying to pick a lock. On a high roll you might be able to successfully pick the lock while also making no noise doing so. But what if you don't roll so well? Should you successfully pick the lock but make a noise? Or perhaps the player would rather fail at picking the lock but also not make any noise? Maybe the player could... choose?

So here's the idea I had. When you roll, you get a "currency" of sorts equal to the value of your roll. You then have a list of potential outcomes that you can "purchase" with that currency. Some of these outcomes would be to create some kind of positive effect, while other outcomes would be to mitigate negative effects. Picking the lock would be an example of a positive effect, while making a noise would be an example of a negative effect. By rolling sufficiently high you'll be able to mitigate most or all negative effects while taking the most important positive effects. Or perhaps you might choose not to mitigate a negative effect so that you can afford to purchase an additional positive effect. With a middling or lower roll, you'll have to make some tough choices. Is it better to tank a negative effect in order to get a positive effect you need, or to mitigate as many negative effects as you can and just try again on your next turn?

The downside to this is that it could be difficult to come up with a list of possible outcomes for every roll. Some things can be standardized, like combat and spellcasting, but it could be very taxing for the DM to freeform something not directly covered in the rules. The advantage is that it improves player agency by giving them both a choice before the roll (whether they roll normally, cautiously, or recklessly) and a choice after they roll (what outcomes to purchase), which can make players feel more engaged and make the game feel more fair. Even when they roll poorly, they still have some power to choose how they get screwed over.

Perhaps a way to alleviate the downsides is to have other players at the table come up with a positive or negative outcome (in addition to the default pass/fail) while the DM figures out appropriate prices for each outcome. This helps further increase player engagement by getting other players involved even when someone else is rolling. The DM can come up with some "default" DC value that indicates a standard success with no negative outcomes, then split that value between each individual outcome (e.g. DC 15 to pick a lock, picking the lock costs 10 while not making noise costs 5), so that it doesn't really matter how many negative outcomes are listed, as the overall cost has to be split between them. This way, you could end up with a more slapstick system with lots of potential negative outcomes, e.g. it only costs 2 to hit an enemy with an attack, but it also costs 2 to not break your weapon, not stab yourself, not stab an ally, not throw a weapon away, etc. so that you still have to roll around a 15 or so to hit with no negative outcomes.

One last thing are criticals and fumbles. In the thread linked earlier, I posited that perhaps criticals would occur on a roll of 19 or 20, while fumbles would occur on a roll of 1 or 2. Because of the range of values that can be rolled on 3d6, 2d10, and 1d20, this gives each type of roll a unique interaction with fumbles and criticals (3d6 can't roll either, 2d10 is more likely to crit than to fumble, 1d20 is equally likely and has much higher odds). But honestly, I wonder if critical and fumble rules might be unnecessary. Rolling low already means having to tank a negative outcome, which is basically what fumbling usually is. Rolling high means being able to afford additional positive outcomes, which is what a crit usually is. Fumbles and crits are basically already built into the system and don't need to be explicitly codified.

How does this sound? Is this a terrible idea that will bog down the game? Or is this a major improvement over a standard pass/fail dice roll? Any ways I can streamline this to make it faster and easier to run?

[NichG]

7th Sea 2ed has a system like this. It has some awkward rough edges, but I really like the idea and have stolen bits and pieces of it to use in a more controlled fashion elsewhere.

In 7th Sea, rather than rolling per specific action you take, you choose a general direction per scene and you roll once for the entire scene using the dice pool determined by that choice of direction - e.g. 'I'm going to do athletic things to resolve this scene', 'I'm going to do intellectual things to resolve this scene', etc. The result of the roll gives you a certain number of 'raises', which you can then spend to accomplish things and avoid consequences in the scene. The example in the book is that you're trying to rescue someone from a burning building, but maybe there's also some evidence documents you might want to save. You can spend 1 raise to successfully rescue the person, spend 1 raise to get the documents, but also the fire (which acts sort of like an NPC) might do damage to you a couple of times and each time you could spend a raise to avoid the damage. If the thing you're going to use a raise on doesn't match the direction you've chosen for the scene, you pay double cost.

The awkward bit is that this doesn't really do well with a dynamic back and forth in which you'd naturally want to respond to things in a different way, or things where the action should escalate rather than kind of die off. E.g. if for example you're not just rescuing someone from a burning building, but you're trying to rescue someone when someone else is trying to stop you - well, you sort of just delete each-others' raises I guess? It read a bit flat to me anyhow. So the variation I did was that rather than roll once and spend, you have a series of back and forth bidding war type interactions with any particular adversarial force (either a person using moves and abilities, an environmental condition, etc - they act slightly differently with the environmental things always getting a certain flat number of raises, but that's an aside). In these back and forth bids, as long as the 'aggressor' has something left to try, they can push for a new round of bidding; the defender has to basically respond with some action and meet the aggressor's pool or the aggressor lands whatever consequence it was trying to do (but as a defender, you can always opt to 'take a wound' to increase your results).

Anyhow, where this 'buying off' thing remains is that the system under which characters create moves lets you either buy up the dice pool of the move or (more cheaply) add fixed-cost side-effects that must be bought off or, win or lose, the other side will be impacted by them, but with a restriction that the side-effects are always less serious than the stakes of the interaction (in a tabulated way). So for example if you're trying to kill someone that might have a total Stakes rating of 6 (both a permanent consequence and a severe consequence for +3 each) - that means your side-effects could only be up to Stakes 4, which could be a severe consequence that lasts for just the next bidding round, or it could be a permanent but relatively mild consequence. Both the aggressor and defender can add side-effects, and since its cheaper to scale powers with side-effects than by buying up their core dice pool, a lot of the interplay ends up being which things you decide to let land and which things you need to buy off.

So that's two variations of the idea.

One thing that occurs to me given your reckless/cautious dichotomy could be that rather than just getting a numerical result on a roll, you roll a dice pool where each dice you roll has both a positive contribution towards what you're trying to do and a negative contribution (maybe the dice and its value is handed off to whoever the DM has roll something against the character next as a default setting, but the DM also has the option of narrating a consequence or side-effect). Then after you roll, you can choose to discard some dice. The way it'd work might be that this is like a pool of (stat)d6, but for your bonus you just count how many kept dice are 4,5,or 6 and for your penalty you count the number of kept dice which are even. If you keep at least one 'bonus' point, you can succeed at the thing you were trying to do at a basic level; extra bonus points would be like a crit. But if you keep at least one 'penalty' point then something *also* goes wrong or there's some kind of unfortunate side-effect other than the success/failure of what you were trying to do. Maybe you also can't choose to keep zero dice?

The dynamic of handing off the 4s and 6s to the next person means you could have these riding curse dice that just make a fight more brutal - no one misses an attack, but the cost of that is that the enemy also gets a free hit. Not sure if that's good or bad? Certainly someone could choose to end the curse by taking a miss or by rolling and getting a 5...

[Biggus]

One last thing are criticals and fumbles. In the thread linked earlier, I posited that perhaps criticals would occur on a roll of 19 or 20, while fumbles would occur on a roll of 1 or 2. Because of the range of values that can be rolled on 3d6, 2d10, and 1d20, this gives each type of roll a unique interaction with fumbles and criticals (3d6 can't roll either, 2d10 is more likely to crit than to fumble, 1d20 is equally likely and has much higher odds).


One way to work criticals/fumbles which gives them an even chance for all rolls but keeps the cautious/normal/reckless aspect intact would be to redefine them as "the highest/lowest possible outcome for that dice roll". For 3d6 that would be a 1 in 216 chance of either, for 2d10 a 1 in 100 chance, and for 1d20 a 1 in 20 chance.