D&D General 1s and 20s: D&D's Narrative Mechanics

[Lanefan]

If you want to keep 1 in 20 dumb luck, enjoy your game. It makes for a poor experience when I've seen it because, guess what, 95% of the time the player FAILS and that hardly seems like "fun"--but maybe you enjoy that? Only you can determine that.

Sometimes 95% failure becomes worth it when that 5% success comes through.

That, and despite everything the players can do to migitate the odds, the game is still at its heart a gamble. Embrace that, and it all makes far more sense.

 

[Theory of Games]

None of that is what I was suggesting. This is an unnecessary alarmist reply.
The 1 and 20 is supposed to describe the failure and success not the ending the scene or gaining some sort of mechanic benefit. The rules of the game are still followed, you don't just get to declare decapitation or dismemberment.
The GM will let you know if the monster dies or loses a limb (i.e. roper tentacle) according to the creature's hit points or damage tally inflicted.

"Alarmist" suggests what I wrote couldn't happen, but I've been a GM and player at sessions where it did

I don't know who you play with, but the people I play with wouldn't do this. I've done this kind of delegation as GM plenty of times and never had this kind of response.

I play with Earthlings and they do crazy stuff. Sometimes it's fun and other times it can be "problematic"

Um, no. Auto failure is still a thing. Not every longshot has a 1 in 20 chance of success. Players can't use it to accomplish anything they feel like. If you misunderstand what it's about that badly, no wonder you think it shouldn't be done.

Players can do whatever they want IF the GM gives them the space to do it. Which is why I suggest rolling a "1" or a "20" should be nothing special. Or if those rolls are defined as "crits", what happens should be clearly determined during Session 0

I'm posting exclusively for the novice GMs who could run into issues. Experienced GMs will probably have a handle on this. Maybe.

 

[Lanefan]

No. It is how most games play in my experience anyway. For checks and saves, a 1 doesn't always fail and a 20 does always succeed. I know many groups probably extend the combat rules into checks and saves, but I don't, my groups don't, and frankly never have.

Do I want my STR 8 PC with no proficiency in Athletics to be able to leap a 20-foot chasm (at DC 20) just because they rolled a natural 20? No. The group and PCs have to find another way. Think of it this way: if I DID want that... what are the results if I don't roll the natural 20? I fall to my death maybe! No thank you.

Hell, if it's a choice between a 1-in-20 chance of safely jumping the gap vs a 0-in-20 chance of surviving what's chasing me on this side then I'm trying that jump every time.

Allowing 20s and 1s a bit of latitude reflects, in a way, the extreme ends of the bell curve of success probability.

If you've ever looked at the 1e combat matrix it gives an example of what I mean. As the target's AC increases the numeric progression (i.e. the roll) required to hit rises in lockstep - until it gets to 20, where it plateaus over several increasing AC ratings until resuming its climb to 21, 22, etc.

 

[Flights of Fancy]

Simple answer: if the in-game characters, as reflected by the player's knowledge, don't realize a task is beyond them (which can often be the case) and the roll comes up 4, they don't and shouldn't know whether they failed due to bad luck or failed because they couldn't succeed. They just know - both in and out of character - they failed this time and have to try something else.

The problem with this is meta-currency. The player knows it was bad luck when they rolled a 4. You might think they shouldn't know why, but the truth is the player knows. Depending on the scene, they might feel it is pretty stupid and resent failing due simply to bad luck.

Sometimes 95% failure becomes worth it when that 5% success comes through.

That, and despite everything the players can do to migitate the odds, the game is still at its heart a gamble. Embrace that, and it all makes far more sense.

I have no issue with 5% success, in and off itself. I take issue when the numbers say there should be 0%, and people want the "nat 20" to work anyway. If a roll of 20 can't make the numbers work for success, there should never be a roll. That is what I am saying.

Hell, if it's a choice between a 1-in-20 chance of safely jumping the gap vs a 0-in-20 chance of surviving what's chasing me on this side then I'm trying that jump every time.

But it isn't a 1-in-20, it is a 0-in-20. A roll of 20, -1 for STR 8, is 19 which fails the DC 20.

Allowing 20s and 1s a bit of latitude reflects, in a way, the extreme ends of the bell curve of success probability.

Sure, and I can see why some groups embrace that. Unless a DM is strict about when to call for checks and doesn't allow absurdly ridiculous thing, I can't get behind the 1 always failing and 20 always succeeding. I like to keep it simple: if the roll + bonuses makes it, you succeed. Otherwise you fail. I don't care if the roll is 20--if you don't have the bonuses to make the extra amount, no good.

If you've ever looked at the 1e combat matrix it gives an example of what I mean. As the target's AC increases the numeric progression (i.e. the roll) required to hit rises in lockstep - until it gets to 20, where it plateaus over several increasing AC ratings until resuming its climb to 21, 22, etc.

Great example. That natural 20 on the combat matrices (?) in 1E was to allow someone to hit higher ACs without bonuses. Once the numbers went over 20, you needed a natural 20 AND a bonus of +1 (for 21) or better (for higher numbers).

I suppose you could extrapolate that to 5E-mechanics, so if natural 20+bonus is "within 5" of the DC, then you can do it. In the jumping example above, that would be such a case. But if the DC was 25, then even the natural 20 would fail.

But this gets to a level of nuiance that many players don't care to embrace I would think.

 

[Blue]

Maybe so, but if they don't know they can't succeed then IMO the roll should happen anyway; and only on a nat. 20 should they realize the task is in fact beyond them.

If I go to move something heavy, I can get the idea that it's not going to move and I won't be able to pick it up consistently, I don't only figure it out with my knee-bursting best. Why would PCs be less effective than the players at figuring that out?

 

[soviet]

"Alarmist" suggests what I wrote couldn't happen, but I've been a GM and player at sessions where it did

I play with Earthlings and they do crazy stuff. Sometimes it's fun and other times it can be "problematic"

Players can do whatever they want IF the GM gives them the space to do it. Which is why I suggest rolling a "1" or a "20" should be nothing special. Or if those rolls are defined as "crits", what happens should be clearly determined during Session 0

I'm posting exclusively for the novice GMs who could run into issues. Experienced GMs will probably have a handle on this. Maybe.

I'm not sure how many novice GMs are likely to be reading this thread, but if any are I submit that 'you can't trust your players not to take advantage' is not a great lesson to pass on to them.

 

[Lanefan]

The problem with this is meta-currency. The player knows it was bad luck when they rolled a 4. You might think they shouldn't know why, but the truth is the player knows. Depending on the scene, they might feel it is pretty stupid and resent failing due simply to bad luck.

Which is why, if that meta-knowledge would matter, I do those rolls in secret.

I have no issue with 5% success, in and off itself. I take issue when the numbers say there should be 0%, and people want the "nat 20" to work anyway. If a roll of 20 can't make the numbers work for success, there should never be a roll. That is what I am saying.

What I want, somehow, is for the game to support there being a 1% chance of success. As the designers have shoehorned everything into the much less granular d20, the only ways I can see of achieving this is allowing '20' some latitude or (the unpopular option) maybe adding a second, confirm roll.

But it isn't a 1-in-20, it is a 0-in-20. A roll of 20, -1 for STR 8, is 19 which fails the DC 20.

Sure, and I can see why some groups embrace that. Unless a DM is strict about when to call for checks and doesn't allow absurdly ridiculous thing, I can't get behind the 1 always failing and 20 always succeeding. I like to keep it simple: if the roll + bonuses makes it, you succeed. Otherwise you fail. I don't care if the roll is 20--if you don't have the bonuses to make the extra amount, no good.

That's the sort of binary thinking the designers are after. We'll probably differ on our opinions of such.

Great example. That natural 20 on the combat matrices (?) in 1E was to allow someone to hit higher ACs without bonuses. Once the numbers went over 20, you needed a natural 20 AND a bonus of +1 (for 21) or better (for higher numbers).

I suppose you could extrapolate that to 5E-mechanics, so if natural 20+bonus is "within 5" of the DC, then you can do it. In the jumping example above, that would be such a case. But if the DC was 25, then even the natural 20 would fail.

But this gets to a level of nuiance that many players don't care to embrace I would think.

The players wouldn't need to worry about it. This would be DM-side arithmetic. But the within-5 idea on a natural 20 is solid enough to account for some really lucky moments.

Flip side: I'm quite happy with a 1 always failing regardless of bonus. Nobody's perfect.

 

[Hriston]

I don't do critical success/failure on ability checks or saves. If the roll succeeds, the character achieves their intent. If it fails, they do not and suffer a consequence I probably stated before the roll was made.

I don't know what you mean by narrative mechanic, but I assume it has something to do with it having "a more powerful impact on the fiction" than the above, which sounds okay. It's just not how I do it.

 

[Flights of Fancy]

Which is why, if that meta-knowledge would matter, I do those rolls in secret.

But it always matters because it impacts the players' experiences.

What I want, somehow, is for the game to support there being a 1% chance of success. As the designers have shoehorned everything into the much less granular d20, the only ways I can see of achieving this is allowing '20' some latitude or (the unpopular option) maybe adding a second, confirm roll.

I've never had any issue with confirmation rolls. I loved the threat ranges in 3E, for example! What I don't like is for there to be a 1% chance when there really is a 0% chance. If something just can't be done, don't allow it.

That's the sort of binary thinking the designers are after. We'll probably differ on our opinions of such.

Seems to be.

The players wouldn't need to worry about it. This would be DM-side arithmetic. But the within-5 idea on a natural 20 is solid enough to account for some really lucky moments.

Flip side: I'm quite happy with a 1 always failing regardless of bonus. Nobody's perfect.

In a less bounded system, I could see something like this working. However, in 5E, when the PCs are mostly likely to encounter higher DCs, odds are they don't need nat 20's to hit them. In this respect, I often see this as a moot point. The DMG itself says how you can run a very challenging and appropriate game just using DC 10 to 20 and nothing higher.

As for failure, the confirmation failure roll would also work for me, but not a flat 5%.

 

[Scribe]

Would have been interesting to see poll results on this. If people attribute narrative to 1/20 or just the mechanical aspects.

(Mechanics here.)