Why don't 1s/20s autofail/pass on Saving Throws?

[Amnestic]

Fairly self-explanatory title, hopefully.

I understand it for ability checks to a certain extent. At least, moreso than saves. If you want to display a level of competence, having a 'floor' for your ability checks helps convey that, and they're typically less important for the undeniable mechanical core of the game: combat.

Attack rolls have automiss/crit on 1s/20s, so we've already got the concept in this edition. This isn't an appeal to tradition, but I will note that 3.5 and 4e* both had autofail/crit on saves. I don't know if older editions did as well. With bounded accuracy being the name of the game it just strikes me as weird that saves can be 'outleveled' where attack rolls, by virtue of the nat 20, cannot. One of the most common saves you'll see characters 'outlevel' is concentration, since a solid % of the time the DC will be a mere 10, and that's something you can guarantee between a save boost or two (i.e. stone of good luck), proficiency, and a moderate amount of constitution. Chances are if you want it, a caster can have Concentration guaranteed on most attacks by tier 3.

Likewise, there are some higher CR enemies with DC21+ save features that some characters may be literally unable to pass, ever. One example is Asmodeus in...the Chains of Asmodeus. Now he's CR30, so he should be scary, but he's got a trio of features with DC24 saves - one targeting Wis, one Cha, and one Int (plus some other dex saves knocking around elsewhere).

*(okay 4e's not saving throws exactly, but it was 'attack vs. reflex/will/fort', so essentially the same thing for 5e)

So yeah, the question of the thread is, I guess:-
Does anyone know (for example did the devs ever discuss) why they moved saves away from the autofail/pass that they had in previous editions? Is there any behind-the-scenes insight onto the decision behind it?

And also

Would you prefer it if they were autopass/fail or not? Personally, after getting hit (well, almost - it got counterspelled) by a DC23 Psychic Scream that my cleric could never pass the save for and locked me down forever guaranteed, my vibe is leaning towards "actually nat 20s/1s on saves are good for the game" but I am definitely feeling my bias on that one, so curious how other people feel, or even if they care at all.

[stoutstien]

IMO all save DCs should drop a good amount but the ability to stack save values should also be curtailed.

I also think that 4e has a better system all together for handling them as a generic approach. First "save" was checked by a passive score and the following ones you rolled for.

* A 25%-75% chance to save seems about right for me. Could make a siding scale so you can't get stuck but I rather keep it simple.

[Mindflayer_Inc]

Honestly, once you hit like level 3 or 5, the auto miss on a 1 should go out the window.

The auto-hit should lower.

At least for the two martial (non-magicals / non-spellcasting) classes.

But more to the point, WorC isn't really known for having rules be comparable across the board. So the answer is "shhh don't think about it, WotC likes it that way".

[JLandan]

My table allows for auto fail/succeed on nat 1/20 on saves. It's a holdover from 3e and Pathfinder.
It's a very common house rule.

[KorvinStarmast]

Fairly self-explanatory title, hopefully.

Have you considered asking the dev team?

I don't see the point, and since very few save DCs are a 20 or more anyway ... nor do I see the utility.

And FWIW, as regards edge cases: if you have a wisdom of 10 and you are up against an ancient red dragon, his frightening presence ought to make you flee unless you have some kind of magical help.

(Not sure I care for the auto miss/auto hit either ... but it does keep low level mooks at least dangerous enough in later tiers)

I don't know if older editions did as well.

They did not. (If AD&D 2e added them later on, I am happy to be corrected)
EGG was pretty strong in his position against that.
Mind you, plenty of non official stuff about crits appeared in Dragon Magazine, and plenty of tables used stuff like that. One of our OD&D GMs (plus greyhawk/blackmoor/eldritch wizardry) dragged in a crit hit table from Dragon.

He also had the deal of "if you hit by more than a certain amount over TN, it was a crit" as a feature.
Fiddly? Yes. However, all of us were in technical degrees and mental math/arithmetic wasn't hard for any of us. Did it slow play down a little bit? A bit, but we played pretty up tempo most of the time, so I am not sure if anyone noticed.

I remember one of our PCs going down pretty hard when a crit landed on him and he lost a leg...

[brainface]

The only thing I have real strong opinions about here is that autofailing saves on a 1 is terrible in CRPGS, just due to how they tend to be more hack & slash than even the hackiest and slashiest of tabletop games, so you go from making maybe 6 saves a session to 6 a round. This might be the reason for the change, but it'd be a bit odd to change paper rules to make crpgs run better. \o/

Outside of that it feels like in almost all situations a 1 already fails and a 20 already succeeds, so it also seems mostly irrelevant. I guess it comes up more in higher level play?

[Unoriginal]

Would you prefer it if they were autopass/fail or not?

Abilities that demand a saving throw are usually once a turn, sometime once a round. They scales by the effect becoming nastier and nastier.

A significant part of attacks scaling is letting the person do more attacks.

IMO it's a very good thing that saves don't have autopass/autofail.

Personally, after getting hit (well, almost - it got counterspelled) by a DC23 Psychic Scream that my cleric could never pass the save for and locked me down forever guaranteed, my vibe is leaning towards "actually nat 20s/1s on saves are good for the game" but I am definitely feeling my bias on that one, so curious how other people feel, or even if they care at all.

Well, there is three things you have to consider here:

1) Imagine that your character was hit by that DC 23 Psychic Scream, and rolled anything but a 20. Would you have been more or less annoyed than if you were sure to fail?

2) Imagine that your character was hit by that DC 23 Psychic Scream, and rolled the required 20. Your PC would have passed the save... but wouldn't have Psychic Scream be far less scary then?

Remember that today's adversities are tomorrow's battle tales. Fighting people when their supposedly wallop-packing power fails 5% of the times and it only happens once per turn isn't that enjoyable.

3) You have to consider the thing from the other side of the equation, too:

You're a Rogue with 20 DEX. At lvl 13, you're straight up immune to all DEX saving throws DC 11 or less.

Now imagine your DM telling you that actually, you have 5% chances of failing your DEX saving throw against a CR 1/2 Magma Mephit's breath.

Do you think it'd feel good?

[Amnestic]

Abilities that demand a saving throw are usually once a turn, sometime once a round. They scales by the effect becoming nastier and nastier.

In the cases of damage, sure. In the case of control...eh. Command, Hideous Laughter, Suggestion and Hold Person are all debilitating effects at 1st and 2nd level. If low level effects were all 'soft' control and all higher level effects were 'hard' control then I'd maybe buy that, but they're not nearly so neatly delineated.

1) Imagine that your character was hit by that DC 23 Psychic Scream, and rolled anything but a 20. Would you have been more or less annoyed than if you were sure to fail?

2) Imagine that your character was hit by that DC 23 Psychic Scream, and rolled the required 20. Your PC would have passed the save... but wouldn't have Psychic Scream be far less scary then?

Remember that today's adversities are tomorrow's battle tales. Fighting people when their supposedly wallop-packing power fails 5% of the times and it only happens once per turn isn't that enjoyable.

3) You have to consider the thing from the other side of the equation, too:

You're a Rogue with 20 DEX. At lvl 13, you're straight up immune to all DEX saving throws DC 11 or less.

Now imagine your DM telling you that actually, you have 5% chances of failing your DEX saving throw against a CR 1/2 Magma Mephit's breath.

Do you think it'd feel good?

I mean, just to hit the three questions in order:-
Probably less still, since "you are permanently stunned forever" is way more debilitating than "maybe you can eventually free yourself".

Not really, because 20s are...you know, rare. It's a moment of relief at avoiding the very scary thing, it doesn't diminish how scary it would have been otherwise.

And I wouldn't really mind failing a dex save vs. a magma mephit? Not only would it not do much damage (Evasion still halves it!) but it's the same as getting hit by a crit attack when you're at AC25 or whatever. It happens.

[Unoriginal]

Not really, because 20s are...you know, rare. It's a moment of relief at avoiding the very scary thing, it doesn't diminish how scary it would have been otherwise.

20s aren't actually that rare.

I mean, if you were offered a dish that had 1 chance out of 20 to send you to the hospital, would you eat it?

And I wouldn't really mind failing a dex save vs. a magma mephit? Not only would it not do much damage (Evasion still halves it!) but it's the same as getting hit by a crit attack when you're at AC25 or whatever. It happens.

Wasn't the best example.

As you mentioned, saves are often followed by nasty effects (including of the "battlefield control" variety), and with AC your character inherently becomes less affected by attacks due to having more and more HPs.

Autofail or autopass would change the math for how likely PCs are to resist those effects a lot, and as a result that could throw the efficiency of the concerned NPCs out of whack.

In the reverse, it could also make the used-by-PCs saving throw abilities go out of whack, or at least feel like it. Especially when the PCs take care to target NPCs with saves they can be reasonably sure the NPCs are bad at.

[Jerrykhor]

After DMing for a party of Barbarians who went up against some Mind Flayers, being permanently stunned out of the whole fight is very unfun. I dont care for any arguments, people are here to play D&D, and having 0 chance to break free just stops you from playing the game. Thats why I'll always advocate for auto fail/pass on 1/20s.

Actually i do have an argument. In a game where bounded accuracy is one of the main features, having 0% and 100% chance doesn't make sense.

[Mastikator]

After DMing for a party of Barbarians who went up against some Mind Flayers, being permanently stunned out of the whole fight is very unfun. I dont care for any arguments, people are here to play D&D, and having 0 chance to break free just stops you from playing the game. Thats why I'll always advocate for auto fail/pass on 1/20s.

Actually i do have an argument. In a game where bounded accuracy is one of the main features, having 0% and 100% chance doesn't make sense.

A character with 8 intelligence will have to roll a 16 on the D20 to save against mind blast. Adding auto-success on 20 changes nothing, except maybe the player's feeling so maybe that makes it worth it. I'd hardly call a 20% chance to fail a 0% chance to fail (though it may feel like it).

AFAIK you don't really get into the 20+ DCs on monsters until tier 4, auto-success/failure on 1/20 won't really change the game, I strongly suspect you can add and change nothing about the math in 99% of the games. For instance, a character would need +14 to their intelligence saving throw to save against a mindblast on a natural 1, a 9th level wizard with 20 intelligence will only have +8, even if they're standing next to a 9th level paladin with 20 charisma that would "only" be +13, you'd need to give them a cloak of protection as well. Bounded context remains intact enough for most of the game that a auto-success/fail mechanic doesn't change anything about most campaigns.

Edit- Another example is a nilbog, DC 12 charisma save to be allowed to attack it. A 5th level bard with 18 charisma would have +7 to charisma saves, they'd have to roll 5 or higher to not praise the nilbog. A 6th level paladin with 18 charisma would have +9 to charisma saves (pretty good), they'd still have to roll 3. A 2 on the D20 fails, adding a rule that says you automatically fail on a 1 when a 2 doesn't cut it seems like an emotionally driven choice.

[OvisCaedo]

DC 20+ saves being rare/late game doesn't really change whether or not they're a silly issue that should be addressed, though. If it's a rule proposal for addressing edge cases that only really comes up in edge cases, that's... kind of the point. Usually the ideal for an edge case rule, even!

Though I don't think it's the best answer to how obnoxious some save paradigms get at high levels, I don't see much reason to object to it in a vacuum. I'd be more in favor of a larger change so that whether or not a 20 auto-passes doesn't come up at all in a system that's meant for bounded accuracy.

Nat 1s being an auto-fail, on the other hand, is something I'm not sure should even apply to attack rolls, much less any other mechanic. I suppose for a typical attack there's always some believable chance that the opponent stumbles out of the way of even the most competent warrior's blows.

And on the very specific case of Terrifying Presence on ancient dragons and the like, I'm pretty sure low wisdom should be the requirement for not being scared of one. Wanting to charge up to it and hit it with a sword is easily one of the least wise things one could do!

[Amnestic]

seems like an emotionally driven choice.

Is that an issue? Rules stemming from "this feels good/bad" is hardly a bad thing.

20s aren't actually that rare.

I mean, if you were offered a dish that had 1 chance out of 20 to send you to the hospital, would you eat it?

They're rare when I'm fishing for crits.

Autofail or autopass would change the math for how likely PCs are to resist those effects a lot, and as a result that could throw the efficiency of the concerned NPCs out of whack.

In the reverse, it could also make the used-by-PCs saving throw abilities go out of whack, or at least feel like it. Especially when the PCs take care to target NPCs with saves they can be reasonably sure the NPCs are bad at.

Is smoothing the curve of NPCs bad? Bounded accuracy should kinda lend itself towards that, I thought.

If the idea is that "everything should be dangerous", it feels weird that at some point a CR2 Cult Fanatic's spells just have 0 chance of landing because their DC is only 11 and your Wissave is +10.

"Why are you fighting CR2 enemies in tier 3/4?" Because they're minions of a bigger boss, prime fireball fodder to make people feel good, but I still want them to do something while they're alive.

Hope I'm not coming across as too disagreeable here, just trying to lay out my thought process for it.

[Mastikator]

Is that an issue? Rules stemming from "this feels good/bad" is hardly a bad thing.

Yeah, since it accomplishes/harms/changes nothing besides making players feel better it might be worth the change.

[KorvinStarmast]

After DMing for a party of Barbarians who went up against some Mind Flayers, being permanently stunned out of the whole fight is very unfun. I dont care for any arguments, people are here to play D&D, and having 0 chance to break free just stops you from playing the game. Thats why I'll always advocate for auto fail/pass on 1/20s.

Actually i do have an argument. In a game where bounded accuracy is one of the main features, having 0% and 100% chance doesn't make sense.

If you choose to go out and fight magical monsters and eschew magical help (an arcane or divine caster) as a part of your party, you are leaning into your weakness as a strategy.
The game was built around a party of mixed talents and classes.

"Hi, we intentionally gimped the whole party, and when we got PWNED it stunk."

Yes. Of course it did. You may be better off Blaming the WotC saving thow system, (which unlike TSR system does not globally improve as you go up in level).
If the mindflayer had any minions, they will have made mincemeat out of your party's barbarians, just as the party's barbarian (me DM) in a group that I run tends to make mincemeat out of enemies who are stunned by the party monk. When those two get in synch, it's like watching blender.

[JNAProductions]

If you choose to go out and fight magical monsters and eschew magical help (an arcane or divine caster) as a part of your party, you are leaning into your weakness as a strategy.
The game was built around a party of mixed talents and classes.
"Hi, I intentionally gimped the whole party, and when we got PWNED it stunk.

Yes. Of course it did.
If the mindflayer had any minions, they will have made mincemeat out of your party's barbarians, just as the party's barbarian (me DM) in a group that I run tends to make mincemeat out of enemies who are stunned by the party monk.

How exactly do you expect the Barbarian to make better saves, even with casting help?Bless would help. A bit. But that's eating the Cleric's Concentration, when they might want to concentrate on another spell.
Barbarians have save issues-especially against fear effects. The solution to that is not "Partner with a Paladin to get a big boost to saves", the solution is "Give the Barbarian features that can let them power through at a cost, or something similar."

To the premise of the thread: I don't think adding autofails and autosuccesses is needed.
Saves generally shouldn't go past 19, ESPECIALLY on hard control abilities. That they do is an issue to be solved, not that you can sometimes always pass or fail a save. A "Knock Prone" effect could reasonably have a higher than 19 DC, since being knocked prone isn't hard control. But a domination effect? stun effect? even something weaker, like Mind Whip? Shouldn't have a DC exceeding 19, and even 19 is pretty high.

On the opposite end, autopassing. The main low DC save you might see at higher levels is DC 10 Constitution saves, for Concentration. Most other relevant saves will be 14+.
And achieving a +9 to saves is pretty hard. Using Point Buy, the earliest I can see achieving it is level 6, for a Variant Human Paladin with Resilient (Constitution) and a 16+ in Constitution and Charisma.
For a dedicated caster that can't guarantee standing next to a Paladin, earliest is level 9, assuming you beeline for Constitution 20 over your casting stat and have proficiency in Constitution saves from class or feat.
Realistically, I doubt you'd see anyone outside some Paladins be autopassing before T3. And once they CAN autopass DC 10 Concentration saves? They get to feel like a badass against weaker blows, knowing they can't possibly fail to keep their spell up.

[sithlordnergal]

After DMing for a party of Barbarians who went up against some Mind Flayers, being permanently stunned out of the whole fight is very unfun. I dont care for any arguments, people are here to play D&D, and having 0 chance to break free just stops you from playing the game. Thats why I'll always advocate for auto fail/pass on 1/20s.

Actually i do have an argument. In a game where bounded accuracy is one of the main features, having 0% and 100% chance doesn't make sense.

yeeeah, I'm with Korvin here...your example isn't really a good argument for making 1/20s an autopass/fail. It'd be like saying "All enemies need to have +10 to their attack bonus because my entire party has a base AC of 30". Is it possible for that situation to occur? Sure, I have an AL Character that has a base AC of 30 due to magic items and the Defense Fighting Style, and I can spike my AC to 37 with spells. Is that within the realm of "normal" gameplay? No, and you shouldn't make design choices based on such a potential event.

If you decide to fight creatures that target Int Saves with a party that has -1 to Int Saves, don't get surprised if you end up getting stomped. A similar thing happened to the Barbarian in my game. He knew he was in a position where the party couldn't reach him, he knew he was fighting a Mind Flayer, he knew he had a -1 to Int saves...and still got salty when his Barbarian was stunned for several rounds and nearly crit to death by the brain eating ability. That's not the fault of the game or the DM, that's on the player's poor planning.

How exactly do you expect the Barbarian to make better saves, even with casting help?Bless would help. A bit. But that's eating the Cleric's Concentration, when they might want to concentrate on another spell.
Barbarians have save issues-especially against fear effects. The solution to that is not "Partner with a Paladin to get a big boost to saves", the solution is "Give the Barbarian features that can let them power through at a cost, or something similar."

Actually...I do expect the Cleric to provide Bless. Especially if they notice the Barbarian hasn't made their save in a few turns. Hell, I do that as a Bard. If I notice someone is in need of some assistance, I'll drop concentration on a spell in order to give a buff, and toss a Bardic Inspiration. That's part of playing a support class like a Cleric.

[Beni-Kujaku]

Yeah, since it accomplishes/harms/changes nothing besides making players feel better it might be worth the change.

"It changes nothing but makes players feel better" is a really good thing to have in a game. It's not a highly competitive environment, we're just here to have fun.

Also, I'm having trouble imagining many scenarios where that would even occur. Even barbarians with Int 3 would save against the DC 15 mind flayer stun when rolling a 20. Even a level 20 wizard with 22 Int would fail that same save with a natural 1. If that situation occurs repeatedly, then the party is facing overwhelming odds unprepared, and in that case making a save with a nat 20 will not save them in the long run.

[Segev]

While I am not arguing one way or the other, it would be interesting to see if we could figure out WHY it was determined that attack rolls have auto fail/succeed on natural 1/20, but saving throws do not.

Ability checks not autofailing actually makes a lot of sense, as it provides a point where the characters specializing in something can truly just be cool all the time. But saving throws are distinct from ability checks for roughly the same reason that attack rolls are: they're a function primarily of combat, where the randomness is meant to be a little more pronounced. Not a lot, but a little. The question, then, of why saving throws share ability checks' lack of auto fail/success rather than attack rolls' presence thereof is an interesting one, in terms of design choices.

[Mindflayer_Inc]

While I am not arguing one way or the other, it would be interesting to see if we could figure out WHY it was determined that attack rolls have auto fail/succeed on natural 1/20, but saving throws do not.

Ability checks not autofailing actually makes a lot of sense, as it provides a point where the characters specializing in something can truly just be cool all the time. But saving throws are distinct from ability checks for roughly the same reason that attack rolls are: they're a function primarily of combat, where the randomness is meant to be a little more pronounced. Not a lot, but a little. The question, then, of why saving throws share ability checks' lack of auto fail/success rather than attack rolls' presence thereof is an interesting one, in terms of design choices.

I personally don't see any difference between being skilled in Athletics and doing a contest and being Proficient with a weapon and performing an attack.

I think it's one of those sacred cows that should be slaughtered.

Especially once you hit like 5th level. You can still miss but auto-miss? What? No, just no.

Those hordes of goblins will still be a threat even if you can reliably take one out per attack.

[stoutstien]

While I am not arguing one way or the other, it would be interesting to see if we could figure out WHY it was determined that attack rolls have auto fail/succeed on natural 1/20, but saving throws do not.

Ability checks not autofailing actually makes a lot of sense, as it provides a point where the characters specializing in something can truly just be cool all the time. But saving throws are distinct from ability checks for roughly the same reason that attack rolls are: they're a function primarily of combat, where the randomness is meant to be a little more pronounced. Not a lot, but a little. The question, then, of why saving throws share ability checks' lack of auto fail/success rather than attack rolls' presence thereof is an interesting one, in terms of design choices.

Honestly it's redundant. If a 1 would already miss or if a 20 would hit, it's unnecessary. Sure you *might* have those very extreme corner cases where only 20s can overcome AC but you are playing so far outside the expected curve you probably have much bigger issues to deal with.

The only thing it does is have a way to add crits in some form but it's a lot of text and rules for such a simple and low impact mechanic.

[Rukelnikov]

IMO bigotry is the only explanation

[sithlordnergal]

While I am not arguing one way or the other, it would be interesting to see if we could figure out WHY it was determined that attack rolls have auto fail/succeed on natural 1/20, but saving throws do not.

Ability checks not autofailing actually makes a lot of sense, as it provides a point where the characters specializing in something can truly just be cool all the time. But saving throws are distinct from ability checks for roughly the same reason that attack rolls are: they're a function primarily of combat, where the randomness is meant to be a little more pronounced. Not a lot, but a little. The question, then, of why saving throws share ability checks' lack of auto fail/success rather than attack rolls' presence thereof is an interesting one, in terms of design choices.

I feel like its a bit of a hold over from a previous edition. My 3.5 is rusty, but if I remember correctly, you had to make two rolls in order to see if you crit on a natural 20. The first nat 20 guarantees that you hit, regardless of the target's AC. Then you make a second attack roll using the same bonuses, and if that hits then you managed to crit the target.

[Jerrykhor]

If you choose to go out and fight magical monsters and eschew magical help (an arcane or divine caster) as a part of your party, you are leaning into your weakness as a strategy.
The game was built around a party of mixed talents and classes.

"Hi, we intentionally gimped the whole party, and when we got PWNED it stunk."

Yes. Of course it did. You may be better off Blaming the WotC saving thow system, (which unlike TSR system does not globally improve as you go up in level).
If the mindflayer had any minions, they will have made mincemeat out of your party's barbarians, just as the party's barbarian (me DM) in a group that I run tends to make mincemeat out of enemies who are stunned by the party monk. When those two get in synch, it's like watching blender.

My players didn't expect the Illithids, their appearance was a surprise. To be fair, it was a one shot session so they decided to have some fun with the party lineup. I had a bad feeling when they announced their decision, but i decided not to reveal the monsters they would face.

Sure, my example was not meant to prove that the houserule would have saved them (they made it through anyway). Its meant to be a feel good rule. The difference between 5% and 0% is that 5% is a possibility, while 0% is a certainty. In a game where rolling dice is 99% of the mechanics, knowing that your rolls dont matter feels bad. In fact, there should be MORE house rules to prevent anti-fun situations. Being stuck doing nothing for the entire fight is incredibly frustrating, whether due to bad luck or the rules.

I dont get the argument of 'if they break out of it, then its not scary'. Players dont feel scared when they are frightened/stunned/paralyzed, they feel bored. Nothing is more boring than the feeling of helplessness and the impending doom that is to come slowly but surely. And about the 5% failure thing... nobody complains about their 20 Dex character feeling bad, thats what Inspiration is for. I always feel that if you roll a Nat1 and then reroll another Nat1, its a story to tell. Every master can recall the time they fail a simple task, it happens.

In video game design, perma-stun exists and players tend to complain especially when it happens to them, and they only control one character. If you control multiple characters then its less obnoxious.

[Cheesegear]

Chances are if you want it, a caster can have Concentration guaranteed on most attacks by tier 3.

Okay.

Likewise, there are some higher CR enemies with DC21+ save features that some characters may be literally unable to pass, ever. One example is Asmodeus in...the Chains of Asmodeus. Now he's CR30, so he should be scary, but he's got a trio of features with DC24 saves - one targeting Wis, one Cha, and one Int (plus some other dex saves knocking around elsewhere).

Ideally your example of why something is bad, shouldn't be a campaign-ending God battle that will literally only happen once.

Would you prefer it if they were autopass/fail or not?

I don't even like that critical attacks exist.

A character with +11 can still roll a regular 13. Plus 11 makes 24. Not enough to hit a DC 25 check. If DM is dropping a DC 25 check, it's because either
a) The DM might expect the party to expend resources; Spell slots, magic item charges, etc. Or,
b) The DM is trying to reward the player for having Expertise, and creating a challenge that rewards their investment in their character. The fact that they failed on the first roll, might encourage them to try again, but differently. Or, having failed the first time, we go back to a). But, there was still a chance that the +11 could have worked...It just didn't. Dice are gonna dice.

However, if everyone has a flat 5% chance to succeed - even on the DC 21+ checks - then that means that everyone can roll on every check, forever.

Can I roll Arcana?
You're a Barbarian with -1 Int. I don't think you kn-
I already rolled and a I got a natural 20. Tell me what it is.
I didn't even say you c-
I always have a 5% chance of passing, I can always be lucky. You can never disallow me to roll on anything, ever. Because I can potentially always pass, and you have to give me that opportunity if it makes sense. And it always makes sense, because I can always be lucky.
I don't really think Arcana works that w-
I rolled a 20. I know I have -1 Int. But I pass the DC 25 Arcana check. What don't you understand.

Now, expand that to every player at the table. Now, depending on time, you can turn that into a fun roleplaying bit; How does the character with -1 Int pass the DC 25 challenge? Narrate it and come up with something now or you're a bad player. I demand that you retcon your entire backstory to incorporate this. The DM is throwing DC 21+ challenges so I can assume we're mid Tier 2, if not Tier 3, and how has this particular bit of something you know literally never come up before now?

Conversely, auto-pass, similarly, devalues what you try hard at, and encourages dogpiling. It also strongly encourages "Group Checks as Individuals."

Saying that you will always fail 5% of the time devalues actually trying to be good at something. A Level 8 character with Expertise has +11 to a Skill check. Actually **** you, you fail an Easy or Very Easy task. Not one in 400 times, not one in 100 times...1 in 20, you fail. At my employment, a flat 5% failure rate would get me fired.

You come across a locked door. Everyone roll a Dex check until someone passes. Not a group check. We just go 'round the table. We can start with the Rogue, sure. He might pass first go. But if the Rogue fails, then everyone rolls.

Now, the DM must come up with a reason for why if one character fails, the challenge is broken in that capacity. Otherwise, the "Everyone can pass" argument goes into effect and everyone rolls on everything all the time. This leads to horrifically slow gameplay because every skill challenge requires every character to roll, because everyone rolling, has the chance of rolling a 20.
"I'd like to roll a check, and if I fail, we go to our next solution." turns into
"I'll roll a check, and if I fail, then you roll, and if you fail, then you roll [...] until all of us have roll on the Skill check, then we go to the next solution."
The party can't move on until every roll has been exahausted.
...How exhausting.

Alternatively, instead of dogpiling, it merely turns every DC 21+ challenge into a group check.

I have +11 to Persuade, so I'd like to Persuade the Guard to do something that goes directly against both his orders and better judgement. The DC is 21 or more? Makes sense. I have +11. This is my jam.
Actually everyone roll. Some of you might roll 20s.
No but, it's just me. My +11 in Persuade means nothing because this group isn't good at it, which means half of us will fail, which means we'll fail the Persaude check, even though I have +11, and I've got the best shot at doing this by myself.
Nope. Group check. Everyone can pass, so everyone must roll. Just in case.

The only thing that matters is; Are there consequences to failing a DC 21+ check, and are those consequences worth not rolling at all? If the answer is "No" and/or "No", then every DC 21+ check is a group check. If your players are canny enough, they can expand this to be able to roll on anything, at any time. The DM can never moderate who can and can't make skill checks.

Conversely, every DC 5 check is also a group check, because everyone can roll a '1'.

[Amnestic]

Okay.



A character with +11 can still roll a regular 13.

It's a little weird how I acknowledged ability checks in my first post and both thread title and first post were about saving throws specifically, but you still wrote a lot about ability checks, and not saving throws.

[Rukelnikov]

snip

If its impossible for a character to do something, they don't have to roll.

If a character is rolling they need to have a chance for success.

[Mastikator]

If its impossible for a character to do something, they don't have to roll.

If a character is rolling they need to have a chance for success.

While you are technically correct (the best kind), the DM can't possibly keep all character's every skill modifier in mind to know whether a particularly difficult task is possibly for any given character. There may be a DC 30 check in the game, that might be possible to succeed in one party but not another. There are cases where what is possible for the party is not known to the DM.

Actually, I think in most cases the DM will only have a vague guess of what the party is capable off. The DM's best bet is to throw things at the wall and see what sticks.

[Segev]

I like that ability checks can get to the point that your bonus means you succeed on a 1. I even like it that some characters just can't succeed on some tasks; the Strength 8 gnome is never going to lift the DC 20 portcullis with his bare hands and brute strength alone.

I like that, in combat, even the Ultra-Kraken has a chance that his flailing tentacle (with its +45 to hit bonus) will miss the plucky octopus familiar with an attack (on a natural 1). I also like that even the bog-standard goblin with his shlrtbow can get lucky and land an arrow in the squishy part of a fold of the Ultra-Kraken's suckers, mildly irritating it (but enough of those will eventually make it reconsider the wisdom of the fight).

In 3.5, both of these (skill checks and attack rolls, respectively) behaved the same as they do in 5e, at least wrt auto fail/success on natural 1/20. In 3.5, saving throws also had auto fail/success. In 5e, saving throws do not automatically fail or succeed on a natural 1 or 20. A DC 20 Strength save to avoid a froghemoth's tongue's grab will never succeed if the target is an owl familiar with a strength of 8.

I think I actually prefer how 5e handles it, because (with the exception of Dexterity saves) the Saving Throw means you HAVE been 'hit' by the effect, and this is about your ability to resist it. The vagueries of combat that allow a freak miss or a lucky blow don't apply; if you're already grabbed or poisoned or mind-wham led, and you lack the strength, constitution, or force of will to stand up to the overwhelming force bearing down on you, you don't luckily prove immune to that particular dose of elephant tranquilizer that would LD50 a blue whale.

[Oramac]

Skipping a lot of the discussion, I'll just jump in and say I would fully support Saving Throws being a 1/20 autofail/success.

I have been in similar situations to the OP, and frankly, I don't give a damn what the math says. Knowing that your character has literally zero chance of doing anything is not fun. Worse, it puts all of the other characters/players in the horrible position of either fighting shorthanded or trying to come up with some way to help the impossible saving throw.