What exactly should a Roll Represent?

[kyoryu]

Look at NBA Free Throws then, or NFL field goals. They both have a strict success/fail binary (missing wide right gives you the same number of points as bouncing off the post), and even the pros miss them all the time. I only used bowling as an example because I've actually gone bowling myself.

I've never made a field goal myself, but as a viewer I can confirm that good kickers miss easy field goals all the time.

So let's get clear. Yes, people fail tasks. That's a thing that happens.

Since we're playing games that are, in general, about cool people doing cool things, I generally prefer to have them fail due to external factors rather than simply bumbling. I think it goes over better with players, and makes sessions more fun.

Sometimes you can't. And, honestly, field goals and free throws are pretty much that situation.

This line specifically is interesting. Maybe I'm just being pedantic but would something routine have a 50% chance of failure? I guess it does depend on the system, in DnD you can just set a trivially low DC, but in Apocalypse world you don't have that ability. I would argue though, that if you're attempting a routine task it probably doesn't justify a roll. You can just succeed.

I'm not talking about making things easier. I'm talking about how we narrate failure. No more, no less. I'm not advocating for fewer rolls, or not making failure an option. I'm just saying that, in general, I prefer to narrate failure as due to some kind of external factor, rather than "haha you suck". My line here is more about when I start to think that "nope, it was just hard, you didn't do it" becomes more acceptable to me, and when I prefer to lean on "wow, you totally would have had it but..."

And, yes, even for a 50/50 I prefer to narrate failures as external rather than a lack of competence.

[NichG]

So let's get clear. Yes, people fail tasks. That's a thing that happens.

Since we're playing games that are, in general, about cool people doing cool things, I generally prefer to have them fail due to external factors rather than simply bumbling. I think it goes over better with players, and makes sessions more fun.

Sometimes you can't. And, honestly, field goals and free throws are pretty much that situation.



I'm not talking about making things easier. I'm talking about how we narrate failure. No more, no less. I'm not advocating for fewer rolls, or not making failure an option. I'm just saying that, in general, I prefer to narrate failure as due to some kind of external factor, rather than "haha you suck". My line here is more about when I start to think that "nope, it was just hard, you didn't do it" becomes more acceptable to me, and when I prefer to lean on "wow, you totally would have had it but..."

And, yes, even for a 50/50 I prefer to narrate failures as external rather than a lack of competence.

This seems to me more of something you need when there's a mismatch between what the odds are, and what the odds feel like they should be.

Like, if instead you think of things with a 20% chance of failure as challenging (and actually adjust the mechanics to align that with the character fiction), would it still feel necessary to use external explanations?

Basically the odd thing for me is that you're describing 50/50 cases narratively as things a character is good at, where for me 50/50 reads as quite difficult. Like, if I'm jumping across a stream or gap between rocks while hiking, I just wouldn't even try it if I felt less than 99% sure I could do it successfully. A jump I actually had only a 80% chance to pull off would be 'perilous'.

[Quertus]

So let's get clear. Yes, people fail tasks. That's a thing that happens.

Since we're playing games that are, in general, about cool people doing cool things, I generally prefer to have them fail due to external factors rather than simply bumbling. I think it goes over better with players, and makes sessions more fun.

Sometimes you can't. And, honestly, field goals and free throws are pretty much that situation.


And, yes, even for a 50/50 I prefer to narrate failures as external rather than a lack of competence.

If Hawkeye / Bullseye / whoever you consider “really cool” only has a 50/50 chance of hitting the target? That’d better be an eye slit in a 500’ tall giant in a Hurricane, else the rules have Captain Hobo‘d the character’s coolness. And, if the rules haven’t Captain Hobo’d the character’s coolness, then you don’t need to narrate extra reasons why they missed - they missed because they were attacking an eye slit 500’ up in a Hurricane.

Also, if you’re playing with players who try to flip a coin, need heads but get tails, and need some “external factor” excuse for why they failed? That’s a failure to understand probability, or terrible self-esteem issues. Play with better players.

[kyoryu]

This seems to me more of something you need when there's a mismatch between what the odds are, and what the odds feel like they should be.

Like, if instead you think of things with a 20% chance of failure as challenging (and actually adjust the mechanics to align that with the character fiction), would it still feel necessary to use external explanations?

Basically the odd thing for me is that you're describing 50/50 cases narratively as things a character is good at, where for me 50/50 reads as quite difficult. Like, if I'm jumping across a stream or gap between rocks while hiking, I just wouldn't even try it if I felt less than 99% sure I could do it successfully. A jump I actually had only a 80% chance to pull off would be 'perilous'.

But.... like, why?

What is the advantage of narrating as "you just can't?"

Why is "you try to cross the stream but miss the rock" better than "you try to cross the stream but slip on a wet and mossy part of the rock?" Why is "you shoot an arrow at the guy and miss wildly" better than "you shoot an arrow at the guy but he moves at the last second?"

What's the advantage?

[BRC]

But.... like, why?

What is the advantage of narrating as "you just can't?"

Why is "you try to cross the stream but miss the rock" better than "you try to cross the stream but slip on a wet and mossy part of the rock?" Why is "you shoot an arrow at the guy and miss wildly" better than "you shoot an arrow at the guy but he moves at the last second?"

What's the advantage?

The main reason is efficiency.

Combat often means a lot of rolls. If your Master Archer has a string of bad luck and keeps missing, it can get tedious to have the GM keep saying "The Arrow flies true but he brings up his shield at the last second" and "He turns his head at the right moment and the arrow soars past" ect ect. The issue with the External Factors approach is that you need to keep coming up with new ones each time.


"You just happened to miss" is kind of the default narrative, and the Null Approach of not providing a narrative explanation, just saying "You fail" or "You hit" will kind of default to that, technically. You don't get to add any flavor and the game can quickly become divorced from the narrative, but not elaborating also doesn't rub in "Your master archer missed the shot" the way a more elaborate explanation does.

[NichG]

But.... like, why?

What is the advantage of narrating as "you just can't?"

Why is "you try to cross the stream but miss the rock" better than "you try to cross the stream but slip on a wet and mossy part of the rock?" Why is "you shoot an arrow at the guy and miss wildly" better than "you shoot an arrow at the guy but he moves at the last second?"

What's the advantage?

I mean, one aspect has nothing even to do with how you choose to narrate, but is rather about noticing that the system isn't reinforcing the fiction that its trying to represent well. If you find yourself rolling for things where failure due to the character 'just not pulling it off' would feel like it depicts that character's skill incorrectly, that's a good sign that the system is misaligned with the fiction. It's saying in one place 'this person is more athletic in all ways than the world's top olympian athletes' and in another place 'this person has a 10% chance of messing up a long jump that is easily within Olympic records'. So absent everything else, I don't want to ignore that signal of misalignment - I'd rather adapt the system to the fiction or the fiction to the system so that in the long run there's lower overall dissonance. Or even adapt how I'm presenting situations - if the rules demand a 50/50 chance for metagame reasons, I should be narrating the jumps as being longer in the first place if it feels weird for that character to have a 50/50 chance for that jump.

Its not even that using an external factor now and then to rescue a really inane result is going to be terrible. That's a pragmatic solution to something going wrong that you hadn't anticipated, which is going to be necessary now and then. But in the long run, I would not want to be satisfied with that as a continual state of operation. Because as an ongoing thing, if that's the normal way things are resolved, it introduces these weird and predictable metagame dependencies which I really would only want to have if I were specifically running a 4th wall breaking campaign. Namely that because characters of differing skill (narrative) levels choosing to attempt a thing require different external consequences to explain away a success or failure, it is possible to influence the state of the external world outside of the instance of a single roll by choosing who makes the roll and when to make it. If external factors are only used inconsistently for explanation, that's not really an issue at all, and the pragmatics win. But if it becomes the standard of how the system is run, then that sort of effect potentially becomes an ongoing consideration from the point of view of the players, and that's (again, not game-breaking in any sense) a bit less good than the alternative of not having that dependency exist.

So basically, to me its a small flaw unless I specifically want to telegraph that kind of metagame consideration. If I can fix that flaw by better aligning the system with the fiction, I'd prefer to fix it than to lean into the flaw and build around it.

And I do think having something that feels like the person should succeed 99% of the time be a 50/50 chance because the rules say so is actually more of a major flaw, but again that's independent of the question of how you narrate it. But I would worry that if I find myself needing to narrate external factors, I might be blind to that sort of more extreme dissonance.

[Quertus]

The main reason is efficiency.

Combat often means a lot of rolls. If your Master Archer has a string of bad luck and keeps missing, it can get tedious to have the GM keep saying "The Arrow flies true but he brings up his shield at the last second" and "He turns his head at the right moment and the arrow soars past" ect ect. The issue with the External Factors approach is that you need to keep coming up with new ones each time.


"You just happened to miss" is kind of the default narrative, and the Null Approach of not providing a narrative explanation, just saying "You fail" or "You hit" will kind of default to that, technically. You don't get to add any flavor and the game can quickly become divorced from the narrative, but not elaborating also doesn't rub in "Your master archer missed the shot" the way a more elaborate explanation does.

Very much this. I’m in the “don’t waste my time with purple prose” camp, in the “let’s handle this in (close to) the most efficient manner possible” camp. Now, if Hawkeye and Bullseye both miss, but Aunt May happens to smack the villain with her purse, and the table has a skilled storyteller, and said storyteller happens to see some great moment here? Then, yeah, they’re welcome to slow things down with a, “guys, do you realize that…” moment.

Or even if someone has a “wait… didn’t we put the McGuffin / nitroglycerin / alien squeaky toy / whatever in Aunt May’s purse?” That’s an acceptable time to slow down, too.

[BRC]

Very much this. I’m in the “don’t waste my time with purple prose” camp, in the “let’s handle this in (close to) the most efficient manner possible” camp. Now, if Hawkeye and Bullseye both miss, but Aunt May happens to smack the villain with her purse, and the table has a skilled storyteller, and said storyteller happens to see some great moment here? Then, yeah, they’re welcome to slow things down with a, “guys, do you realize that…” moment.

Or even if someone has a “wait… didn’t we put the McGuffin / nitroglycerin / alien squeaky toy / whatever in Aunt May’s purse?” That’s an acceptable time to slow down, too.

Even if you don't want to take the Null approach, the "You missed" approach is a lot more efficient.
"You don't quite make the shot" or "The arrow doesn't penetrate their armor" is a complete statement on it's own. Coming up with external factors needs a little more elaboration and creativity.

[gbaji]

But.... like, why?

What is the advantage of narrating as "you just can't?"

Why is "you try to cross the stream but miss the rock" better than "you try to cross the stream but slip on a wet and mossy part of the rock?" Why is "you shoot an arrow at the guy and miss wildly" better than "you shoot an arrow at the guy but he moves at the last second?"

What's the advantage?

I think the big problem with narrating failure as an environmental factor external to the character rather than the character just failing is that it more or less curbstomps on the fact that most gaming systems have different levels of skill for characters (written on their sheets even!). I get that you're just saying to narrate things this way, but this suggests to us the somewhat absurd assumption that the guy with rank1 in archery is just as good at hitting as someone with rank5 in archery, but his targets just move more often, or things randomly get in the way more often because... reasons.

It also somewhat reduces the value of skill increases as well (at least narratively). I guess this somewhat depends on how you view skills in a game. You could assume that everyone is incredibly skilled and capable of always succeeding unless something happens that causes failure, and that higher skill represents being able to better account for environmental factors (what you're basically arguing for). Or we could assume that people start out being really terrible at what they are attempting to do, and then become gradually more competent over time and even good enough to counteract environmental factors as well. Both are viable methods, but mechanically most games allow the GM to apply some sort of difficulty factor (usually an adjustment to the target you have to roll, right?). Um... That's what's supposed to represent environmental factors, so using your method, you basically have two mechanisms that do more or less the same thing.


The first method also fails to allow for cases where people are just starting out, or not terribly good at something. What if you want to play a character who's never picked up a sword in his life, usually runs and hides, whatever. But this character is thrust into an environment where he has to fight. He's got to start somewhere, right? So he's going to suck at first. Missing a lot, just because he's not that good. But over time, if he survives, and perhaps with some assistance from his friends, he gets better. A skill system where low skill really just represents "low skill", and getting better means you actually get better (even at basic stuff), allows for this character development over time. Assuming that anyone with any skill at all is "perfect", and will always succeed unless something external causes them to fail somewhat steps on this as a game play mechanism.

I happen to think that the "you start out pretty terrible at things, and that's what low skill levels represent" is a more realistic way of running the game. But that's just my opinion. And frankly, I think that better matches skill development and progression in the "real world". There's a reason why boxers hit bags over and over, basketball players practice shooting over and over, baseball players practice hitting balls over and over, etc. You don't start out being able to hit a static bag cleanly and powerfully at all at first. You don't start out being able to get a ball in the basket even semi-regularly at all at first. And yeah, even just hitting a ball on a T is difficult when first starting out. A skill system that starts at the bottom and then progresses better simulates this than the other way around.

[Quertus]

Combining those last two posts, when Hawkeye misses his 50/50 shot at the eye slit of the 500’ giant in a Hurricane, “it bounced off the armored helmet”; when Aunt May missed her 0% success rate at the same shot, “the arrow missed the giant entirely”.

EDIT: and, again, the external factors of “it’s an eye slit, 500’ off the ground” and “there’s a bloody Hurricane” have already been narrated.

[NichG]

Funnily enough, I don't really agree with the reasons given by the people who are agreeing with me on the conclusion here. But I guess that happens.

[Quertus]

Funnily enough, I don't really agree with the reasons given by the people who are agreeing with me on the conclusion here. But I guess that happens.

It just shows that there’s lots of reasons to go down that path. Although I’m pretty sure I agree with you on this one bit, at least:

I mean, one aspect has nothing even to do with how you choose to narrate, but is rather about noticing that the system isn't reinforcing the fiction that its trying to represent well. If you find yourself rolling for things where failure due to the character 'just not pulling it off' would feel like it depicts that character's skill incorrectly, that's a good sign that the system is misaligned with the fiction.

[Fiery Diamond]

I get what people are saying about misaligned fiction and system, but that's not always something that's solvable. I mean, if I were to make a game where the fiction and system aligned, "untrained, has no idea what they're doing" would have a success rate of at least fifty percent for anything humanly possible in the real world, while "moderately competent" would have a success rate of closer to ninety percent for Olympic-level tasks, and "specialist" would never have a success rate for anything physically possible unopposed below ninety-nine percent. And that's for, in D&D terms, characters no higher than level five. And who else would actually want to play in that game? Nobody, that's who. Narrating away failure as external allows people who see even the untrained as hyper-competent to visualize their specialist as, you know, being good (by their own terms), without compromising the "it's a game, rolls actually matter" part so that people will want to play.

You have to realize that not everyone wants the same things out of the game, and that a misalignment between real probability and imagined probability is pretty much inevitable for a lot of people. If you can't fix the misalignment, your options are 1) play a game with drastically different mechanics or 2) use the almighty power of narrating roll results to reinforce the fiction in spite of the mechanics. I understand that many people prefer option 1, but that doesn't make option 2 any less valid. What if I want to play a game with D&D character abilities and spells, using D&D settings, but want the fiction to have the competence levels dialed up by an order of magnitude? If I can find a balance point where suspension of disbelief isn't broken for me, that's all that really matters, regardless of whether the mechanics are actually truly suited to my imagined levels of competence.

[NichG]

I get what people are saying about misaligned fiction and system, but that's not always something that's solvable. I mean, if I were to make a game where the fiction and system aligned, "untrained, has no idea what they're doing" would have a success rate of at least fifty percent for anything humanly possible in the real world, while "moderately competent" would have a success rate of closer to ninety percent for Olympic-level tasks, and "specialist" would never have a success rate for anything physically possible unopposed below ninety-nine percent. And that's for, in D&D terms, characters no higher than level five. And who else would actually want to play in that game? Nobody, that's who. Narrating away failure as external allows people who see even the untrained as hyper-competent to visualize their specialist as, you know, being good (by their own terms), without compromising the "it's a game, rolls actually matter" part so that people will want to play.

You have to realize that not everyone wants the same things out of the game, and that a misalignment between real probability and imagined probability is pretty much inevitable for a lot of people. If you can't fix the misalignment, your options are 1) play a game with drastically different mechanics or 2) use the almighty power of narrating roll results to reinforce the fiction in spite of the mechanics. I understand that many people prefer option 1, but that doesn't make option 2 any less valid. What if I want to play a game with D&D character abilities and spells, using D&D settings, but want the fiction to have the competence levels dialed up by an order of magnitude? If I can find a balance point where suspension of disbelief isn't broken for me, that's all that really matters, regardless of whether the mechanics are actually truly suited to my imagined levels of competence.

You can absolutely have games that don't hinge on there being high uncertainty dice rolls. Older versions of D&D were like this where basically 'if you have to roll for something, you already messed up'. You can use costs instead of binary success/failure for the resolution system. You can have 'roll to determine your damage multiplier' rather than 'roll to determine whether you hit'. You can anticipate that the players want Lv5 characters to represent the pinnacle of human competency and simultaneously show that by having the characters trivially and without failure do things which would be hard for real people, but also have lots of opportunities where they might have to stick out their neck and do things which would be hard even for mythological figures. Don't have the pinnacle of physical skill struggle with jumping over a 2 foot creek, have them struggle with jumping across a 15 foot chasm.

Have those external factors that you'd make up post-hoc to explain a weird failure already be there before the roll is even attempted. This guy doesn't just have a helmet with a narrow visor slit, he's got some sort of helmet that constantly moves and shifts around and uses optics to continually maintain his line of sight even as the weakspot flickers from place to place.Or bring the concept of the game down and treat even high level characters as within the bounds of real-world human ability so that it actually makes sense that even the Lv20 Bard has only a 50/50 chance of hitting a hole in one on a par 4 golf course.

But basically, being coherent is usually worthwhile in the long run. It's not always worthwhile to insist on coherency on every single little thing, but if you sacrifice it altogether, the ability to think about the world of a game as 'a place that makes some kind of (internal) sense' degrades more the longer you play.

[Satinavian]

I get what people are saying about misaligned fiction and system, but that's not always something that's solvable. I mean, if I were to make a game where the fiction and system aligned, "untrained, has no idea what they're doing" would have a success rate of at least fifty percent for anything humanly possible in the real world, while "moderately competent" would have a success rate of closer to ninety percent for Olympic-level tasks, and "specialist" would never have a success rate for anything physically possible unopposed below ninety-nine percent.

How is that fiction and system aligning ? If those aligned, you would expect success chances that seem reasonable.

Yes, people complained about chances too low for competent characters doing routine tasks, but chances too high are a problem as well.

And yes, systems aligned to fiction tend to have significantly less randomness and more importance of skill than D&D. But there are enough of them out there.

[Quertus]

And yes, systems aligned to fiction tend to have significantly less randomness and more importance of skill than D&D.

I mean, I consider 2e D&D to be the best RPG of all time, and even I will admit that being unable to get off Arangee for skill checks is to its detriment.

Best I could ever do was to align the fiction to the mechanics, and create spoof worlds where NPCs failed to get dressed as often as not, where the “ancient ruins” was actually built last week on a failed skill roll, and Lawful ruled, as society was essential to ensure that someone succeeded at getting a fire started, or at cooking a meal for the community.

[Easy e]

Slightly off-topic, but I think many GMs try to narrate too much with purple prose, instead of just saying the result.

For example, "You missed" or "You did not succeed yet" is perfectly fine. There is no need to go into more detail, unless the player asks for it. Typically, players do not care beyond the success/failure of the action.

Really Off-Topic, GMs in general explain way more than they need to and instead should try "less is more" in more cases.

[kyoryu]

Slightly off-topic, but I think many GMs try to narrate too much with purple prose, instead of just saying the result.

For example, "You missed" or "You did not succeed yet" is perfectly fine. There is no need to go into more detail, unless the player asks for it. Typically, players do not care beyond the success/failure of the action.

Really Off-Topic, GMs in general explain way more than they need to and instead should try "less is more" in more cases.

I think there's a large, large middle ground between "pure mechanics" and "purple prose."

"The orc barely parries your attack" takes only marginally more time than "you missed", and, to me, keeps the game grounded in the fictional reality. You may not care about that, and that's also fine. I do, and I know a lot of people that do. I also know people that don't.

We can both agree that a result like "the orc sees your attack coming, bringing up his blade to meet yours in a shower of sparks. He grits his teeth and glares at you, an orcish bellow coming from his barrel chest as he raises his fist defiantly" is completely unnecessary.

[Quertus]

"The orc barely parries your attack" takes only marginally more time than "you missed", and, to me, keeps the game grounded in the fictional reality.

The problem (aside from the base Time spent) is when that narration doesn’t match the fictional reality: “but this sword can’t be parried”. Or, perhaps worse, when it does match the fictional reality, but triggers its own Rules calls: “so… when he parried my Doom blade, his sword rusted to uselessness, and he caught on fire?”

I’ve just Time and again seen negative value in GM narration, not only in the GM wasting my time, but in setting themselves up for failure. Once in a blue moon thought-through narration has value. Constant mindless drivel? Not so much.

Now, maybe you’re used to a better cut of GM than I am. Maybe you’re used to GM’s who could handle such things without painting themselves into a corner by parrying the Blade of Doom with the McGuffin we’re looking for. Who could introduce a boat, and handle physics to within an order of magnitude of believability when a Dragon lands on it. Who could remember the characters and effects they’ve placed in the game.

But my experience says, the less opportunity you give the GM to mess up, the better.

Of course, on the main topic, “a roll should represent all the things you’ve already narrated into the scene” - there’s rarely a need to be redundant, to repetitively state what the players already know, to reiterate the details of the scene. There’s an exception, when the GM uses this to good value, letting the PCs see something via “showing, not telling”, particularly when it’s not painfully obvious that they’re doing so because they’ve kept a consistent level of description up throughout the campaign.

Of course, again, I’ve never had a GM show that level of skill (no, not even myself), so I still see it as having strictly negative mechanical value.

[Telok]

The problem (aside from the base Time spent) is when that narration doesn’t match the fictional reality: “but this sword can’t be parried”. Or, perhaps worse, when it does match the fictional reality, but triggers its own Rules calls: “so… when he parried my Doom blade, his sword rusted to uselessness, and he caught on fire?”

I would just like to note in general that the above is only an issue in D&D-likes that don't allow the characters to take defensive reactions like parries & dodges. Systems where being a master sword user means you're better at parrying than an untrained mook will also have rules & tags & guidance on what can or cannot be parried and what special effects might then happen.

I have a half worked up system where I'm considering making all attacks opposed rolls. You can oppose an attack with weapon skill (bonus for a shield), dodgyness, or your armor rating. Still considering how it should interact with reaction actions & spending effort to get additional reactions.

[BRC]

The problem (aside from the base Time spent) is when that narration doesn’t match the fictional reality: “but this sword can’t be parried”. Or, perhaps worse, when it does match the fictional reality, but triggers its own Rules calls: “so… when he parried my Doom blade, his sword rusted to uselessness, and he caught on fire?”

I’ve just Time and again seen negative value in GM narration, not only in the GM wasting my time, but in setting themselves up for failure. Once in a blue moon thought-through narration has value. Constant mindless drivel? Not so much.

Now, maybe you’re used to a better cut of GM than I am. Maybe you’re used to GM’s who could handle such things without painting themselves into a corner by parrying the Blade of Doom with the McGuffin we’re looking for. Who could introduce a boat, and handle physics to within an order of magnitude of believability when a Dragon lands on it. Who could remember the characters and effects they’ve placed in the game.

But my experience says, the less opportunity you give the GM to mess up, the better.

Of course, on the main topic, “a roll should represent all the things you’ve already narrated into the scene” - there’s rarely a need to be redundant, to repetitively state what the players already know, to reiterate the details of the scene. There’s an exception, when the GM uses this to good value, letting the PCs see something via “showing, not telling”, particularly when it’s not painfully obvious that they’re doing so because they’ve kept a consistent level of description up throughout the campaign.

Of course, again, I’ve never had a GM show that level of skill (no, not even myself), so I still see it as having strictly negative mechanical value.

I feel like we need to step back a moment and think about different sorts of tests, because Combat is kind of a weird exception in two ways.

1) Combat usually means making a lot of tests in quick succession, with each representing a single discreet action (An arrow fired, a sword swung), with fully defined mechanics for resolving it.

2) The turn-based nature of Combat means that failing a test in combat has consequences built in (You spent your turn).

So, while narration in combat can provide flavor and keep the game from devolving into "I throw my number at their number until they run out of number", it's less necessary than it is elsewhere.

Part of the point of narration is to make sure everybody is on the same page as the scene transitions from the scene BEFORE the roll, to the scene AFTER the roll. every roll should advance the scene somehow.

"Bob fails to chop a hole in the fence before the zombies get here" can mean a lot of things. There may or may not be hole in the fence, the Zombies could be at the alley mouth or their rotting fingers could be closing around Bob, the fence could be unscratched or there could be a hole in it that Bob could squeeze through if he could get away from the Zombies.

For something like the BAFZ scenario, where the effort takes a period of time, it's relevant how long the character tries to succeed before they know they've failed. Do they have a chance to try something else, or do we assume they keep trying until the moment the consequences catch up with them. The GM's narration needs to bridge that gap and firmly set the new scene, and so the details about WHY the effort succeeded or failed are important.

In combat, that's not really necessary, The state of the scene before the attack + "You missed" tells us everything we need to know about the current state of the scene, because the rules already say how things get resolved.

That said, narration CAN be a tool to provide information. If you narrate the attack bouncing off the enemy's armor vs if the enemy is described as dodging tells you about the nature of the foe. And of course, some people enjoy a good cinematic narrative of the fight.

[Quertus]

I would just like to note in general that the above is only an issue in D&D-likes that don't allow the characters to take defensive reactions like parries & dodges. Systems where being a master sword user means you're better at parrying than an untrained mook will also have rules & tags & guidance on what can or cannot be parried and what special effects might then happen.

I have a half worked up system where I'm considering making all attacks opposed rolls. You can oppose an attack with weapon skill (bonus for a shield), dodgyness, or your armor rating. Still considering how it should interact with reaction actions & spending effort to get additional reactions.

Um… I think you’ve got it backwards: it’s the systems where weapons have tags that say “can’t be parried” where GM’s notoriously and hilariously **** this up. Where “parrying” ought to be an action, but they use it as fluff.

Note that 2e D&D did have rules for parrying, that were technically better than “block sword with face”, but not by much.

[Telok]

Um… I think you’ve got it backwards: it’s the systems where weapons have tags that say “can’t be parried” where GM’s notoriously and hilariously **** this up. Where “parrying” ought to be an action, but they use it as fluff.

Note that 2e D&D did have rules for parrying, that were technically better than “block sword with face”, but not by much.

Are we talking about the same sorts of systems? The ones I'm thinking of make stuff like parrying a roll. No roll = no parry = no saying parry & no interaction with a "may not parry" tag or other tags. Is there a particular system where someone declares a parry and its not an active thing?

Been there, works fine:
P1, "Attack" <roll> "is a hit?"
P2, "Parry" <roll> "successful parry"
P1, "Power weapon"
P2, "... crap, you hit then... does half a chainsword counting as a club sound ok?"

Also fine:
P1, "Attack" <roll> "is a hit?"
P2, "Parry" <roll>
P1, "Flail, no parry. Want a take back?"
P2, "Change to dodge?"
P1, "Ok"
P2, <roll> "Yuck... and that's why I wanted to parry. You hit."

[Quertus]

Are we talking about the same sorts of systems? The ones I'm thinking of make stuff like parrying a roll. No roll = no parry = no saying parry & no interaction with a "may not parry" tag or other tags. Is there a particular system where someone declares a parry and its not an active thing?

Been there, works fine:
P1, "Attack" <roll> "is a hit?"
P2, "Parry" <roll> "successful parry"
P1, "Power weapon"
P2, "... crap, you hit then... does half a chainsword counting as a club sound ok?"

Also fine:
P1, "Attack" <roll> "is a hit?"
P2, "Parry" <roll>
P1, "Flail, no parry. Want a take back?"
P2, "Change to dodge?"
P1, "Ok"
P2, <roll> "Yuck... and that's why I wanted to parry. You hit."

I’m pointing out the level of fail I’ve seen and expect from GM’s, who cannot remember that “parry” is a keyword in this system, and use it as fluff anyway.

P: “Attack” <roll> “I missed”.
GM: “it parries the blow out of the way, blah blah blah”
P: “but… it can’t be parried…”

The less rope we hand GM’s, the less they’ll hang themselves, IME.

EDIT: and, yes, if I were dumb enough to hand you a sword in 2e D&D with the property “can’t be parried”, and dumb enough to narrate combat, I’d likely be dumb enough to narrate someone parrying your blade at some point. That’s just the way such things go.

[Fiery Diamond]

How is that fiction and system aligning ? If those aligned, you would expect success chances that seem reasonable.

Yes, people complained about chances too low for competent characters doing routine tasks, but chances too high are a problem as well.

And yes, systems aligned to fiction tend to have significantly less randomness and more importance of skill than D&D. But there are enough of them out there.

I think you may have misinterpreted my use of the word "fiction." I didn't mean "like the stories it is inspired by," I meant "the agreed-upon mental construct of how competent the characters are within the world they inhabit." It's not about "reasonable" success chances. It's about "success chances that match what the players envision their characters to be capable of," with no regard for realism at all. Different people have different desires for where this sweet spot is, and "hyper competent power fantasy" is a perfectly legitimate sweet spot for someone to want. It's not what you want, clearly, but that's kind of irrelevant to the point.

[gbaji]

I would just like to note in general that the above is only an issue in D&D-likes that don't allow the characters to take defensive reactions like parries & dodges. Systems where being a master sword user means you're better at parrying than an untrained mook will also have rules & tags & guidance on what can or cannot be parried and what special effects might then happen.

Yup. D&D is a bit "odd" that it conflates several components of a combat hit attempt into one thing. All "hits" represent striking the opponent, well enough, and hard enough, to avoid any attempts at evading/dodging, also get through any assumed parry/block/whatever, *and* penetrate whatever armor the target is wearing, doing damage as a result. There's a single AC target number that is a combination of all different defensive things. It does simplify things a bit, but also leave an odd "all damage or no damage" element (excepting damage reduction effects in some versions).

Other systems break this down a bit. I play most often in RuneQuest, in which the attacker simply rolls his attack roll, based on the percentage with the weapon he's using. So simple. No real calculation. Look at the number on your sheet, roll a percentile die. If you roll under the percentage skill with the weapon, you hit. The defender chooses to dodge or parry. He looks at his sheet, picks the skill he's using, and then also attempts to succeed at rolling under that skill. If it's a parry, then the parry blocks X amount of damage from the hit, based on the weapon/shield being used (also written on the sheet). The remaining damage, if any, get through, at which point any armor gained from magic reduces the damage (protection and shield spells specifically). Any remaining damage then goes through and hits the armor on the location struck, and any damage remaining from *that* goes through and does damage to the location.

Dodges are a bit different, in that it's level of success that matters, but you either miss entirely, or hit entirely, with the reaming damage going on to hit magic defenses, then worn armor (oh, and I suppose skin armor if the target has that), then does damage to the location.

While it's a few extra steps involved (honestly, just two die rolls, one from each person), the results can be extremely well calculated. You know whether the weapon impacted the parrying shield/weapon, and if it has some special effect that does something in that case, it applies. You know if the damage got through the parry (or dodge) and impacted the magical defenses of the target (and could possibly have effect on those, for some strange artifact that maybe dispels stuff it touches). You then also know if the weapon impacts the worn armor (and could have yet more effects, like acid melting, corrosion, whatever). And you know how much exactly gets through and does damage to the target in the form of hps.

One extra roll, and a few simple bits of math and you can simulate a wide array of different effects accurately.

I have a half worked up system where I'm considering making all attacks opposed rolls. You can oppose an attack with weapon skill (bonus for a shield), dodgyness, or your armor rating. Still considering how it should interact with reaction actions & spending effort to get additional reactions.

Honestly, the biggest problem with "to hit" based systems is the concept that all defensive things tend to be "all or nothing" effects. So heavier armor just increases the odds that you take zero damage, and better skill at blocking/dodging does the same. But yeah, if that's already imbedded in the system you are using, then you're kinda stuck with it. Giving an option to spend an action (or half action, or whatever depending on the system you are using), to apply a defense and thus a bonus to AC/whatever works. It can be tricky to balance that into an existing combat system though. We implemented a "desperate defense" option into our game, where you could use that instead of attacking that round, had to retreat a space, but as long as you made your defensive skill, you were able to avoid all damage from opponents that round (new roll required for each attacking opponent though). It works ok, and allows for survival when characters find themselves in really really tough spots and just need to survive long enough for someone else to come and bail them out.

Some other roll rules for skill based (specifically percentage based) systems:

One of the house rules I added to the base RQ rules was that for weapons, natural skill over 100% could be used to subtract from the opposed persons skill (in both directions, so an attacker could make a parry or dodge miss, or a defender could make the attacker simply miss, which means that the attacker didn't even hit the parrying item. You're just that good). This allows the system to normalize skill rolls within a 100% range, while scaling up to relatively high skill levels. Allows for trivializing fights between extremely skilled people and "wimpy" opponents, while allowing for the same "do I hit or miss" dynamic in fights between similarly skilled folks. I find that this sort of mechanism works well in skill based (especially percentage skill based) systems. In RQ specifically, I limit this to "natural skill" (what's actually written on the character sheet), because there are a number of spells that can increase your chance to hit. I didn't want those being used. They still act to increase the chances for special or critical hits (and can soak up reductions from a higher skilled opponent), but can't themselves be used to subtract. Oh, and obviously, subtractions always reduce the opponents natural skill first, then any magic (otherwise you get bizarre cases where the resulting skills change based on who declares a subtraction first).

For non-combat skills, but still opposed skills, I do something similar. Basically, the first person rolls their skill. The amount they make it by becomes the difficulty factor for the opposed skill (feel free to round to easy numbers if you feel like making the math easier). So if the PC is attempting to sneak up on an NPC, you have the PC roll their sneak skill (applying environmental factors like light, cover, etc), then if they make it, subtract the difference from the NPCs spot chance. This also allows for scaling skill levels (otherwise everyone over 100% is like "I made my sneak" and "I made my spot", er,... "was I spotted, or not?". Most systems (RQ does this) introduce some sort of "level of success" bit, but I personally hate those. They work "ok" at low level, but at high skill level it basically turns into "lucky die roller wins". And still leaves us wondering which skill should "win" if both achieve the same level of success.

I've seen other percentage based systems that use a "best level wins, and highest roll at the same level wins", method. Which I also personally hate with a passion. Yeah. Technically, if you have a 40% skill, and I have a 60% skill, I can roll anywhere from 41-60 and always beat you, but I could also roll a 22, and you roll a 25 and now you win. Just because I rolled... lower? In a system where rolling lower is supposed to be "better"? And when you have levels of success (like RQ does), it gets worse. I rolling 41-60 win, unless you roll a 01-08, in which case you win. I rolling a 13-40 win, if you roll 41-00 *or* if you roll a 09-40 *and* I roll higher than you, but I lose if you roll a 01-08. I rolling a 04-12 win, if you roll 09-00 *or* if you roll 03-08 *and* I roll higher than you, but I lose if you roll a 01-02. I rolling a 01-03 win, if you roll 03-00 *or* if you roll 01-02 *and* I roll higher than you. OMG is that not the most obnoxiously silly set of rules and produces incredibly silly odds calculations? Yes, it is.

Yes. My method involves some math, but is not really a new concept. Most skills systems include some sort of difficulty factor that the GM can impose on the skill (that lock is particularly easy/hard, that climb is particularly easy/hard, etc). You apply plusses or minuses to the skill being attempted. This just gives you the same thing when using opposed rolls. And yeah, it naturally scales linearly based on skill rather than having odd bumps and odds shifts based on where the skills are relative to an arbitrary cap number. Oddly, in many ways, it's very similar to D&D's system of skill rank versus target roll, but allows that dynamic for a percentage based skills system. The "odds" are not based on the actual skill rank as a base, but the linear numerical difference between the skill ranks being used (one creates the target number for the other). It's elegant, relatively easy to apply in practice, and produces absolutely clear and consistent final success odds.


Oh. And someone earlier talked about issues with people failing simple tasks due to skill rules. Same concept applies though. The GM can assign plusses as well as minuses for target numbers. So, even in such systems, you just assign really high plusses to doing things that are "easy". So putting on your pants doesn't become about failing round after round. It's an easy thing. You can do it. You need a strength roll to pick up that tea cup? Well, it's got a +200 difficulty factor, so you just can't fail, no matter how weak you are. There. Done.

Um. I also introduced linear values at which it was increasingly less likely for "auto fail" or "auto success" numbers to work. This is heavily dependent on game system, but for D&D, it could be something like "if your target roll is double a 20, then a nat 20 no longer succeeds", and "if your target roll is -20, then a nat 1 no longer fails" (feel free to scale as needed though). In RQ, there's a concept that's the same 01-05 always succeeds, and 96-00 always fails. For certain opposed checks (stat checks typically), base percent if the totals are the same is 50%. For each one point higher/lower, the percent chance move 5% higher/lower. So being +10 points, gives you a 100% chance, but that's limited to 95% due to the whole "96-00 fails" bit. At +30 (effectively 200%), we reduce the auto fail to 00% only (1% chance to fail). At +50 (real chance 300%), there no chance to fail. The idea being that if you have 50 more strength than the other guy, you kinda can't lose. And the same occurs in the other direction. -30 means your chance to auto succeed drops to 01%, and at -50, you have no chance.

That does great things like prevent the 05% chance to "OMG I just hit that god with my spell" nonsense. And makes arm wrestling with a giant a bad idea. But, just like with everything else, it allows the system to scale up to giants arm wrestling with each other (or characters with a whole lot of strength enhancing magic).

Dunno. Lots of ways to do this, but I do tend to be a fan of various forms of linearly scaling probability rules for rolls in my games. I just find that it scales better than any other method, while still keeping the math relatively simple, *and* has the virtue of maintaining the same "high/low roll is always better" dynamic through all skill levels.

[martixy]

A roll should represent a moment of minor DRAMA.

If there is no drama to be had, there should not be a roll.

If the drama is too major, it should not be a roll.


Rephrased a bit:
You should never hang the fate of the world on a die roll.
And you should be looking for every opportunity to avoid die rolls.

[Quertus]

A roll should represent a moment of minor DRAMA.

If there is no drama to be had, there should not be a roll.

If the drama is too major, it should not be a roll.


Rephrased a bit:
You should never hang the fate of the world on a die roll.
And you should be looking for every opportunity to avoid die rolls.

Rephrased a bit more?
“A roll should represent things we cannot resolve otherwise, despite our best efforts.”

I feel like we need to step back a moment and think about different sorts of tests, because Combat is kind of a weird exception in two ways.

1) Combat usually means making a lot of tests in quick succession, with each representing a single discreet action (An arrow fired, a sword swung), with fully defined mechanics for resolving it.

2) The turn-based nature of Combat means that failing a test in combat has consequences built in (You spent your turn).

So, while narration in combat can provide flavor and keep the game from devolving into "I throw my number at their number until they run out of number", it's less necessary than it is elsewhere.

Part of the point of narration is to make sure everybody is on the same page as the scene transitions from the scene BEFORE the roll, to the scene AFTER the roll. every roll should advance the scene somehow.

"Bob fails to chop a hole in the fence before the zombies get here" can mean a lot of things. There may or may not be hole in the fence, the Zombies could be at the alley mouth or their rotting fingers could be closing around Bob, the fence could be unscratched or there could be a hole in it that Bob could squeeze through if he could get away from the Zombies.

For something like the BAFZ scenario, where the effort takes a period of time, it's relevant how long the character tries to succeed before they know they've failed. Do they have a chance to try something else, or do we assume they keep trying until the moment the consequences catch up with them. The GM's narration needs to bridge that gap and firmly set the new scene, and so the details about WHY the effort succeeded or failed are important.

In combat, that's not really necessary, The state of the scene before the attack + "You missed" tells us everything we need to know about the current state of the scene, because the rules already say how things get resolved.

That said, narration CAN be a tool to provide information. If you narrate the attack bouncing off the enemy's armor vs if the enemy is described as dodging tells you about the nature of the foe. And of course, some people enjoy a good cinematic narrative of the fight.

“Narration should represent necessary ‘change of state’ data we cannot otherwise assume the players know”?

If we move a mini on a battlemat, it’s probably safe to say that the players know that the corresponding creature moved, there’s probably no need to narrate that… unless that creature had previously just been scenery - a tree, rock, or building, for example - but now was an active participant in the fight. Darn Treebeard!

[Stonehead]

If Hawkeye / Bullseye / whoever you consider “really cool” only has a 50/50 chance of hitting the target? That’d better be an eye slit in a 500’ tall giant in a Hurricane, else the rules have Captain Hobo‘d the character’s coolness. And, if the rules haven’t Captain Hobo’d the character’s coolness, then you don’t need to narrate extra reasons why they missed - they missed because they were attacking an eye slit 500’ up in a Hurricane.

Also, if you’re playing with players who try to flip a coin, need heads but get tails, and need some “external factor” excuse for why they failed? That’s a failure to understand probability, or terrible self-esteem issues. Play with better players.

That's kind of what I was trying to get across by talking about "making the dc easier," which only ended up distracting from my point.

You made it much clearer, thanks.

But.... like, why?

What is the advantage of narrating as "you just can't?"

Why is "you try to cross the stream but miss the rock" better than "you try to cross the stream but slip on a wet and mossy part of the rock?" Why is "you shoot an arrow at the guy and miss wildly" better than "you shoot an arrow at the guy but he moves at the last second?"

What's the advantage?

Realism is an advantage, but only a few people care about that.

Everyone else has mentioned different playstyle preferences, but one concrete, non-subjective answer is that environmental factors often eliminate the possibility to try again while realistic luck doesn't. If you try to jump the stream but slip on the water, the water will still be there if you want to try again, or if the other party members want to try. If Robin Hood shoots at the arrow slit and misses, if the DM narrates that the weather is storming and making it impossible, it'll still be storming if he wants to try again. If he missed because everyone misses hard shots some time, he can try again in the future.

Now, sometimes you don't want to allow the players to keep trying until they get it right, in which case this becomes a drawback rather than an advantage. I personally like letting people keep trying if there's no time pressure or punishment for failure or anything. And if it's something that they aren't rushed on, and can try as many times as they want, I don't even ask for a roll, because it j

[Tanarii]

If someone fails because of an environmental factor already present, it should either already be an automatic failure or affect the chance of success from the go. Not retroactively start because one person failed.

Choose an Consequence that changes the state of the challenge if you want to make further retries for anyone impossible. If you fail to pick a lock, it should be something like it broke. Not that it suddenly turned into a Class IV Kingstomb Adamantium Lock because you failed.

Or make it a state-of-the-character check if you want just one character to not be able to retry. Like a 3e/4e Knowledge check, you know the thing or you don't know the thing, that character can't try again. Or a BECMI pick locks, you can't try again until you go up a level, because your character's skill is insufficient.