Theory: Categorizing non-binary action outcomes

[PhoenixPhyre]

NB For this post, I'm not really worried about mechanics (ie how the system decides which outcomes to pick under what circumstances, dice systems, etc.). This is more thinking about breaking down the categories themselves. Different systems may decide to implement this differently...or not at all.

-------------

In many TTRPGs, one of the fundamental roles of the mechanics is uncertainty resolution. Deciding which event out of several possibilities will actually occur. Does that action succeed? How well does it succeed? In real life, the spectrum of outcomes is extremely multi-dimensional. Some of those are discrete variables (taking on specific values only) and others are continuum variables, taking on an effectively continuous spectrum of results. That's really really hard[1] to do in a TTRPG, at least with humans doing the resolution. So we tend to chunk things into smaller "chunks". Counting successes, rolling vs a TN, even down to reducing it to a binary pass/fail result.
----------
I see 6 primary "bins", at a high level. These have lots of different names, but I'm going to go with descriptors that answer the question "I do <thing>, trying for <success outcome>. Do I get that?"

Yes And. This "bin" of outcomes are where the answer to the question is "yes" (ie you get the result you wanted) but also you get some additional desired thing. E.g. a D&D critical hit. You hit AND you did more damage. Or maybe you knew the answer to the question and got additional information. This bin is only half-bounded--the "additional goodies" might range from a single small thing (a few extra points of damage) to multiple and/or very large things.

Yes. This is your basic binary "yes, you get <success outcome>".

Yes, But. In this bucket, you get what you wanted...but a complication ensued. This might be "success at a cost" (ie you managed to make the jump, but something you had strapped to you fell off. Or you opened the lock but set off an alarm/trap) or it might be a partial success (you wanted the king to send 3 battalions of elite troops, but he agreed to send 1 battalion of regulars + a squad of mages). Better than a failure, but still not exactly what you wanted.

No, But. This is the flip side of Yes But, and the lines are often blurry--you didn't get what you wanted, but it wasn't as bad as it could have been. Or you got something else. The king didn't send any troops at all, but you figured out who the traitor was or got a referral to someone else who could help. This is basically a "softened" failure in many cases. Or, in some cases, it's the best possible result--instead of the tyrant ordering you all executed when you ask him nicely to abdicate the throne to you, a bunch of nobodies, he simply laughs and has you kicked out. Either way, it's still a failure.

No. Your basic binary failure. Whatever the consequences for failure are[2], you get them.

No, and. You not only failed, you made things worse than they would have been otherwise. Instead of just failing to open the lock, now it's jammed and you broke your picks. Or instead of disarming the bomb, you made it tick faster. This is bad news city.
------------

With those bins, it seems that there are a set of possible "check types":

Regular Binary. Some checks just don't have meaningful extensions, usually very straightforward tasks.
"Extended" Binary. These have at least 3 of the four "outer" bins, but don't have the two yes but/no but possibilities. You can either Critically Succeed, Succeed, Fail, or Critically Fail. Stock D&D with crit successes/fumbles plays like this most of the time.
Degrees of Success. This one isn't asking "did it work" but "how well did it work". You only have the 3 "yes..." results. Or maybe only the top two + regular "No." I've seen these most often for "what do I know about X" questions--either you know nothing at all (failure, uncommon) or you know some number of things depending on how good your result was. Often, if the only real consequence for failure would be taking the time to try again, systems will speed this up by some sort of "take 20" rule that costs extra time but assures success.
Degrees of Failure. This one only has the 3 negative options. Here, you're bound to fail and the only reason we have uncertainty to resolve is that we're not sure how badly you've screwed it up. The example of asking the tyrant king to step down to a bunch of nobodies is one of these--no matter what you say, he's going to say no. The only thing to be determined is whether he tries to have you executed, imprisoned, or just laughs and kicks you out/sarcastically tries to hire you as court jesters.
"Full" non-binary. This one has all 6. TBQH, I'm not sure of situations that would call for this, as it gets really hard to distinguish the two "... But" options when both are possible. I guess it's possible though?

Question for anyone who made it this far: are there situations you've encountered that don't fit into one of these 6 buckets or 5 combinations of outcomes? If so, please elaborate. If there are particularly elegant implementations of these non-binary checks out there, I'd love to hear about them as well.

[1] I won't say impossible. Very few things are truly impossible. But I really don't want to have to try to implement such a fully-continuous, multi-dimensional outcome model in a game I'm running or playing in. Too much like my real job.

[2] I hope you have consequences for failure that aren't just "nothing happens, try again"? If not...is there really uncertainty to be meaningfully resolved? Is it really worth even trying to parse these things out at that point? I'd rather just go straight to "you succeed" and skip the pointless rerolls. That's my opinion on the matter, anyway.

[Telok]

Look basically complete. Hits the usual hi points. Reminds me of what I wrote for rolling advice in DtD40k7e some years ago.

Only thing I'd add is that in a rpg you're only rolling to answer a question, and the question usuall falls into 1 of 3: can it be done, how well/badly did it succeed/fail, and how long does it take. There one combo question of can you succeed, by how much, before time runs out.

[Thane of Fife]

There exists out there an (at this point fairly old) RPG called Freeform Universal RPG, or FU RPG. It uses a simple system that would be recognizable as similar to 5e's Advantage/Disadvantage, though FU predates 5e. Basically, you ask the question of what you want to do, tally up the number of advantages or disadvantages you have in the situation, roll the difference in d6's (+1), and take either the best or worst die as the result, where the dice faces correspond directly to the six outcomes you posit above.

So, for example, Jessica d'Uniac is racing to Paris to stop a Spanish assassin from attacking the Cardinal. The question is "Can she get there in time to intercept the assassin?" She is an Expert Rider with a Loyal Steed, but due to earlier misadventures, she is Wanted by the Law. That's two advantages and one disadvantage, so a net of +1. She rolls two dice and takes the better of them - in this case, she gets a 2 and a 3. Three is "No" and 2 is "Yes, but...," so the result is "Yes, but...." We decide she rushes onto the scene in the nick of time and tackles the Cardinal out of the path of the assassin's shot. For the "but," perhaps she takes the bullet herself and is now Wounded.

In my experience with FU, the only questions that are hard to answer while allowing all six options are ones where you're already assuming one of the results. "Can I defeat the orcs?" is easy to answer. "Can I defeat the orcs without getting injured?" is where it gets harder, because you're kind of answering two questions at once. I don't agree that "Yes, but" and "No, but are hard to distinguish from each other.

[PhoenixPhyre]

There exists out there an (at this point fairly old) RPG called Freeform Universal RPG, or FU RPG. It uses a simple system that would be recognizable as similar to 5e's Advantage/Disadvantage, though FU predates 5e. Basically, you ask the question of what you want to do, tally up the number of advantages or disadvantages you have in the situation, roll the difference in d6's (+1), and take either the best or worst die as the result, where the dice faces correspond directly to the six outcomes you posit above.

So, for example, Jessica d'Uniac is racing to Paris to stop a Spanish assassin from attacking the Cardinal. The question is "Can she get there in time to intercept the assassin?" She is an Expert Rider with a Loyal Steed, but due to earlier misadventures, she is Wanted by the Law. That's two advantages and one disadvantage, so a net of +1. She rolls two dice and takes the better of them - in this case, she gets a 2 and a 3. Three is "No" and 2 is "Yes, but...," so the result is "Yes, but...." We decide she rushes onto the scene in the nick of time and tackles the Cardinal out of the path of the assassin's shot. For the "but," perhaps she takes the bullet herself and is now Wounded.

In my experience with FU, the only questions that are hard to answer while allowing all six options are ones where you're already assuming one of the results. "Can I defeat the orcs?" is easy to answer. "Can I defeat the orcs without getting injured?" is where it gets harder, because you're kind of answering two questions at once. I don't agree that "Yes, but" and "No, but are hard to distinguish from each other.

I guess it depends on how specific you get defining what "success" and "failure" mean. And who gets to decide them. The more specific the desired outcome, the harder it gets (I think?) to divide "got what I wanted but not entirely" and "didn't get what I wanted but got part of it".

Maybe some of it also depends on how the framing is set? Are you resolving things at the scene level (get there, get the Cardinal out of the way), where the underlying uncertainty being resolved is broad and complex? If so, it's easier to have "central" goals (the Cardinal is safe, therefore success) and secondary goals (without me being injured, therefore partial success or complication). Or are you operating at a much more narrow, task level, where those complex scenes are made up of potentially many independent actions (some of which may be uncertain and some may not be), and so success/failure are more "atomic" and it's harder to have a single resolution that shows the difference.

To be honest, most of my TTRPG experience is with more "task" oriented games, rather than the more freeform ones (which I have a loose notion are more "scene" oriented). So that's why, for me, those two middle ones can be hard to distinguish.

Thanks for the different take!

[NichG]

NB For this post, I'm not really worried about mechanics (ie how the system decides which outcomes to pick under what circumstances, dice systems, etc.). This is more thinking about breaking down the categories themselves. Different systems may decide to implement this differently...or not at all.

-------------

In many TTRPGs, one of the fundamental roles of the mechanics is uncertainty resolution. Deciding which event out of several possibilities will actually occur. Does that action succeed? How well does it succeed? In real life, the spectrum of outcomes is extremely multi-dimensional. Some of those are discrete variables (taking on specific values only) and others are continuum variables, taking on an effectively continuous spectrum of results. That's really really hard[1] to do in a TTRPG, at least with humans doing the resolution. So we tend to chunk things into smaller "chunks". Counting successes, rolling vs a TN, even down to reducing it to a binary pass/fail result.
----------
I see 6 primary "bins", at a high level. These have lots of different names, but I'm going to go with descriptors that answer the question "I do <thing>, trying for <success outcome>. Do I get that?"

Yes And. This "bin" of outcomes are where the answer to the question is "yes" (ie you get the result you wanted) but also you get some additional desired thing. E.g. a D&D critical hit. You hit AND you did more damage. Or maybe you knew the answer to the question and got additional information. This bin is only half-bounded--the "additional goodies" might range from a single small thing (a few extra points of damage) to multiple and/or very large things.

Yes. This is your basic binary "yes, you get <success outcome>".

Yes, But. In this bucket, you get what you wanted...but a complication ensued. This might be "success at a cost" (ie you managed to make the jump, but something you had strapped to you fell off. Or you opened the lock but set off an alarm/trap) or it might be a partial success (you wanted the king to send 3 battalions of elite troops, but he agreed to send 1 battalion of regulars + a squad of mages). Better than a failure, but still not exactly what you wanted.

No, But. This is the flip side of Yes But, and the lines are often blurry--you didn't get what you wanted, but it wasn't as bad as it could have been. Or you got something else. The king didn't send any troops at all, but you figured out who the traitor was or got a referral to someone else who could help. This is basically a "softened" failure in many cases. Or, in some cases, it's the best possible result--instead of the tyrant ordering you all executed when you ask him nicely to abdicate the throne to you, a bunch of nobodies, he simply laughs and has you kicked out. Either way, it's still a failure.

No. Your basic binary failure. Whatever the consequences for failure are[2], you get them.

No, and. You not only failed, you made things worse than they would have been otherwise. Instead of just failing to open the lock, now it's jammed and you broke your picks. Or instead of disarming the bomb, you made it tick faster. This is bad news city.
------------

With those bins, it seems that there are a set of possible "check types":

Regular Binary. Some checks just don't have meaningful extensions, usually very straightforward tasks.
"Extended" Binary. These have at least 3 of the four "outer" bins, but don't have the two yes but/no but possibilities. You can either Critically Succeed, Succeed, Fail, or Critically Fail. Stock D&D with crit successes/fumbles plays like this most of the time.
Degrees of Success. This one isn't asking "did it work" but "how well did it work". You only have the 3 "yes..." results. Or maybe only the top two + regular "No." I've seen these most often for "what do I know about X" questions--either you know nothing at all (failure, uncommon) or you know some number of things depending on how good your result was. Often, if the only real consequence for failure would be taking the time to try again, systems will speed this up by some sort of "take 20" rule that costs extra time but assures success.
Degrees of Failure. This one only has the 3 negative options. Here, you're bound to fail and the only reason we have uncertainty to resolve is that we're not sure how badly you've screwed it up. The example of asking the tyrant king to step down to a bunch of nobodies is one of these--no matter what you say, he's going to say no. The only thing to be determined is whether he tries to have you executed, imprisoned, or just laughs and kicks you out/sarcastically tries to hire you as court jesters.
"Full" non-binary. This one has all 6. TBQH, I'm not sure of situations that would call for this, as it gets really hard to distinguish the two "... But" options when both are possible. I guess it's possible though?

Question for anyone who made it this far: are there situations you've encountered that don't fit into one of these 6 buckets or 5 combinations of outcomes? If so, please elaborate. If there are particularly elegant implementations of these non-binary checks out there, I'd love to hear about them as well.

[1] I won't say impossible. Very few things are truly impossible. But I really don't want to have to try to implement such a fully-continuous, multi-dimensional outcome model in a game I'm running or playing in. Too much like my real job.

[2] I hope you have consequences for failure that aren't just "nothing happens, try again"? If not...is there really uncertainty to be meaningfully resolved? Is it really worth even trying to parse these things out at that point? I'd rather just go straight to "you succeed" and skip the pointless rerolls. That's my opinion on the matter, anyway.

So first something from Don't Rest Your Head has a resolution method where when you determine success or failure, you also determine primacy between four thematic forces of which dominates in the scene - discipline, exhaustion, madness, and pain. There are things you can do to increase your chances of success but which also increase the chance that madness dominates, which has longer term consequences. So it sort of is a 'yes, but' thing, except that there's never not one of the themes dominating so its more like 'yes, how?'.

Next something from 7th Sea 2ed. Here when you roll it gives you a pool of 'raises' that you can spend over the course of a scene to buy off bad things that happen or buy into good opportunities. So its sort of a 'degrees of success' kind of thing, but where it avoids the difficulty of 'how is it different to critically succeed at opening a lock?' issues with binding degree of success to a more atomic action. Instead you're rolling on broad approaches to the scene - are you moving quickly or fighting or being highly perceptive or what? And so in that case, a high roll lets you be free to choose to take advantage of more opportunities later on, whereas a low roll means you have to be much more mission-focused if you want to actually accomplish your main goal.

Now we're getting more into my own personal homebrew, but I rather like things along the line of 'you can always choose whether to succeed or fail, and the dice just determine the current cost'. I ran a few systems where characters have pools of points they can spend rather than having things like 'hitpoints' that are reduced by external action. You can always spend points from your pools to boost any defensive rolls, but most 'heavy' attacks or effects will have serious consequences if they're allowed to hit - one or two-shotting most characters basically - while there are also 'light' effects that are cheap to make but have less serious effects (but are just as expensive to 'block' as the heavy attacks). So when you roll a defense, you're not rolling to see if you block (and if you fail you take some ablative damage), you're usually rolling to determine how much you're going to have to pay to not be one-shot by the attack. I initially went this direction so you could have very hypercompetent characters would could still be killed by e.g. a dagger to the eye - they can basically ensure that they 100% dodge every hit for longer than less competent characters, but if for some reason they choose not to defend (or aren't allowed to defend), they die just like anyone else. But then I liked the structure of 'roll to determine cost'. My current design project is going to use a sort of hybrid system here - its roll and keep like 7th Sea, but you can increase your kept dice after rolling by spending Stamina or Focus up to a maximum of your rolled dice. So it has one extreme end of 'I succeed for free', one extreme end of 'no matter how much I pay I can't succeed', and a wide middle of 'I will probably have to pay from 1 to 3 Stamina or Focus to make this thing go, but its my choice between failure or cost'.

If we go into a more extreme end of design space, I can imagine outcomes that work by drawing tarot cards and interpreting the card in the context of the attempted action. Or, less extreme, consider something like wild magic charts in D&D or other things where an action provokes the generation of some sort of open-ended scene aspect. There you're rolling for more than just degree of success or failure - a different roll might be the difference between becoming giant-sized or being teleported to the nearest city or having your left arm permanently become 3 inches longer than your right arm or randomly inheriting property from a relative who retroactively was brought into existence or ... In those cases, to better integrate the player's intent and purpose, you could have mechanics that let the player choose between multiple random outcomes (so with more choices, they can try to get randomness they can do something useful with); or ones that let the player nudge the result up or down the chart by some amount at a cost, or other such things. In that case, the answer of the resolution system isn't directly 'you succeed' or 'you fail', its more like 'what do you have to work with?'. In that sense its kind of like the randomness in deckbuilding/card drawing games...

[Thane of Fife]

I guess it depends on how specific you get defining what "success" and "failure" mean. And who gets to decide them. The more specific the desired outcome, the harder it gets (I think?) to divide "got what I wanted but not entirely" and "didn't get what I wanted but got part of it".

Personally, I think of it less as getting more or less of you want, and more as you sometimes get extra (which may be good or bad). So, in the d'Uniac example, to me, it's less that she got there with less time to spare because of the "but," and more that she got there in time (which was what she wanted), but she also get shot (that's the extra).


Maybe some of it also depends on how the framing is set? Are you resolving things at the scene level (get there, get the Cardinal out of the way), where the underlying uncertainty being resolved is broad and complex? If so, it's easier to have "central" goals (the Cardinal is safe, therefore success) and secondary goals (without me being injured, therefore partial success or complication). Or are you operating at a much more narrow, task level, where those complex scenes are made up of potentially many independent actions (some of which may be uncertain and some may not be), and so success/failure are more "atomic" and it's harder to have a single resolution that shows the difference.

In a lot of cases, the "but" or "and" can just come down to a modifier to some future action. Can Ragnar hit the orc with his sword? Yes or no determines if he deals damage, and the "and" or "but" may result in one side or the other getting a bonus to their next attack (because the other person got stunned, tripped, disarmed, whatever). Or being pushed back a short distance.

Can Mialee dominate the giant's mind? Yes or no might give us the answer, and maybe the "and" or "but" says she has to spend more or less of her magic power to do it. Maybe the giant resists her control and fights less effectively, or can no longer feel pain and fights more effectively. Maybe the giant resists being controlled but is still dazed a bit, or maybe Mialee gets hit by some kind of magic backlash that dazes her a bit.


To be honest, most of my TTRPG experience is with more "task" oriented games, rather than the more freeform ones (which I have a loose notion are more "scene" oriented). So that's why, for me, those two middle ones can be hard to distinguish.

I'm not sure I would describe those as freeform, but there are definitely differences between RPGs that use rolls that cover some amount of time (e.g. this roll represents one time unit of fighting) vs those where a roll covers things until they change (e.g. this roll represents all of the fight up until a meaningful change occurs).

[Pauly]

There is a 7th outcome. A null outcome, in that the PC’s action did nothing to alter the status quo. From a gaming/story telling perspective thus is a boring outcome so most systems fast forward to the point where the PC’s action falls into a success or failure bin.

[Vahnavoi]

I don't think it's fully possible to ignore the actual resolution method - whether the resolution method is analog or digital seems directly relevant to the distinctions you're making. This becomes more apparent when you move beyond dice- and number-based resolution systems.

Furthermore, systems that are binary on level of individual checks can be shown to be able to emulate more complex systems via strings of checks. The same method can be used to show that in a lot systems, null or "nothing happens" results are illusory.

A simple example from turn-bases combat: a check is binary between "hit" or "do nothing". But every "do nothing" result passes the turn to the opponent, giving them the opportunity to make a counter-attack. So as turns pass, all named types of result can happen as combinations of check results.

As a side note, the distinction between "yes, but..." and "no, but..." becomes easier to conceptualize when you break a decision to its components: what, how and why (and when and where and so on). The "yes" or "no" applies to "What", the "but..." applies to "how" or "why". So, in the simple combat scenario, the "what" is "hit the enemy", the "how" may be "with a sword in close combat" and the "why" may be "to stop their advance". So "yes, but..." could be "yes you hit, but the enemy is still advancing", and "no, but..." could be "no, you missed, but you manage to close a door on them". So on and so forth.

Of course, once you make this distinction, similar to point made by NichG, you can break down each of the six categories into numerous sub-categories based on which components of decision-making they weigh.

[kyoryu]

It's workable.

I think it's broad enough that most things can be coerced into one of the categories - for instance the "nothing changes the status quo" can be thought of as a "flat no" category. Things like "you make progress" can be thought of as a "yes", "yes, but" or "no, but", but feel different than any of the above. But I mean, that's how models work, right? Like, in general, you can coerce things to fit into any sufficiently broad model, and that says more about you than the model.

[truemane]

It's an interesting exercise. And a fairly complete framework. I think you're going to run into a lot of trouble once you try to go granular with this and try to make it 'mean' something within the confines of a narrative or system. Dice get rolled for so many reasons, in so many circumstances, for a dizzying array of mechanical, narrative, thematic, philosophical reasons. What are dice even for? What are rules for? Are we emulating or simulating? To what degree does the desired experience dictate the dice rolls and in what ways the reverse?

It's a real muddle. And lots of games, in my experience, don't even shoe evidence that they've even considered all the ways that dice and rules and story intersect.

Anyway. I'm not sure if it counts as elegant, but the FFG Star Wars system uses all six buckets. A given dice can return one or more each of: Success, Advantage, Failure, Threat, Triumph, or Despair. Successes cancel Failures, Advantages cancel Threats. You add them all up and work with the net result. So you can succeed with Threats, fail with Advantages, and/or any combination of those with Despair and/or Triumph.

So you can have any result from "Yes, And" to "No, And."

Also, Powered by the Apocalypse (PtbA) games offer: "Yes", "Yes, But" and "No, And." You could even say that the "Yes" is actually a "Yes, And" and there is no straight "Yes." Depends on how you define your terms.

For me the key tension point is the degree to which the game system proscribes the results of a non-binary check, or expects me to make it up. One of the complaints about the FFG Star Wars system is that sometimes you just want to know if you hit the guy, and you don't want to have to create a whole story to explain Every. Single. Roll.

PtbA games deal with this by restricting dice rolls to only circumstances where the result will saliently impact the fiction, and then each roll tells you what happens for each result. PtbA games don't really have rolls to passively gather information, or perform mundane tasks. And there's no such thing as a roll that just "fails" and so nothing happens. Any time you roll dice in a PtbA game, something happens.

[aimlessPolymath]

Question for anyone who made it this far: are there situations you've encountered that don't fit into one of these 6 buckets or 5 combinations of outcomes? If so, please elaborate. If there are particularly elegant implementations of these non-binary checks out there, I'd love to hear about them as well.

I'd like to throw out various PbtA 'ritual' moves. Taking the following example, from the Dungeon World wizard:

Ritual

When you draw on a place of power to create a magical effect, tell the GM what you’re trying to achieve. Ritual effects are always possible, but the GM will give you one to four of the following conditions:

It’s going to take days/weeks/months.
First you must .
You’ll need help from .
It will require a lot of money
The best you can do is a lesser version, unreliable and limited
You and your allies will risk danger from .
You’ll have to disenchant to do it.


(This is an example of the general category; if the scope of the ability were narrowed from 'you can do anything', I think GM adjudication could get replaced by a die roll. I don't have an example of that offhand, though)

This is a variant in which the only* possible result is "Yes, but...", in which the "degree of success" is how many requirements or restrictions there are in order to carry out the original proposed effect.


I'd also like to toss out Schema as an example of the full non-binary. In it, your dice pool is used to 'purchase' success plus extra benefits (by spending your 5's and 6's) as well as buy off negative consequences (by spending your 1's and 2's).

*Admittedly, in one case, "no, but...". I think the broader point is made, though.

[Witty Username]

This may fall under 'Yes' but there is also progress.
Where a check represents progress on a longer-term project rather than true success.

Complex skill checks from 3.5 D&D, and Clocks in Blades in the Dark get in on this. HP and damage as well if you are willing to squint at it.

[PhoenixPhyre]

This may fall under 'Yes' but there is also progress.
Where a check represents progress on a longer-term project rather than true success.

Complex skill checks from 3.5 D&D, and Clocks in Blades in the Dark get in on this. HP and damage as well if you are willing to squint at it.

I'd only really be willing to call the entire multi step process a single "task", just one that needs several rolls. And even then it feels a bit strained. Any task the can only have one outcome doesn't feel uncertain enough to worry about. Now if it's time sensitive and success/greater success makes things go faster and/or failure makes you lose time, then yeah. That's generally something like degrees of success. If time doesn't really matter, then it's just as clock that ticks up.

Or Skill Challenges from 4e, where it's a single (in that case binary yes/no) task with multiple resolution steps.

Speaking of which, I think you could greatly improve Skill Challenges by making them dynamic. Each "turn", success or failure, has some kind of consequence that changes the nature of the subsequent checks. Things like "you can use Intimidate here for an easy success (lower DC), but that's going to make the guy hate you so subsequent checks that try to be all nice and kind will just fail." Or "you can push over that statue, making it easier to bar the door, but failing is going to throw your back out a bit so further physical stunts will have a penalty for a bit." And you can include non binary outcomes--if you succeed at doing 3 of the 4 subtasks but then failed enough to fail the check, you get a no but situation. Or conversely, if you just barely scraped by, you might get a yes but, while if you blew it away, you might get yes and (or vice versa for a really crappy job).

The beauty of having relatively small, discrete steps is that you can emulate complex resolutions without being tied to them for simple things. Call it emergent depth.

But I'm rambling now.

[Telok]

Something I do for long duration tasks using a single skill/stat is a number of points over a target number.

Recently it was hauling an unconsious big critter into a cage before its mate got to the boat. Thing was size 12 so target number 12 athletics+strength, which is not difficult for an average str & no training person to hit at a bit over 50% success, but not trivial to move a wet furry 14 foot long critter. Then needing 60 over the 12, because the system uses stat*5 for a range of stuff and it works out well. First person rolled a 24 so 24-12=12 and 60-12=48 more to go. Hauled about 1/5th or the head into the cage. After a couple turns of people rolling 8-28 range results the mate got there before they caged it. It would't actually matter unless the boat capsized (which it didn't), but the players were nervous and excited.

Just a nice way to represent a larger than a one-&-done task in mid combat. It really shines with things like opening huge doors or raising a drawbridge while others hold off a horde. Gives a nice decision point on if the strong melee types fight or pull, because weaker characters won't reliably deliver the bigger numbers over the target number. Works better with bell curve systems than flat distributions, but the method doesn't fail under any particular dice system.

[Witty Username]

And even then it feels a bit strained. Any task the can only have one outcome doesn't feel uncertain enough to worry about. Now if it's time sensitive and success/greater success makes things go faster and/or failure makes you lose time, then yeah. That's generally something like degrees of success. If time doesn't really matter, then it's just as clock that ticks up.


That's fair, the times I have seen stuff like this used successfully, time sensitivity is a significant factor. Either in long term like good vs bad represents up to a week of down time. Or short term, enabling an action that wouldn't otherwise be resonable.