This rule really most strongly falls apart when there's unknown information involved.
Insight checks are often made opposed to a Deception check the DM makes in secret, even if they person being checked is telling the truth, so that the player doing the insight check doesn't learn something out-of-character that they didn't learn in character (if the DM only makes a Deception check when it's actually a lie, then the players know it's a lie even if the lying NPC succeeds in beating the Insight check with their Deception check). This way, the player doing the Insight check only learns whether their character thinks what the NPC said is true, but without magic they don't know whether it's actually true or whether the other person is just a really good liar.
But with this rule, if the player making the insight check gets a 20, they know for a fact that if they're told something is true, it's true.
Perception checks to detect traps are often made regardless of whether there are any traps and whether the character has a chance to detect them. If you just tell them 'there are no traps' when they ask to look for traps without making them roll, then they'll know for a fact that there are traps if you do make them roll. And if you don't let them roll because they have no chance to detect the trap then they will know there is a trap and that they can't detect where it is. It's really hard not to meta-game with that in the back of your head.
But with this rule, if the player rolling Perception to check for traps rolls a 20 and you say "You don't see any traps." they know for a fact that there are no traps.
And if you then still spring a trap on them (say, a trapdoor that causes them all to fall into a part of the dungeon that no one's ever been to and where they will find a magical artefact that will kick off the main plot), they will get upset at you for not telling them it was there even though they rolled a 20. Or if you do tell them it's there they'll avoid it and never kick off the main plot (unless you railroad them. Fun!)
A Stealth check is usually rolled against the passive perception of the creatures that one is trying to sneak past (or up on). The spell Pass Without Trace is therefore pretty powerful, as the +10 it gives to Stealth checks makes sneaking into an automatic success against the passive perceptions of most creatures (at least for any character with a positive Dex score).
But with this rule, even a level 10 Rogue with Expertise in Stealth and 20 Dex (for a total of +13 on Stealth) can still fail to sneak past a pack of Kobolds (passive perception: 8 !) after having Pass Without Trace.
Honestly though, the most important reason why this rule is bad is the reason why the Take 20 rule exists: If there's no time pressure and no price for failure, players will keep rolling until they succeed. And with 20 being an automatic success, the will always succeed (eventually)...
This rule is just going to lead to players constantly demanding to be allowed to reroll on any task that they didn't succeed at that isn't time sensitive.
It's going to lead to dejected DMs either allowing the party to succeed at absolutely everything that would require a roll without actually asking for a roll, because they'll just reroll until they get an automatic success or overstressed Players having to deal with their DM constantly putting time pressure on absolutely everything in order to prevent them from rerolling on everything until they automatically succeed.
'Automatic failure on 1, automatic success on 20' works as a house rule, because house rules are optional and can be taken away if they're being abused. But it doesn't work as a core rule, because then you have to deal with it in instances where players will abuse it. Like, say, when running an Adventurers League game with a new player who hasn't yet learned to treat D&D as a collaborative storytelling game, instead of treating it like an analogue video game they want to win.
Ok so first of all if players are insisting they re roll the same task over and over that is just poor DM management and table etiquette. If the DM intends the party to achieve the thing regardless then they should do so and the dice roll indicates only how long it took, for instance the massive boulder in the way of the party, the party have to overcome it because it is the only way to the dragon. That kind of a dead end situation where the DM expects the DC to be made but is then stumped when it is failed and the party can't go any further is bad DM planning.
I have no issue with a nat 1, yes that rogue with pass without a trace has a silly +massive stealth roll, and they roll a 1. There are 2 things here.
First of all if the whole party is sneaking then the right approach is a party wide test, so one failure does not mean the whole party fail, what it does allow is a bit of in character banter as the cleric in there plate armour is the one to grab the rogue and stop him making a ton of noise. The Rogue will never live that down, it's a great little moment and can then carry on for several sessions, I have seen these things help build a party chemistry.
Or the Rogue is on there own, they sneak up, they roll a 1, the Kobolds sense something, they are aware, but as a DM "auto failiure" does not have to mean they spin round and see him. Given there high + to stealth I may tell them that future stealth rolls are with disadvantage, I might tell them narratively that they wait, 10 mins, 20 mins for the Kobolds to settle back down, It depends on the situation, cover available, the actions the player takes, yes the roll has failed, but that doesn't always have to mean the Kobolds turn around see the character and attack. maybe it does make sense to have that happen, but usually a nat 1 roll simply makes things trickier and the player then has to think there way out of that situation.
An example recently from a real session.
Party was trying to get past some Skum in a chamber underground, The Tunnel out of the chamber was the other side opposite and 3 skum where patrolling the centre of the chamber. I set up the encounter expecting a combat but, the players decided to try and sneak past. A combination of illusionary sounds, stealthy rogue with Mage Hand throwing stones and dust of disappearance meant that the party of 7 crept there way through, all except the Rogue, his He rolled a double 1 but, I didn't then instantly have that mean that the skum knew where he was, he was invisible, they heard a noise from his direction, they moved that way to investigate, for a round they had advantage on the perception roll (so a straight roll) and he had disadvantage on his stealth rolls the next 2 rounds, not he was rolling straight, he was rolling at full disadvantage. He changed the direction he was moving, and did some things to cast distraction and daw the Skum away The players watched this with bated breath, from a resources point of view they wanted to avoid a fight. But there characters had no idea, after all, they where all invisible as well. That nat 1 made for a really tense moment in the game, something that would never have happened had I simply told the player "oh you auto sneak through don't bother rolling".
Now as for Insight etc, as a DM I never roll hidden, my players always see all my dice rolls, so they know when they have neaten an insight check, or made a deception check and, guess what, they are really good at roleplaying ignoring the dice roll. So, they might roll really high on Insight but, still doubt what I told them because it makes sense for the moment and there character. If players are metagaming dice rolls then there is a much much bigger table issue at play then Nat 20's and Nat 1's
Ok so first of all if players are insisting they re roll the same task over and over that is just poor DM management and table etiquette.
I don't disagree, but that doesn't mean it doesn't happen. In fact as far back as before 3rd edition it happened all the time, which is why 3rd edition introduced Taking 20: you take ten times as long as task would normally take and get the result you would get as if you had rolled a nat 20, which can still be too low to beat the DC you need, but at least you know that you didn't fail because of an unlucky low roll.
The fact of the matter is that when there is no time pressure on an action, it just makes sense for a character to keep trying, and therefore a player to keep rolling, until they're absolutely sure that they can't succeed. The Take 20 rule is there to facilitate that.
But if a 20 is an automatic success, you can't 'Take 20' if you would still fail on a 20, because rolling a 20 is an automatic success even if you wouldn't have a chance to succeed otherwise.
I know that a lot of people say "If it can't succeed, you shouldn't ask for a roll in the first place."
But the thing is... If that were the case the '20 is an automatic success' thing is pointless. If you can only ever roll if you have a chance to succeed, then of course a 20 is going to be a success, a 20 being a success would be the minimum requirement to be allowed to roll. If the modifiers made it impossible for you to succeed on a 20, then you wouldn't be allowed to make the roll that would let you roll that 20 for an automatic success that ignores those modifiers.
It's like including a rule that you automatically do maximum damage on your attack with a shortsword if you roll a 6 on your damage roll. Sure, it's true, but it would be true regardless of whether that was an explicitly listed rule or not.
So the very existence of this rule implies that the official position of WotC is that a DM should let players roll on things they would normally have no chance to succeed at, otherwise having this rule at all is just wasting space in the rulebook.
If players are metagaming dice rolls then there is a much much bigger table issue at play then Nat 20's and Nat 1's
Not really. It's something that's pretty much unavoidable.
Even your players metagame, if you pay attention. You'll find that if they roll really high on their Insight, it will affect how they deal with that information compared to if they rolled really low and therefore don't know if it's genuinely true out-of-character. They may roleplay still doubting the information, but their out-of-character decision making will definitely be affected.
You can see this at any table when you tell a player "There are no traps, as far as you know." on a low roll versus saying that on a high roll. Some players will be a lot more careful if they rolled low than if they rolled high. But other players will be far more reckless if they rolled low than if they rolled high (because they're trying not to metagame and overcompensating). Just because their reaction to the roll isn't the 'optimal' reaction from a 'playing to win' point of view, doesn't mean it's not still a reaction to out-of-character information.
I don't think the change is a huge deal. The rules in the PHB and DMG both just need to be really clear that IF there is no chance at success, or no chance at failure there shouldn't be a dice roll.
If a player has a 11 STR and no proficency in athletics, and tries to force open a door that requires a DC21 Athletics check, there shouldn't be a call for a d20 roll. The DM should just say they aren't strong enough to force it open. Or, if they just LOVE a dice roll, say "you're not going to be able force it open, but roll a d20 to see how cool you look while failing."
D&D One just needs to be very clear about it being fine for a DM not to allow a d20 roll sometimes. If both success and failure are not options, there's no need for a dice roll.
I know that a lot of people say "If it can't succeed, you shouldn't ask for a roll in the first place."
But the thing is... If that were the case the '20 is an automatic success' thing is pointless. If you can only ever roll if you have a chance to succeed, then of course a 20 is going to be a success, a 20 being a success would be the minimum requirement to be allowed to roll. If the modifiers made it impossible for you to succeed on a 20, then you wouldn't be allowed to make the roll that would let you roll that 20 for an automatic success that ignores those modifiers.
Finally, we've dug deep enough to strike the root. Extremely well articulated.
It's like including a rule that you automatically do maximum damage on your attack with a shortsword if you roll a 6 on your damage roll. Sure, it's true, but it would be true regardless of whether that was an explicitly listed rule or not.
Being charitable, I think maybe the thing is that the rule applies to all tests, but it's only meaningful for attacks and saves. In the interest of brevity or aesthetic consistency they've included ability checks as well, though it means nothing.
Even being charitable, we can disagree on whether that's a good idea -- though I suspect we don't. If you ask me, it's an attempt at clean simplicity that inadvertently results in more mess and more confusion.
Memories can be selective, but basic math tells us 1/20 rolls on average will be a 1. Unless the die you use is severely flawed, you should be rolling a 1 once every 20 rolls. That is not how it works. It isn't like if you roll a 20 the first time that 20 is removed from the probability the next roll and so on. The probability of rolling a number on a 20 sided die is 1/20. The next roll is 1 in 20. The roll after that is the same. A deck a cards is different if you take that result out. Goes 1/52 to 1/51 and down the line.
How would people feel if the auto success was lower on difficulty. Right now a test has to be on a 5 to 30. What if it was 5-20 or 5-25?
You could still roll but there is no chance for auto success. Then you can still say on some 30 DC you don't even roll. So the Bard trying to con the King is 30 or more.
I don't think anyone will ever change my mind that there is always a failure chance even for the best of the world. The Undefeated team can beat the team with no wins. The best cook in the world can burn a mean, the best ice skater in the world can slip on the ice and break their leg. Rare but happens.
5-30 is still not a great range for allowing rolls as 30 is a very low max cap and 5 is too low as a bottom cap. 10-40 would be a better coverage in my mind or a variable scale based on proficiencies or other factors.
There's also a psychological effect that you get when you say the top end is 40, which is that anyone who can't hit 40 feels impotent. I imagine this could be wrapped into explaining the proliferation of Expertise in recent years, but I'm too lazy to do the work.
Plus, it sneakily raises the average DC. The middle point between 10 and 40 is not the same thing as the average DC, but now it looks like it is.
If a player has a 11 STR and no proficency in athletics, and tries to force open a door that requires a DC21 Athletics check, there shouldn't be a call for a d20 roll. The DM should just say they aren't strong enough to force it open. Or, if they just LOVE a dice roll, say "you're not going to be able force it open, but roll a d20 to see how cool you look while failing."
I won't claim that's wrong, but it's not 100% correct.
Being a good DM for a given group can require you to work with or against their expectations out-of-character.
Let's say you have a door that's magically warded. The door itself is just a worn, slightly rickety looking wooden door, but it's spelled to such an extent that it's functionally indestructible and can only be opened by speaking the right command word.
With some players, not letting them roll and simply telling them "You try to force the door, but it's a lot sturdier than it looks" when they say they want to try to force it open is fine and they'll understand that you're saving everyone some time by not asking for a roll that's guaranteed to fail.
But that's not every player. Some players don't feel like their character is doing something unless they are doing something, even if it's as simple as rolling a die. And that feeling is valid. It can be that rolling is something they have agency over (even if the result isn't) when real life denies them agency just a little too often. It can be that part of how they learned to role play is to let the dice direct them and how they describe what their character does in situations like this depends on whether their result is high or low, regardless of whether it succeeds. It can simply be that since rolling a die is simply symbolic of their character doing something and it gives them the feeling that they're affecting the world and being part of the story.
If you tell that kind of player that you're not allowing them to roll, you're telling them that you're not allowing their character to interact with (that element of) the world. It makes them feel restricted and in the example of the door above it'll probably lead to them deciding that you're railroading them and don't want them going through that door. So they ignore it, instead of trying to find a different way of opening it after brute force (and picking the lock) fails.
But, you know, if a nat 20 is always automatically a success, then telling them they still don't succeed even after getting one feels like railroading too.
I know that a lot of people say "If it can't succeed, you shouldn't ask for a roll in the first place."
But the thing is... If that were the case the '20 is an automatic success' thing is pointless. If you can only ever roll if you have a chance to succeed, then of course a 20 is going to be a success, a 20 being a success would be the minimum requirement to be allowed to roll. If the modifiers made it impossible for you to succeed on a 20, then you wouldn't be allowed to make the roll that would let you roll that 20 for an automatic success that ignores those modifiers.
Finally, we've dug deep enough to strike the root. Extremely well articulated.
It's like including a rule that you automatically do maximum damage on your attack with a shortsword if you roll a 6 on your damage roll. Sure, it's true, but it would be true regardless of whether that was an explicitly listed rule or not.
Being charitable, I think maybe the thing is that the rule applies to all tests, but it's only meaningful for attacks and saves. In the interest of brevity or aesthetic consistency they've included ability checks as well, though it means nothing.
Even being charitable, we can disagree on whether that's a good idea -- though I suspect we don't. If you ask me, it's an attempt at clean simplicity that inadvertently results in more mess and more confusion.
It's definitely not pointless. It's just very clear. If the DM calls for a roll, and it's a 20, it's a success. If it's a 1, it's a fail.
It's still the DMs choice to call for or allow a roll. So in the example I gave above, if the DC was 21 to open the door but the character could only max out a 20, the DM could still allow a roll if they wanted to. And if they got a 20, it would succeed automatically. Everything is in the DMs discretion still. "The DM determines whether a d20 Test is warranted in any given circumstance." If they want to still give the player a shot even though it's not possible on paper, that's their choice to make. The DM just needs to be OK with the fact that if they're allowing a roll, there is a chance to succeed. So in those cases where the DM really feels like there is no chance at success, then "You fail and seducing the dragon, Mr. Bard." However, if they call for the roll and it's a 20, the 20 bypasses the difficulty modifiers that would turn what should be a failure into a success.
It's fine this way. It's fine the way it was before. It's not a huge change IMO. It will just need a slight shift in player and DM expectations. The benefit of the change is it results in the d20 feeling more impactful, while still giving the DM authority to disallow impossible acts. The downside is it will take some adjustment at first for some players (my table doesn't play this way currently), but as long as the DM and players are on the same page, that will resolve itself quickly.
If a player has a 11 STR and no proficency in athletics, and tries to force open a door that requires a DC21 Athletics check, there shouldn't be a call for a d20 roll. The DM should just say they aren't strong enough to force it open. Or, if they just LOVE a dice roll, say "you're not going to be able force it open, but roll a d20 to see how cool you look while failing."
I won't claim that's wrong, but it's not 100% correct.
Being a good DM for a given group can require you to work with or against their expectations out-of-character.
Let's say you have a door that's magically warded. The door itself is just a worn, slightly rickety looking wooden door, but it's spelled to such an extent that it's functionally indestructible and can only be opened by speaking the right command word.
With some players, not letting them roll and simply telling them "You try to force the door, but it's a lot sturdier than it looks" when they say they want to try to force it open is fine and they'll understand that you're saving everyone some time by not asking for a roll that's guaranteed to fail.
But that's not every player. Some players don't feel like their character is doing something unless they are doing something, even if it's as simple as rolling a die. And that feeling is valid. It can be that rolling is something they have agency over (even if the result isn't) when real life denies them agency just a little too often. It can be that part of how they learned to role play is to let the dice direct them and how they describe what their character does in situations like this depends on whether their result is high or low, regardless of whether it succeeds. It can simply be that since rolling a die is simply symbolic of their character doing something and it gives them the feeling that they're affecting the world and being part of the story.
If you tell that kind of player that you're not allowing them to roll, you're telling them that you're not allowing their character to interact with (that element of) the world. It makes them feel restricted and in the example of the door above it'll probably lead to them deciding that you're railroading them and don't want them going through that door. So they ignore it, instead of trying to find a different way of opening it after brute force (and picking the lock) fails.
But, you know, if a nat 20 is always automatically a success, then telling them they still don't succeed even after getting one feels like railroading too.
That's why knowing your players, being adaptable, and setting expectations is really important for the DM. If the table agrees they want it done another way, that's fine. Otherwise, just communicate very clearly the way you're going to play it, and explain that sometimes when a character tries something impossible, they don't get to roll a dice.
5-30 is still not a great range for allowing rolls as 30 is a very low max cap and 5 is too low as a bottom cap. 10-40 would be a better coverage in my mind or a variable scale based on proficiencies or other factors.
Nah, a range of 5-30 is plenty good. A 20th-level character with a maxed statistic and expertise only has +17 bonus to their die roll. This means they need a 13 on the d20. That's a 40% chance of success; upped to 64% if they have advantage. But most players won't get that far, so it's fine for extreme challenges. A DC 40 check would be damn near insurmountable. You'd need specific magic items to overcome that, which isn't a guarantee.
30 is not a good cap for investigative or other contested stuff though. As it's pretty easy for stealth, deception and sleight of hand to go way over 30 so not allowing people to roll to check because the dc is over 30 is weird. Also lowering the dc down to 30 would be wrong as well in mind.
5-30 is still not a great range for allowing rolls as 30 is a very low max cap and 5 is too low as a bottom cap. 10-40 would be a better coverage in my mind or a variable scale based on proficiencies or other factors.
Nah, a range of 5-30 is plenty good. A 20th-level character with a maxed statistic and expertise only has +17 bonus to their die roll. This means they need a 13 on the d20. That's a 40% chance of success; upped to 64% if they have advantage. But most players won't get that far, so it's fine for extreme challenges. A DC 40 check would be damn near insurmountable. You'd need specific magic items to overcome that, which isn't a guarantee.
Pass Without Trace adds a flat +10 to stealth rolls, no magic items needed. Throw in some Bardic Inspiration, Guidance, and a few other bonuses and you can pretty cleanly hit stealth rolls in the 50+ range.
Pass Without Trace adds a flat +10 to stealth rolls, no magic items needed. Throw in some Bardic Inspiration, Guidance, and a few other bonuses and you can pretty cleanly hit stealth rolls in the 50+ range.
I expect pass without trace to become "You have advantage on stealth rolls, and opponents have disadvantage on perception rolls" -- i.e. identical to a cloak of elvenkind.
5-30 is still not a great range for allowing rolls as 30 is a very low max cap and 5 is too low as a bottom cap. 10-40 would be a better coverage in my mind or a variable scale based on proficiencies or other factors.
Nah, a range of 5-30 is plenty good. A 20th-level character with a maxed statistic and expertise only has +17 bonus to their die roll. This means they need a 13 on the d20. That's a 40% chance of success; upped to 64% if they have advantage. But most players won't get that far, so it's fine for extreme challenges. A DC 40 check would be damn near insurmountable. You'd need specific magic items to overcome that, which isn't a guarantee.
Pass Without Trace adds a flat +10 to stealth rolls, no magic items needed. Throw in some Bardic Inspiration, Guidance, and a few other bonuses and you can pretty cleanly hit stealth rolls in the 50+ range.
Most Dexterity (Stealth)checks don't need to get that high. After all, you're rolling against someone else's Wisdom (Perception).
And that's only good for one person. The whole party can't benefit from guidance, so now you're looking at a contrived scenario where only one person needs to make this check no one else could possibly accomplish. It reeks of poor design.
How about picking a lock on a door? That's a Dexterity check with Thieves' tools. Some [magicitem]gloves of thievery[/item] will help, with a +5 bonus, but then you need Expertise and a high Dexterity. The best you can hope for is +22 before other bonuses come into play. And, again, that's for a high-level campaign. Most adventures don't need that. And if the idea is to have a DC where only very few people could ever hope to accomplish it, and only through the application of highly-specific magic, then...again...bad design.
It doesn't matter if a 50 or more is possible. You want it to be likely. You want your players to succeed. And they shouldn't need to be specialized to succeed. Yes, specialization can and should be rewarded. And the system will take care of that on its own. It does so already without gatekeeping.
I would disagree the upper end of dcs really shouldn't be achievable by most people. That's the point if the dc is high enough that you don't allow a roll then you're saying it is impossible. Dc 30 is way too low and achievable to be believable as an impossible dc. 40-50 is way more out of reach and believable as an impossible dc that people shouldn't even roll for.
Memories can be selective, but basic math tells us 1/20 rolls on average will be a 1. Unless the die you use is severely flawed, you should be rolling a 1 once every 20 rolls. That is not how it works. It isn't like if you roll a 20 the first time that 20 is removed from the probability the next roll and so on. The probability of rolling a number on a 20 sided die is 1/20. The next roll is 1 in 20. The roll after that is the same. A deck a cards is different if you take that result out. Goes 1/52 to 1/51 and down the line.
How I described it is how it works. Note that I said "on average." If you rolled an infinite number of times, you will see a 1 appear 20% of the time. Never once did I say that a number is removed if you rolled it.
1 in 20 or 5% is how often you roll a Nat 1, assuming a large population of rolls. It is not that uncommon or unlikely, actually quite a bit more than how often it feels like it appears. That is partly why I don't like this auto fail/success rule to apply to Saving Throws and Ability Checks. 1/20 is actually quite significant statistically speaking. If you have the modifier to succeed on a Nat 1 you should be allowed to do so.
Most of the time the auto fail on Nat 1 is meaningless as usually a 1 is a failure due to modifiers generally not being enough to succeed. However, if someone actually does put the effort and investment to get the modifier that high, they should be allowed to succeed on a nat 1.
Maybe this should be something done on a per play opt-in/out basis. Where each player decides if Nat 1/20 are auto fail/success for them. Downside with this though is that it requires the DM to keep track of who opt in/out.
However, I don't believe that everyone should be subject to a 5% auto fail chance if their modifier is enough to succeed on a nat 1.
Memories can be selective, but basic math tells us 1/20 rolls on average will be a 1. Unless the die you use is severely flawed, you should be rolling a 1 once every 20 rolls. That is not how it works. It isn't like if you roll a 20 the first time that 20 is removed from the probability the next roll and so on. The probability of rolling a number on a 20 sided die is 1/20. The next roll is 1 in 20. The roll after that is the same. A deck a cards is different if you take that result out. Goes 1/52 to 1/51 and down the line.
How I described it is how it works. Note that I said "on average." If you rolled an infinite number of times, you will see a 1 appear 20% of the time. Never once did I say that a number is removed if you rolled it.
1 in 20 or 5% is how often you roll a Nat 1, assuming a large population of rolls. It is not that uncommon or unlikely, actually quite a bit more than how often it feels like it appears. That is partly why I don't like this auto fail/success rule to apply to Saving Throws and Ability Checks. 1/20 is actually quite significant statistically speaking. If you have the modifier to succeed on a Nat 1 you should be allowed to do so.
Most of the time the auto fail on Nat 1 is meaningless as usually a 1 is a failure due to modifiers generally not being enough to succeed. However, if someone actually does put the effort and investment to get the modifier that high, they should be allowed to succeed on a nat 1.
Maybe this should be something done on a per play opt-in/out basis. Where each player decides if Nat 1/20 are auto fail/success for them. Downside with this though is that it requires the DM to keep track of who opt in/out.
However, I don't believe that everyone should be subject to a 5% auto fail chance if their modifier is enough to succeed on a nat 1.
That isn't how it works. Roll 1d20 there is 20 different permeations that can happen. Next die roll there is another 20 different permeations. The chance to roll a one doesn't go up or down. Roll it again, same thing. The chance to roll a 1 is the same as every other number.
I just don't get why people are so upset about failure. Again, the best at something can still fail at something. Even if it is simple. Again, my Rogue with a +10 has failed. I am OK with it. It happens. I laugh it off out of game and in game he will often laugh it off.
Note I am just using the rule they wrote in the test.
10 being and average difficulty and 30 being near impossible Totally OK with changing the actually number DC but saying that the range for rolling should be Easy to Hard for auto success or failure. Anything below Easy should not have an auto failure and anything above should not be auto success.
I also believe in variable scale. Then of course there is the what is failure and can you try again. A weaker character may need more than one hit to break down a door than a stronger.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Ok so first of all if players are insisting they re roll the same task over and over that is just poor DM management and table etiquette. If the DM intends the party to achieve the thing regardless then they should do so and the dice roll indicates only how long it took, for instance the massive boulder in the way of the party, the party have to overcome it because it is the only way to the dragon. That kind of a dead end situation where the DM expects the DC to be made but is then stumped when it is failed and the party can't go any further is bad DM planning.
I have no issue with a nat 1, yes that rogue with pass without a trace has a silly +massive stealth roll, and they roll a 1. There are 2 things here.
First of all if the whole party is sneaking then the right approach is a party wide test, so one failure does not mean the whole party fail, what it does allow is a bit of in character banter as the cleric in there plate armour is the one to grab the rogue and stop him making a ton of noise. The Rogue will never live that down, it's a great little moment and can then carry on for several sessions, I have seen these things help build a party chemistry.
Or the Rogue is on there own, they sneak up, they roll a 1, the Kobolds sense something, they are aware, but as a DM "auto failiure" does not have to mean they spin round and see him. Given there high + to stealth I may tell them that future stealth rolls are with disadvantage, I might tell them narratively that they wait, 10 mins, 20 mins for the Kobolds to settle back down, It depends on the situation, cover available, the actions the player takes, yes the roll has failed, but that doesn't always have to mean the Kobolds turn around see the character and attack. maybe it does make sense to have that happen, but usually a nat 1 roll simply makes things trickier and the player then has to think there way out of that situation.
An example recently from a real session.
Party was trying to get past some Skum in a chamber underground, The Tunnel out of the chamber was the other side opposite and 3 skum where patrolling the centre of the chamber. I set up the encounter expecting a combat but, the players decided to try and sneak past.
A combination of illusionary sounds, stealthy rogue with Mage Hand throwing stones and dust of disappearance meant that the party of 7 crept there way through, all except the Rogue, his He rolled a double 1 but, I didn't then instantly have that mean that the skum knew where he was, he was invisible, they heard a noise from his direction, they moved that way to investigate, for a round they had advantage on the perception roll (so a straight roll) and he had disadvantage on his stealth rolls the next 2 rounds, not he was rolling straight, he was rolling at full disadvantage. He changed the direction he was moving, and did some things to cast distraction and daw the Skum away The players watched this with bated breath, from a resources point of view they wanted to avoid a fight. But there characters had no idea, after all, they where all invisible as well. That nat 1 made for a really tense moment in the game, something that would never have happened had I simply told the player "oh you auto sneak through don't bother rolling".
Now as for Insight etc, as a DM I never roll hidden, my players always see all my dice rolls, so they know when they have neaten an insight check, or made a deception check and, guess what, they are really good at roleplaying ignoring the dice roll. So, they might roll really high on Insight but, still doubt what I told them because it makes sense for the moment and there character. If players are metagaming dice rolls then there is a much much bigger table issue at play then Nat 20's and Nat 1's
I don't disagree, but that doesn't mean it doesn't happen. In fact as far back as before 3rd edition it happened all the time, which is why 3rd edition introduced Taking 20: you take ten times as long as task would normally take and get the result you would get as if you had rolled a nat 20, which can still be too low to beat the DC you need, but at least you know that you didn't fail because of an unlucky low roll.
The fact of the matter is that when there is no time pressure on an action, it just makes sense for a character to keep trying, and therefore a player to keep rolling, until they're absolutely sure that they can't succeed. The Take 20 rule is there to facilitate that.
But if a 20 is an automatic success, you can't 'Take 20' if you would still fail on a 20, because rolling a 20 is an automatic success even if you wouldn't have a chance to succeed otherwise.
I know that a lot of people say "If it can't succeed, you shouldn't ask for a roll in the first place."
But the thing is... If that were the case the '20 is an automatic success' thing is pointless. If you can only ever roll if you have a chance to succeed, then of course a 20 is going to be a success, a 20 being a success would be the minimum requirement to be allowed to roll. If the modifiers made it impossible for you to succeed on a 20, then you wouldn't be allowed to make the roll that would let you roll that 20 for an automatic success that ignores those modifiers.
It's like including a rule that you automatically do maximum damage on your attack with a shortsword if you roll a 6 on your damage roll. Sure, it's true, but it would be true regardless of whether that was an explicitly listed rule or not.
So the very existence of this rule implies that the official position of WotC is that a DM should let players roll on things they would normally have no chance to succeed at, otherwise having this rule at all is just wasting space in the rulebook.
Not really. It's something that's pretty much unavoidable.
Even your players metagame, if you pay attention. You'll find that if they roll really high on their Insight, it will affect how they deal with that information compared to if they rolled really low and therefore don't know if it's genuinely true out-of-character. They may roleplay still doubting the information, but their out-of-character decision making will definitely be affected.
You can see this at any table when you tell a player "There are no traps, as far as you know." on a low roll versus saying that on a high roll. Some players will be a lot more careful if they rolled low than if they rolled high. But other players will be far more reckless if they rolled low than if they rolled high (because they're trying not to metagame and overcompensating). Just because their reaction to the roll isn't the 'optimal' reaction from a 'playing to win' point of view, doesn't mean it's not still a reaction to out-of-character information.
I don't think the change is a huge deal. The rules in the PHB and DMG both just need to be really clear that IF there is no chance at success, or no chance at failure there shouldn't be a dice roll.
If a player has a 11 STR and no proficency in athletics, and tries to force open a door that requires a DC21 Athletics check, there shouldn't be a call for a d20 roll. The DM should just say they aren't strong enough to force it open. Or, if they just LOVE a dice roll, say "you're not going to be able force it open, but roll a d20 to see how cool you look while failing."
D&D One just needs to be very clear about it being fine for a DM not to allow a d20 roll sometimes. If both success and failure are not options, there's no need for a dice roll.
Finally, we've dug deep enough to strike the root. Extremely well articulated.
Being charitable, I think maybe the thing is that the rule applies to all tests, but it's only meaningful for attacks and saves. In the interest of brevity or aesthetic consistency they've included ability checks as well, though it means nothing.
Even being charitable, we can disagree on whether that's a good idea -- though I suspect we don't. If you ask me, it's an attempt at clean simplicity that inadvertently results in more mess and more confusion.
How would people feel if the auto success was lower on difficulty. Right now a test has to be on a 5 to 30.
What if it was 5-20 or 5-25?
You could still roll but there is no chance for auto success. Then you can still say on some 30 DC you don't even roll. So the Bard trying to con the King is 30 or more.
I don't think anyone will ever change my mind that there is always a failure chance even for the best of the world. The Undefeated team can beat the team with no wins.
The best cook in the world can burn a mean, the best ice skater in the world can slip on the ice and break their leg.
Rare but happens.
5-30 is still not a great range for allowing rolls as 30 is a very low max cap and 5 is too low as a bottom cap. 10-40 would be a better coverage in my mind or a variable scale based on proficiencies or other factors.
There's also a psychological effect that you get when you say the top end is 40, which is that anyone who can't hit 40 feels impotent. I imagine this could be wrapped into explaining the proliferation of Expertise in recent years, but I'm too lazy to do the work.
Plus, it sneakily raises the average DC. The middle point between 10 and 40 is not the same thing as the average DC, but now it looks like it is.
I won't claim that's wrong, but it's not 100% correct.
Being a good DM for a given group can require you to work with or against their expectations out-of-character.
Let's say you have a door that's magically warded. The door itself is just a worn, slightly rickety looking wooden door, but it's spelled to such an extent that it's functionally indestructible and can only be opened by speaking the right command word.
With some players, not letting them roll and simply telling them "You try to force the door, but it's a lot sturdier than it looks" when they say they want to try to force it open is fine and they'll understand that you're saving everyone some time by not asking for a roll that's guaranteed to fail.
But that's not every player. Some players don't feel like their character is doing something unless they are doing something, even if it's as simple as rolling a die. And that feeling is valid.
It can be that rolling is something they have agency over (even if the result isn't) when real life denies them agency just a little too often.
It can be that part of how they learned to role play is to let the dice direct them and how they describe what their character does in situations like this depends on whether their result is high or low, regardless of whether it succeeds.
It can simply be that since rolling a die is simply symbolic of their character doing something and it gives them the feeling that they're affecting the world and being part of the story.
If you tell that kind of player that you're not allowing them to roll, you're telling them that you're not allowing their character to interact with (that element of) the world. It makes them feel restricted and in the example of the door above it'll probably lead to them deciding that you're railroading them and don't want them going through that door. So they ignore it, instead of trying to find a different way of opening it after brute force (and picking the lock) fails.
But, you know, if a nat 20 is always automatically a success, then telling them they still don't succeed even after getting one feels like railroading too.
It's definitely not pointless. It's just very clear. If the DM calls for a roll, and it's a 20, it's a success. If it's a 1, it's a fail.
It's still the DMs choice to call for or allow a roll. So in the example I gave above, if the DC was 21 to open the door but the character could only max out a 20, the DM could still allow a roll if they wanted to. And if they got a 20, it would succeed automatically. Everything is in the DMs discretion still. "The DM determines whether a d20 Test is warranted in any given circumstance." If they want to still give the player a shot even though it's not possible on paper, that's their choice to make. The DM just needs to be OK with the fact that if they're allowing a roll, there is a chance to succeed. So in those cases where the DM really feels like there is no chance at success, then "You fail and seducing the dragon, Mr. Bard." However, if they call for the roll and it's a 20, the 20 bypasses the difficulty modifiers that would turn what should be a failure into a success.
It's fine this way. It's fine the way it was before. It's not a huge change IMO. It will just need a slight shift in player and DM expectations. The benefit of the change is it results in the d20 feeling more impactful, while still giving the DM authority to disallow impossible acts. The downside is it will take some adjustment at first for some players (my table doesn't play this way currently), but as long as the DM and players are on the same page, that will resolve itself quickly.
That's why knowing your players, being adaptable, and setting expectations is really important for the DM. If the table agrees they want it done another way, that's fine. Otherwise, just communicate very clearly the way you're going to play it, and explain that sometimes when a character tries something impossible, they don't get to roll a dice.
Nah, a range of 5-30 is plenty good. A 20th-level character with a maxed statistic and expertise only has +17 bonus to their die roll. This means they need a 13 on the d20. That's a 40% chance of success; upped to 64% if they have advantage. But most players won't get that far, so it's fine for extreme challenges. A DC 40 check would be damn near insurmountable. You'd need specific magic items to overcome that, which isn't a guarantee.
30 is not a good cap for investigative or other contested stuff though. As it's pretty easy for stealth, deception and sleight of hand to go way over 30 so not allowing people to roll to check because the dc is over 30 is weird. Also lowering the dc down to 30 would be wrong as well in mind.
Pass Without Trace adds a flat +10 to stealth rolls, no magic items needed. Throw in some Bardic Inspiration, Guidance, and a few other bonuses and you can pretty cleanly hit stealth rolls in the 50+ range.
I expect pass without trace to become "You have advantage on stealth rolls, and opponents have disadvantage on perception rolls" -- i.e. identical to a cloak of elvenkind.
Most Dexterity (Stealth)checks don't need to get that high. After all, you're rolling against someone else's Wisdom (Perception).
And that's only good for one person. The whole party can't benefit from guidance, so now you're looking at a contrived scenario where only one person needs to make this check no one else could possibly accomplish. It reeks of poor design.
How about picking a lock on a door? That's a Dexterity check with Thieves' tools. Some [magicitem]gloves of thievery[/item] will help, with a +5 bonus, but then you need Expertise and a high Dexterity. The best you can hope for is +22 before other bonuses come into play. And, again, that's for a high-level campaign. Most adventures don't need that. And if the idea is to have a DC where only very few people could ever hope to accomplish it, and only through the application of highly-specific magic, then...again...bad design.
It doesn't matter if a 50 or more is possible. You want it to be likely. You want your players to succeed. And they shouldn't need to be specialized to succeed. Yes, specialization can and should be rewarded. And the system will take care of that on its own. It does so already without gatekeeping.
I would disagree the upper end of dcs really shouldn't be achievable by most people. That's the point if the dc is high enough that you don't allow a roll then you're saying it is impossible. Dc 30 is way too low and achievable to be believable as an impossible dc. 40-50 is way more out of reach and believable as an impossible dc that people shouldn't even roll for.
How I described it is how it works. Note that I said "on average." If you rolled an infinite number of times, you will see a 1 appear 20% of the time. Never once did I say that a number is removed if you rolled it.
1 in 20 or 5% is how often you roll a Nat 1, assuming a large population of rolls. It is not that uncommon or unlikely, actually quite a bit more than how often it feels like it appears. That is partly why I don't like this auto fail/success rule to apply to Saving Throws and Ability Checks. 1/20 is actually quite significant statistically speaking. If you have the modifier to succeed on a Nat 1 you should be allowed to do so.
Most of the time the auto fail on Nat 1 is meaningless as usually a 1 is a failure due to modifiers generally not being enough to succeed. However, if someone actually does put the effort and investment to get the modifier that high, they should be allowed to succeed on a nat 1.
Maybe this should be something done on a per play opt-in/out basis. Where each player decides if Nat 1/20 are auto fail/success for them. Downside with this though is that it requires the DM to keep track of who opt in/out.
However, I don't believe that everyone should be subject to a 5% auto fail chance if their modifier is enough to succeed on a nat 1.
That isn't how it works.
Roll 1d20 there is 20 different permeations that can happen. Next die roll there is another 20 different permeations. The chance to roll a one doesn't go up or down.
Roll it again, same thing. The chance to roll a 1 is the same as every other number.
I just don't get why people are so upset about failure. Again, the best at something can still fail at something. Even if it is simple.
Again, my Rogue with a +10 has failed. I am OK with it. It happens. I laugh it off out of game and in game he will often laugh it off.
Note I am just using the rule they wrote in the test.
10 being and average difficulty and 30 being near impossible
Totally OK with changing the actually number DC but saying that the range for rolling should be Easy to Hard for auto success or failure.
Anything below Easy should not have an auto failure and anything above should not be auto success.
I also believe in variable scale.
Then of course there is the what is failure and can you try again. A weaker character may need more than one hit to break down a door than a stronger.