My group and I are rookies to D&D and are learning and researching as we go. This question has to do with members of a party making their own perception checks, one at a time over and over again. The same confusion I have applies to any ability check used over and over again by one player or by each player in the party.
In essence, my question is can players try an ability check an infinite amount of times without consequence until it successfully lands?
Example:
Last night when exploring through a mansion basement, I scouted ahead alone and failed a perception check, which meant I missed a door to another room. My buddy saw that our DM's makeshift map revealed a sliver of what looked like a door and said that he wanted to go back to where I was and try a perception attempt of his own. He's not high in perception so he failed too. He wanted to roll another perception check after failing one already because it was clear on the DM's physical map that there was a room beyond the hallway where we both failed a check to see a door, but since the check failed, he returned to the party. That seemed clear enough at the time so we moved on as if we didn't find a door to another room we clearly all knew was there.
However later in the game, we found ourselves in a room full of loot and a shabby, weak treasure chest. My same buddy (Paladin) didn't want to wait for our Rogue to pick the lock and wanted to instead just smash the chest open, which our DM said didn't seem to be protected by magic or particularly physically strong. So he made s strength check to smash it open. Bad roll and it failed. Then he wanted to take another smash, our rookie DM let him, bad roll and that failed too. He wanted to take a third smash, was allowed by our DM and he finally succeed in a strength check to smash the chest open. This seemed like cheating and possibly game breaking but in reality is something one would try to do and was understanable.
My question is, are there rules or etiquete to stop a player from repeatedly smashing a chest as my friend, after each failed throw, over and over again until he achieves what he intends to? Is this solely up to the DM? The same question could be applied to the failed perception check I explained further above and really any ability check of this nature.
What's to stop a player or a party from attempting an infinite amount of check throws to achieve something like lock picking a chest, searching and searching and searching until you finally succeed on any throw to achieve your objective? I feel this has far reaching consequences or advantages in the game if not understood properly.
The simple answer to your main question is: If there's a chance of success and no risk due to failure, don't roll, give it to them. If there is no chance of success, don't roll, explain that it is impossible.
---
You walked by the door, you failed your Perception, you moved on. As a DM I always look at the Passive Perception score first, if it's high enough then a roll isn't necessary, you've already noticed the door.
Your friend did a no no and used knowledge that he was never supposed to have to influence a decision to go back and check for that door again. As a DM I would have asked what prompted him to make this decision, if the answer was justified by an in-game reason I'd allow it. If the answer was based on the fact that I asked for a check from you or, in this case, he felt there was a door there for "reasons", I probably wouldn't allow the check. If I was feeling generous I'd let him walk back to the location and describe him looking for the door and finding nothing, without rolling.
This is a check that has a risk due to failure: You don't find the secret door which leads to something. It is a valid check.
---
In the case of smashing the chest, I wouldn't have had the player roll to see if they could smash the chest open, instead I'd have them roll to see if they damaged anything in the chest. The player could attack that chest all day long without any risk, so they could roll indefinitely until they got a successful roll, as seen by your example. However, they're using a weapon to smash it open, if there were any valuables in there that could be damaged, well a weapon would surely cause damage to them. A high enough STR check would result in the chest being smashed open and some of the treasure being damaged by the blow, a lower STR check would result in the chest bursting open and the treasure being unharmed.
In instances like this, where success is eventually inevitable, without risk, the DM should measure success in different ways. The amount of time it takes, or perhaps a mishap, causing you to succeed at a cost.
If you are looking for a door that is obviously there, you should just find it.
If it's hard to spot, but a thorough search will reveal it, maybe it takes a while.
If you see it on the DM's map, YOU ARE A FILTHY CHEATER AND DESERVE TO STEP ON A PRESSURE PLATE DROPPING A ROCK ON YOUR SCUMMY HEAD.
Moving on.
If you smash open a chest instead of picking it open, perhaps it contained potions among its valuable wealth which shatter due to your inelegant attempts to smash it open, thereby ruining some of the reward.
This kind of method allows your roll to measure success without causing the game to come to a stop until someone eventually rolls high enough.
Use Success at a Cost, and Failing Foreward to keep the story rolling merrily along.
I don't think there's a hard and fast rule, but the players should really try to avoid metagaming. If the character fails a Wisdom (Perception) check to spot something, they don't really know they failed, so there should usually be no reason to try again. On the other hand, if they try to break something open with a Strength (Athletics?) check, and fail to, they know they failed, so there's no reason why they shouldn't want to try again. Bear in mind, though, that each attempt takes time. If time's not really an issue, and it's a check they can succeed at, an option is to just not have the players roll, and just tell them they eventually succeeded. From a realism point of view... if they have all the time in world to attempt something they can conceivably do, they'll eventually get it, so there's no point in checking whether they do. If time's an issue, though... then go ahead and have them roll multiple times, and make sure they're aware how long they've been at it.
But regarding unwarranted repeated attempts (like checks to spot hidden doors, for example), I'd say talk to the players about why their characters are trying again, when they don't know they actually failed their attempt (for all they know, their check succeeded and they found nothing).
Also, regarding multiple characters checking for the same thing, depending on the situation, it might be better handled by considering it a Help action, instead of a separate check. If two players are looking in opposite directions to try to spot hidden enemies, you can treat that as two separate checks, but if they're both trying to force open a door, that's really one of them doing so with Help from the other.
Also, when having players make iterative rolls to attempt to find a needle in a hay stack, Curse of Strahd had a mechanic I liked.
Roll Percentile Dice, with maybe a 15% chance of success. Each time they fail, it takes an hour of search time, but increase the chance of success +10%, so on the second roll, 25%, third, 40%, etc.
This ensures eventual success, and an expected time for it to happen. This is only useful though if they are on a time limit.
This is all really helpful guys. I'm given a lot of important considerations here from you guys on how to perceive the base imagination aspects of this game, which is why it's so unbelievably cool and enduring. Thanks.
I like how it really tasks and empowers the DM to continue to guide a influence the action of the characters and gameplay. Our DM totally has this in his ability, he's cut out for it. But he's new and probably a little timid disrupting a players ideas and attempts so much in our first handful of sessions.
The simple answer to your main question is: If there's a chance of success and no risk due to failure, don't roll, give it to them. If there is no chance of success, don't roll, explain that it is impossible.
I especially like this consideration about Passive abilities. I'm finding our gameplay is going REALLY slow these first few sessions and I was having a conversation with our DM about how we could pick up the pace of our sessions considerably. I believe having the narrative of the DM just simply award action to players due to their high Passive qualities is a good way to cut extra and unnecessary roll time and walking about and help guide us a little bit more. Really interesting.
Except for a few cases the rules define for handling passive Perception and stealth, there's no right or wrong way to use ability checks. They're a tool for DMs to model situations in the story where there's a chance of failure. Different situations can be modeled in different ways. Some allow for retries, some don't.
Smashing open a chest is one of those cases where you can keep trying until it finally breaks or you give up. In those situations the DMG recommends to just skip ahead and assume they succeeded after 10 tries (assuming there was a chance of success in the first place.) That's exactly how I use Perception in my games when players are exploring: I let them examine small objects, containers, or 5'x5' chunks of wall or floor in 1 minute (10 rounds of taking the Search action.) If they're in a time-sensitive situation (like Combat), then I care about the rolls and exactly how many rounds it takes them to succeed.
On the other hand, if someone's trying to fast-talk an NPC and the NPC sees through their ruse, they're not going to be able to try the same lie again.
That said, there's one way of handling ability checks that I'm not particularly fond of, and that's allowing 1 check per character for things where you should be able to try until you succeed. This leads to absurd situations where the 18 Strength barbarian rolls once to break down the door, fails, and the DM declares the door unbreakable by the barbarian; then the 10 Strength wizard tries, rolls a 20, and breaks it down.
But regarding unwarranted repeated attempts (like checks to spot hidden doors, for example), I'd say talk to the players about why their characters are trying again, when they don't know they actually failed their attempt (for all they know, their check succeeded and they found nothing).
I always point out that a standard Perception Check is literally 6 seconds. Character spending more than 6 seconds on something is not unreasonable, but then that gets back to DMThac0's point about lack of consequences. I like the Curse of Strad rule Cecilao, but I think 1 hour is waaaaaayyyyy to long. 1-10 minutes is more appropriate. It's also why I like have Passive being the floor for anything that isn't active contested.
I really liked the 3.5's "Take a 10 and Take a 20". Our standard procedure if we weren't pressured for time was. Make a Perception check to see what we get. Take a 10 (that's 1 minute of time) Take a 20 (that's 2 minutes of time) As long as you were not worried about time, 1-2 minutes to search a room to do the through job isn't that big of a deal. Hell, 1-2 minutes per room isn't even THAT much time as long as orcs aren't banging down the door trying to kill you.
As for smashing the chest, I'm all for disabling chest in not the standard ways. ie: take off the hinges instead of the lock, don't stand IN FRONT of the chest when you open it. If you break the chest and smash the loot, you'll quickly earn yourself on the party's short list.
In the case of smashing the chest, I wouldn't have had the player roll to see if they could smash the chest open, instead I'd have them roll to see if they damaged anything in the chest. The player could attack that chest all day long without any risk, so they could roll indefinitely until they got a successful roll, as seen by your example. However, they're using a weapon to smash it open, if there were any valuables in there that could be damaged, well a weapon would surely cause damage to them. A high enough STR check would result in the chest being smashed open and some of the treasure being damaged by the blow, a lower STR check would result in the chest bursting open and the treasure being unharmed.
I disagree with this, a better roll should result in a better result. A high enough STR check means that you were able to control your strength well enough to break the chest without damaging the contents, a worse roll means you used too much force and damaged the contents.
These are some of the ways I would personally inhibit repeat attempts or "oh my friend failed so I will have my character try also" situations, without trying to spoil any fun or "as DM I say no" things.
Smashing the Chest
Breaking open a chest, according to the Dungeon Master's Guide, is based on an attack, rather than a Strength check (forcing it open from pure strength rather than actually trying to damage it). In this case they make an attack, as per an Attack action, and each time they complete an Attack action you mark off 6 seconds. Missing the AC of the object isn't really a miss: they always hit (unless Nat 1 maybe for comedic effect), but the AC represents how tough the material is to actually damage (like Armour: anything above your 10+Dex means it hits you but strikes your armour and fails to damage it and therefore fails to damage you). Hit points is based on size and whether it is fragile for the material or sturdy. If we assume a medium sized chest made of sturdy wood: AC 15, 18 Hit Points. So, the character would attack, if they get 15 or more they damage it and when brough to 0 it is destroyed. As a home rule I would say if the attack brings it to negative max HP like a "character death" - in this case, -18, the contents are also damaged, I decide how much based on what the contents were (probably some secret rolling) fragile things like potions are probably 50% chance each unless they were important or something.
If the character keeps failing but wants to just "keep going until smashed" and time isn't an issue: I would ask for their attack bonus, I'd check if 10 + attack bonus is enough to hit the AC. If it is, they succeed but I roll 1d20, on a 20 I consider it as them having smashed contents as well (i.e. down to -18 in the example above). Whether they are successful or not I'd consider 1 minute to have passed but I would also have them roll a DC 10 Con Save: success means nothing, failure means a point of exhaustion for tiring themselves out on a chest. This is a homerule and I would ensure they know about this option before they make the attempt - it's an option I offer to them.
Repeated Ability Checks
This depends on what the check is for. Usually, most checks can't be repeated if they're things like Perception checks because it's a "you see it or you don't". Other times a check might be repeated if they could be tried again but they will have disadvantage and whether this is offered will be very situational, and 3rd attempts and more will be auto-fails. Alternatively, depending on the check or situation, I may allow infinite re-attempts: but the DC increases on each fail.
For situations that I feel appropriate, like with the chest above, I may offer the Take 10 option: take 10 + the modifier and this determines the success or fail from multiple attempts. It will take 1 minute, or longer depending on the nature of the situation -- generally, 10 x the normal time. If the task requires considerable physical exertion I might do a DC 10 Con Save to check for Exhaustion. If there is more than one of them trying the same task and working together (which means they normally have advantage) then I add +5 and use the modifier of whoever has the highest.
Take 10 option is only as I offer it, where I feel it is available. This is to prevent overusing this and basically just "take 10"-ing every situation, especially for those with high skill mods (**** you Expertise). This option will appear less if the party has a rogue nearing 11th level so they don't feel like their Reliable Talent is less special (since this 11th level feature is literally just an auto Take 10 without time-cost and probably the reason why Take 10 wasn't included as standard in this edition). The Take 10 option was specifically not included in 5th edition because skills and task DCs work differently and it can be an overpowered option in some circumstances.
In rare moments I might take a cue from the World of Darkness Storyteller system and offer the Succeed At A Cost : they can choose to automatically succeed but there will be a negative consequence determined by me. What that consequence is and whether I let them know what it will be beforehand depends on the situation - but they will know something bad may happen now or later as the cost of auto-success.
My group and I are rookies to D&D and are learning and researching as we go. This question has to do with members of a party making their own perception checks, one at a time over and over again. The same confusion I have applies to any ability check used over and over again by one player or by each player in the party.
In essence, my question is can players try an ability check an infinite amount of times without consequence until it successfully lands?
Example:
Last night when exploring through a mansion basement, I scouted ahead alone and failed a perception check, which meant I missed a door to another room. My buddy saw that our DM's makeshift map revealed a sliver of what looked like a door and said that he wanted to go back to where I was and try a perception attempt of his own. He's not high in perception so he failed too. He wanted to roll another perception check after failing one already because it was clear on the DM's physical map that there was a room beyond the hallway where we both failed a check to see a door, but since the check failed, he returned to the party. That seemed clear enough at the time so we moved on as if we didn't find a door to another room we clearly all knew was there.
However later in the game, we found ourselves in a room full of loot and a shabby, weak treasure chest. My same buddy (Paladin) didn't want to wait for our Rogue to pick the lock and wanted to instead just smash the chest open, which our DM said didn't seem to be protected by magic or particularly physically strong. So he made s strength check to smash it open. Bad roll and it failed. Then he wanted to take another smash, our rookie DM let him, bad roll and that failed too. He wanted to take a third smash, was allowed by our DM and he finally succeed in a strength check to smash the chest open. This seemed like cheating and possibly game breaking but in reality is something one would try to do and was understanable.
My question is, are there rules or etiquete to stop a player from repeatedly smashing a chest as my friend, after each failed throw, over and over again until he achieves what he intends to? Is this solely up to the DM? The same question could be applied to the failed perception check I explained further above and really any ability check of this nature.
What's to stop a player or a party from attempting an infinite amount of check throws to achieve something like lock picking a chest, searching and searching and searching until you finally succeed on any throw to achieve your objective? I feel this has far reaching consequences or advantages in the game if not understood properly.
Thanks for the insight and assistance!
The simple answer to your main question is: If there's a chance of success and no risk due to failure, don't roll, give it to them. If there is no chance of success, don't roll, explain that it is impossible.
---
You walked by the door, you failed your Perception, you moved on. As a DM I always look at the Passive Perception score first, if it's high enough then a roll isn't necessary, you've already noticed the door.
Your friend did a no no and used knowledge that he was never supposed to have to influence a decision to go back and check for that door again. As a DM I would have asked what prompted him to make this decision, if the answer was justified by an in-game reason I'd allow it. If the answer was based on the fact that I asked for a check from you or, in this case, he felt there was a door there for "reasons", I probably wouldn't allow the check. If I was feeling generous I'd let him walk back to the location and describe him looking for the door and finding nothing, without rolling.
This is a check that has a risk due to failure: You don't find the secret door which leads to something. It is a valid check.
---
In the case of smashing the chest, I wouldn't have had the player roll to see if they could smash the chest open, instead I'd have them roll to see if they damaged anything in the chest. The player could attack that chest all day long without any risk, so they could roll indefinitely until they got a successful roll, as seen by your example. However, they're using a weapon to smash it open, if there were any valuables in there that could be damaged, well a weapon would surely cause damage to them. A high enough STR check would result in the chest being smashed open and some of the treasure being damaged by the blow, a lower STR check would result in the chest bursting open and the treasure being unharmed.
In instances like this, where success is eventually inevitable, without risk, the DM should measure success in different ways. The amount of time it takes, or perhaps a mishap, causing you to succeed at a cost.
If you are looking for a door that is obviously there, you should just find it.
If it's hard to spot, but a thorough search will reveal it, maybe it takes a while.
If you see it on the DM's map, YOU ARE A FILTHY CHEATER AND DESERVE TO STEP ON A PRESSURE PLATE DROPPING A ROCK ON YOUR SCUMMY HEAD.
Moving on.
If you smash open a chest instead of picking it open, perhaps it contained potions among its valuable wealth which shatter due to your inelegant attempts to smash it open, thereby ruining some of the reward.
This kind of method allows your roll to measure success without causing the game to come to a stop until someone eventually rolls high enough.
Use Success at a Cost, and Failing Foreward to keep the story rolling merrily along.
Formatting Tooltips, because I always forget
I don't think there's a hard and fast rule, but the players should really try to avoid metagaming. If the character fails a Wisdom (Perception) check to spot something, they don't really know they failed, so there should usually be no reason to try again. On the other hand, if they try to break something open with a Strength (Athletics?) check, and fail to, they know they failed, so there's no reason why they shouldn't want to try again. Bear in mind, though, that each attempt takes time. If time's not really an issue, and it's a check they can succeed at, an option is to just not have the players roll, and just tell them they eventually succeeded. From a realism point of view... if they have all the time in world to attempt something they can conceivably do, they'll eventually get it, so there's no point in checking whether they do. If time's an issue, though... then go ahead and have them roll multiple times, and make sure they're aware how long they've been at it.
But regarding unwarranted repeated attempts (like checks to spot hidden doors, for example), I'd say talk to the players about why their characters are trying again, when they don't know they actually failed their attempt (for all they know, their check succeeded and they found nothing).
Also, regarding multiple characters checking for the same thing, depending on the situation, it might be better handled by considering it a Help action, instead of a separate check. If two players are looking in opposite directions to try to spot hidden enemies, you can treat that as two separate checks, but if they're both trying to force open a door, that's really one of them doing so with Help from the other.
Also, when having players make iterative rolls to attempt to find a needle in a hay stack, Curse of Strahd had a mechanic I liked.
Roll Percentile Dice, with maybe a 15% chance of success. Each time they fail, it takes an hour of search time, but increase the chance of success +10%, so on the second roll, 25%, third, 40%, etc.
This ensures eventual success, and an expected time for it to happen. This is only useful though if they are on a time limit.
Formatting Tooltips, because I always forget
This is all really helpful guys. I'm given a lot of important considerations here from you guys on how to perceive the base imagination aspects of this game, which is why it's so unbelievably cool and enduring. Thanks.
I like how it really tasks and empowers the DM to continue to guide a influence the action of the characters and gameplay. Our DM totally has this in his ability, he's cut out for it. But he's new and probably a little timid disrupting a players ideas and attempts so much in our first handful of sessions.
I especially like this consideration about Passive abilities. I'm finding our gameplay is going REALLY slow these first few sessions and I was having a conversation with our DM about how we could pick up the pace of our sessions considerably. I believe having the narrative of the DM just simply award action to players due to their high Passive qualities is a good way to cut extra and unnecessary roll time and walking about and help guide us a little bit more. Really interesting.
Except for a few cases the rules define for handling passive Perception and stealth, there's no right or wrong way to use ability checks. They're a tool for DMs to model situations in the story where there's a chance of failure. Different situations can be modeled in different ways. Some allow for retries, some don't.
Smashing open a chest is one of those cases where you can keep trying until it finally breaks or you give up. In those situations the DMG recommends to just skip ahead and assume they succeeded after 10 tries (assuming there was a chance of success in the first place.) That's exactly how I use Perception in my games when players are exploring: I let them examine small objects, containers, or 5'x5' chunks of wall or floor in 1 minute (10 rounds of taking the Search action.) If they're in a time-sensitive situation (like Combat), then I care about the rolls and exactly how many rounds it takes them to succeed.
On the other hand, if someone's trying to fast-talk an NPC and the NPC sees through their ruse, they're not going to be able to try the same lie again.
That said, there's one way of handling ability checks that I'm not particularly fond of, and that's allowing 1 check per character for things where you should be able to try until you succeed. This leads to absurd situations where the 18 Strength barbarian rolls once to break down the door, fails, and the DM declares the door unbreakable by the barbarian; then the 10 Strength wizard tries, rolls a 20, and breaks it down.
I always point out that a standard Perception Check is literally 6 seconds. Character spending more than 6 seconds on something is not unreasonable, but then that gets back to DMThac0's point about lack of consequences. I like the Curse of Strad rule Cecilao, but I think 1 hour is waaaaaayyyyy to long. 1-10 minutes is more appropriate.
It's also why I like have Passive being the floor for anything that isn't active contested.
I really liked the 3.5's "Take a 10 and Take a 20".
Our standard procedure if we weren't pressured for time was.
Make a Perception check to see what we get.
Take a 10 (that's 1 minute of time)
Take a 20 (that's 2 minutes of time)
As long as you were not worried about time, 1-2 minutes to search a room to do the through job isn't that big of a deal. Hell, 1-2 minutes per room isn't even THAT much time as long as orcs aren't banging down the door trying to kill you.
As for smashing the chest, I'm all for disabling chest in not the standard ways. ie: take off the hinges instead of the lock, don't stand IN FRONT of the chest when you open it. If you break the chest and smash the loot, you'll quickly earn yourself on the party's short list.
I disagree with this, a better roll should result in a better result. A high enough STR check means that you were able to control your strength well enough to break the chest without damaging the contents, a worse roll means you used too much force and damaged the contents.
Fair enough, either option is valid depending on how you're implementing STR.
It could be said that a higher STR roll means more power was behind the swing or the swing was better controlled as per your example.
These are some of the ways I would personally inhibit repeat attempts or "oh my friend failed so I will have my character try also" situations, without trying to spoil any fun or "as DM I say no" things.
Smashing the Chest
Breaking open a chest, according to the Dungeon Master's Guide, is based on an attack, rather than a Strength check (forcing it open from pure strength rather than actually trying to damage it). In this case they make an attack, as per an Attack action, and each time they complete an Attack action you mark off 6 seconds. Missing the AC of the object isn't really a miss: they always hit (unless Nat 1 maybe for comedic effect), but the AC represents how tough the material is to actually damage (like Armour: anything above your 10+Dex means it hits you but strikes your armour and fails to damage it and therefore fails to damage you). Hit points is based on size and whether it is fragile for the material or sturdy. If we assume a medium sized chest made of sturdy wood: AC 15, 18 Hit Points. So, the character would attack, if they get 15 or more they damage it and when brough to 0 it is destroyed. As a home rule I would say if the attack brings it to negative max HP like a "character death" - in this case, -18, the contents are also damaged, I decide how much based on what the contents were (probably some secret rolling) fragile things like potions are probably 50% chance each unless they were important or something.
If the character keeps failing but wants to just "keep going until smashed" and time isn't an issue: I would ask for their attack bonus, I'd check if 10 + attack bonus is enough to hit the AC. If it is, they succeed but I roll 1d20, on a 20 I consider it as them having smashed contents as well (i.e. down to -18 in the example above). Whether they are successful or not I'd consider 1 minute to have passed but I would also have them roll a DC 10 Con Save: success means nothing, failure means a point of exhaustion for tiring themselves out on a chest. This is a homerule and I would ensure they know about this option before they make the attempt - it's an option I offer to them.
Repeated Ability Checks
This depends on what the check is for. Usually, most checks can't be repeated if they're things like Perception checks because it's a "you see it or you don't". Other times a check might be repeated if they could be tried again but they will have disadvantage and whether this is offered will be very situational, and 3rd attempts and more will be auto-fails. Alternatively, depending on the check or situation, I may allow infinite re-attempts: but the DC increases on each fail.
For situations that I feel appropriate, like with the chest above, I may offer the Take 10 option: take 10 + the modifier and this determines the success or fail from multiple attempts. It will take 1 minute, or longer depending on the nature of the situation -- generally, 10 x the normal time. If the task requires considerable physical exertion I might do a DC 10 Con Save to check for Exhaustion. If there is more than one of them trying the same task and working together (which means they normally have advantage) then I add +5 and use the modifier of whoever has the highest.
Take 10 option is only as I offer it, where I feel it is available. This is to prevent overusing this and basically just "take 10"-ing every situation, especially for those with high skill mods (**** you Expertise). This option will appear less if the party has a rogue nearing 11th level so they don't feel like their Reliable Talent is less special (since this 11th level feature is literally just an auto Take 10 without time-cost and probably the reason why Take 10 wasn't included as standard in this edition). The Take 10 option was specifically not included in 5th edition because skills and task DCs work differently and it can be an overpowered option in some circumstances.
In rare moments I might take a cue from the World of Darkness Storyteller system and offer the Succeed At A Cost : they can choose to automatically succeed but there will be a negative consequence determined by me. What that consequence is and whether I let them know what it will be beforehand depends on the situation - but they will know something bad may happen now or later as the cost of auto-success.
My Homebrew: Races | Subclasses | Backgrounds | Spells | Magic Items | Feats
Need help with Homebrew? Check out this FAQ/Guide thread by IamSposta
See My Youtube Videos for Tips & Tricks using D&D Beyond