Rolling-until-sucess?

[Dark Schneider]

In DMG is said that a character could repeat the task only limited by time, with some exceptions like those that have a reaction or consecuence.

Then, can you rolling-until-sucess? Then opening locked doors is not so hard if no trap is involved.

After reading the PHB about passive skills, I initially though about 1st you roll d20, if roll is below 10 (passive), you can get some time for the task (repeat it) and use your passive value.
For new rolls, then some time should pass, or a new factor involved (like some new info concerning the task), to check again and try to roll greater than 10, or probably rolling until greater than passive, as you already did it with your passive value, but preseving the possibility of rolling a 1.

But after reading the DMG, I have this now. So confused. Seems that the core rules allow to roll as-many-as-want in a rolling-until-sucess style.

[DeTess]

In DMG is said that a character could repeat the task only limited by time, with some exceptions like those that have a reaction or consecuence.

Then, can you rolling-until-sucess? Then opening locked doors is not so hard if no trap is involved.

After reading the PHB about passive skills, I initially though about 1st you roll d20, if roll is below 10 (passive), you can get some time for the task (repeat it) and use your passive value.
For new rolls, then some time should pass, or a new factor involved (like some new info concerning the task), to check again (and try to roll greater than 10, or probably rolling until greater than passive, as you already did it with your passive value).

But after reading the DMG, I have this now. So confused. Seems that the core rules allow to roll as-many-as-want in a rolling-until-sucess style.

If a character has a lot of time, and opening the door is within his ability (remember, skills don't auto-succeed on a 20), he'll indeed eventually open the door. Of course, if he needs to open the door while sneaking into a place, every attempt will cost him more time and increase the risk of being caught. The same logic goes for most action that can be repeated, and if you don't want a PC to do that, introduce time limits.

[Pelle]

In those cases, you don't roll-until-success, you skip the rolling altogether and just declare automatic success.

[Unoriginal]

In DMG is said that a character could repeat the task only limited by time, with some exceptions like those that have a reaction or consecuence.

Then, can you rolling-until-sucess? Then opening locked doors is not so hard if no trap is involved.

After reading the PHB about passive skills, I initially though about 1st you roll d20, if roll is below 10 (passive), you can get some time for the task (repeat it) and use your passive value.
For new rolls, then some time should pass, or a new factor involved (like some new info concerning the task), to check again (and try to roll greater than 10).

But after reading the DMG, I have this now. So confused. Seems that the core rules allow to roll as-many-as-want in a rolling-until-sucess style.


If you attempt a task that is:

a) possible given your capacities
b) have unlimited time to do it
c) done in none stressful circumstances
d) has no interesting consequences for failure

you automatically succeed, after spending some time on it.

In many situations, though, time is of the essence (ex: you're attempting to unlock a prison door between the guards' patrols), so while you can retry until you succeed, but it takes time and it will have consequences (in the example above: having to hide regularly when the patrol shows up)

If you *are* in a stressful situation, like a fight, but that the task is possibe and there is no interesting consequences for failure, you can re-attempt it until you succeed. Ex: trying to unlock a door while your friends are battling a minotaur.

And of course, if a task is impossible for you (ex: a DC 20 STR check when you have 8 in STR), then you can't succeed no matter how much you try.

[Knaight]

In DMG is said that a character could repeat the task only limited by time, with some exceptions like those that have a reaction or consecuence.

Then, can you rolling-until-sucess? Then opening locked doors is not so hard if no trap is involved.

...

But after reading the DMG, I have this now. So confused. Seems that the core rules allow to roll as-many-as-want in a rolling-until-sucess style.
You can roll until success (though a 20 isn't a guaranteed success), which is a good reason not to call for rolls if failure doesn't actually do anything. That locked door? If there's no trap, no guard patrol, no time limit, and no reason not to leave evidence of tampering? Don't roll for it. A trap that goes off, a guard patrol three rounds away, an imminent delivery deadline for a stolen good behind the door, an attempt to plant evidence that works a lot better if you haven't obviously messed with the safe? Roll for it.

[Dark Schneider]

OK that is important for exploration.

Then check roll, and using 10 times you auto-roll 20. I'd like to introduce the use of passive roll for some half-time, as eventually your passive check would be enough to succeed, representing the characters improvement as they don't need to use the full 10 times to succeed on the same task, when passive + bonus is enough.

But not sure if apply the passive at 5 times or some sorter, as 10 is far from being a 20 (you only have 1 possibility to get a 20, but 11 possibilites to get something equal or greater than 10).
For long tasks probably at 2nd try, for short tasks at 2-5 times, not sure.

If rolling have consecuences (like opening a door in combat), then roll and roll, no auto-roll.

[Pelle]

If you *are* in a stressful situation, like a fight, but that the task is possibe and there is no interesting consequences for failure, you can re-attempt it until you succeed. Ex: trying to unlock a door while your friends are battling a minotaur.


Wait, didn't you claim in another thread that you could only re-attempt after the stressful situation ended? I didn't get to check what the book actually say (because it doesn't really matter to me), did we clear that up?

[Pelle]

OK that is important for exploration.

Then check roll, and using 10 times you auto-roll 20. I'd like to introduce the use of passive roll for some half-time, as eventually your passive check would be enough to succeed, representing the characters improvement as they don't need to use the full 10 times to succeed on the same task, when passive + bonus is enough.

But not sure if apply the passive at 5 times or some sorter, as 10 is far from being a 20 (you only have 1 possibility to get a 20, but 11 possibilites to get something equal or greater than 10).
For long tasks probably at 2nd try, for short tasks at 2-5 times, not sure.

Don't worry about it. Do you really track time that detailed?

Just say som significant time passes, like 10 min or 1 hour depending on the task. Or, just define the task according to time before you roll, say DC 20 to spend 1 min on the task, 15 for 10 min, and 10 for 1 hour or something. Or even better, DC 15 to open the door before the guard comes back, DC 10 to do it before the captives are sacrificed.

[Louro]

I like too see this from another perspective.

A failed rolls means the task is beyond your capabilities, so there will is no reroll. You simply don't know how to open that particular lock.

[Dark Schneider]

Time is important as there are effects that last 10 min, 1 hour, etc. Spells, antitoxin.

If you summon an elemental for the dungeon, then have to know the time passed as there is difference between having or not that elemental.

[Unoriginal]

Wait, didn't you claim in another thread that you could only re-attempt after the stressful situation ended? I didn't get to check what the book actually say (because it doesn't really matter to me), did we clear that up?

That was only if the circumstances made attempting another check impossible/unwarranted unless there is a change in situation.

[Pelle]

Just say som significant time passes, like 10 min or 1 hour depending on the task.


Time is important as there are effects that last 10 min, 1 hour, etc. Spells, antitoxin.

If you summon an elemental for the dungeon, then have to know the time passed as there is difference between having or not that elemental.

Exactly, so limit it to 10 min, 1 hour, etc, don't bother with 7 or 32 min. Or just say DC 15 to finish the task before your elemental leaves.

[Mordaedil]

You've never tried to pick locks in real life, have you Louro?

You can roll until success, but there's already a mechanic for auto-success called taking 10 and taking 20, where the latter takes a longer time and the former takes a bit and cannot be done while in combat. It's intentional that the rogue can keep attempting to open the lock during combat, it's to create tense situations where these exist (and they are painfully few in the games I experience) and it is fine if you assume players take 10 outside of combat.

If you decide not to allow the players to do this, you'll get ludicrous things like players destroying the door, or sometimes the wall, since it is often weaker than the door the DM thought would be impenetrable.

[Unoriginal]

OK that is important for exploration.

Then check roll, and using 10 times you auto-roll 20. I'd like to introduce the use of passive roll for some half-time, as eventually your passive check would be enough to succeed, representing the characters improvement as they don't need to use the full 10 times to succeed on the same task, when passive + bonus is enough.

But not sure if apply the passive at 5 times or some sorter, as 10 is far from being a 20 (you only have 1 possibility to get a 20, but 11 possibilites to get something equal or greater than 10).
For long tasks probably at 2nd try, for short tasks at 2-5 times, not sure.

You're better off just hand-waving it unless time is actually a constraint, Dark Schneider. It doesn't matter if it's one minute or 20.


I like too see this from another perspective.

A failed rolls means the task is beyond your capabilities, so there will is no reroll. You simply don't know how to open that particular lock.

While it's a legitimate game design, they didn't go for this for 5e. There are much more auto-success than one used to previous editions would expect, for example.


Time is important as there are effects that last 10 min, 1 hour, etc. Spells, antitoxin.

If you summon an elemental for the dungeon, then have to know the time passed as there is difference between having or not that elemental.

In that case failure DOES have interesting consequences, and you can't just auto-succeed.



but there's already a mechanic for auto-success called taking 10 and taking 20

No there not. You're mixing things up with the 3.X rules.



If you decide not to allow the players to do this, you'll get ludicrous things like players destroying the door, or sometimes the wall, since it is often weaker than the door the DM thought would be impenetrable.

In the published modules at least, the lock is generally much easier to pick than the door to break.

[Pelle]

That was only if the circumstances made attempting another check impossible/unwarranted unless there is a change in situation.

I don't see the difference (here, specifically between trying to climb and trying to open a lock during combat).

[Unoriginal]

I don't see the difference (here, specifically between trying to climb and trying to open a lock during combat).

If you recall our previous discussion, I wasn't claiming you couldn't re-try climbing, just that a DM could say you couldn't. I think I said I would let them try to climb once more for what was a short combat, IIRC.

For example it could be said that if the thing your climb would take too long for a combat, then you can't attempt more than one STR (Athletic) check. Because the check isn't just to pass one particular rough patch, its for the whole climb.

Same way you can have to do a STR (Athletic) check for diving to the bottom of a lake: if you're doing this in the middle of a combat, the DM could say it takes too long to retry while in-combat. If you get what I mean.

The typical "can't retry the check unless the situation changes" is more things like Wis (Survival) (where you can re-try after getting lost, for ex), Int (Religion) or Cha (Persuasion).

[Dark Schneider]

The "taking 20" rule for 5E is on DMG at ability checks section. When no time or consecuences, it takes 10 times.

I think I'd also introduce the passive value for 2nd attempt. Usually is not very important, but in some cases would be applied in an interesting way, I put the main one:
- You have to move slow (half speed) to look for traps. This allows to check roll perception/investigation (depending the trap). In this case you can roll or use your passive (if don't want to roll).
- If you decide to search meticulously, you move 1/4 speed (half than searching), but if you roll lower than your passive, you use your passive. Halving the speed is equivalent to move at normal rate (that is half when searching), roll, stop, and try a 2nd attempt, so moving at 1/4 I think is better for mechanics.

- Notice that moving at 1/20 (1/10 of half speed) is like rolling a 20. It is your max ability but implies you move very very slow. You move a step, and check everything, another step, and check everything...

[Knaight]

I like too see this from another perspective.

A failed rolls means the task is beyond your capabilities, so there will is no reroll. You simply don't know how to open that particular lock.

That's not so much another perspective as another rule - one that D&D chose not to use, for whatever reason. Switching to that is a house rule, and while there's a case to be made for it being a better rule* it's definitely a rule change.

*I'd mostly make that case, though in the specific contexts of a hard time limit or notable variable failure case that doesn't necessarily hold, and there's a definite stylistic argument to be made for emphasizing these.

[jdolch]

Maybe take a page from Shadowrun's extended tests:

If the player is in a Situation where he can try multiple times then let him try multiple times. Set a timeframe for each try (and if there is some element of danger, maybe roll some dice behind the screen to see if he gets discovered or something). Then let him roll and everytime he fails the DC gets a +1 modifier.

So on the first try the DC15 lock is DC15.
On the second try he needs to beat DC16.
On the third try he needs to beat DC17.
etc.

The Test ends when he either succeeds or it becomes mathematically impossible to succeed.

[PhoenixPhyre]

That's not so much another perspective as another rule - one that D&D chose not to use, for whatever reason. Switching to that is a house rule, and while there's a case to be made for it being a better rule* it's definitely a rule change.

*I'd mostly make that case, though in the specific contexts of a hard time limit or notable variable failure case that doesn't necessarily hold, and there's a definite stylistic argument to be made for emphasizing these.

Not really. Players don't call for checks, and DMs are fully authorized to allow or disallow checks at any time for any reason. So it's not a house rule, it's just a DM using his rule-granted discretion.

There are a few cases of interest with attempted tasks:

Success Impossible (whether due to an inappropriate approach or the task simply being too hard): no check needed, just fail.

Success possible, but no interesting consequences for failure and no time pressure: no check needed, just succeed.

Success possible but no interesting consequences for failure except time pressure: One check; on failure allow auto-success at 10x time cost.

Success possible with interesting consequences for failure: One check. Failure means the situation changes so that another check with the same approach is impossible.

---------the ones above are individual checks.

Group endeavor where failure can happen and people can help each other: Group check (50% must pass).

Endeavor where one person can help another: Help action (advantage on the check).

Endeavor that's repeated in slightly difference circumstances or where the existence of a check should be secret from the players: Passive check.

[Pelle]

[...]

Cool, then I better understand what you meant last time, and I think we are on the same page.

The circumstances or situation changing only matters if it affects the task in question. If the Rogue is trying to climb/pick a lock, and then a goblin shows up, starting a combat with the Fighter, that may or may not affect the task the Rogue is doing. The situation has changed, but the Rogue is still doing the same thing. However, what the task represent may change (picking a lock in 10 min vs. 6 sec), and the DM may also rule that the DC should be higher now because the character likely feels more stressed.

[Vogie]

An alternative that I've seen DMs use is while "Taking 20" to do the task, they also have to roll a d20... It's considered to be a success, unless they roll a 1, which means they ultimately fail, regardless of time taken.

[FieserMoep]

For some tasks we allow to "keep" rolling. Regarding the lock picking example its a way of tracking time and other stuff.
If you need 10 attempts a minute may have passed. Sometimes that is relevant, sometimes it is not. It can keep up the tension though if you still insist on rolling for the player does not know if this is a situation in which he can get caught or not. Another thing is that a minute of fiddling with a lock can also cause noise etc. each attempt might require the lock picker to test on stealth versus the perception of a guard etc. each additional roll is a chance of failure or discovery.

Sure, a character might "fish" for the 5% chance to get the task done - if he thinks he even can for he has no idea of the DC - but another character might also lose patience and bash the door open. "Giving up" on a task is imho also important as a choice you leave to a character. They may have to face that they can't do everything even though they tried hard or if it was possible, may have to stand up regarding questions why they did not try harder and gave up so early.

[ChildofLuthic]

So I only do a roll when

1) the task is possibly beyond their capabilities
2) they are trying to do the task as quickly as possible.

In case 1, a failure would mean they cannot pick that lock. No retry, just a super complicated lock that they can't pick. In case 2, they might be trying to break into a room that has a monster in it. On a success, they quickly open the lock and the monster is surprised (or otherwise unprepared.) On a failure, I might say that the lock makes an audible click, and they here a monster perk up. They can continue unlocking it, but they def messed up.

Honestly, if it's the kind of task that they'll eventually succeed at, and there's no consequence for taking their time, why would you make them roll.

[Unoriginal]

So I only do a roll when

1) the task is possibly beyond their capabilities
2) they are trying to do the task as quickly as possible.

In case 1, a failure would mean they cannot pick that lock. No retry, just a super complicated lock that they can't pick. In case 2, they might be trying to break into a room that has a monster in it. On a success, they quickly open the lock and the monster is surprised (or otherwise unprepared.) On a failure, I might say that the lock makes an audible click, and they here a monster perk up. They can continue unlocking it, but they def messed up.

Honestly, if it's the kind of task that they'll eventually succeed at, and there's no consequence for taking their time, why would you make them roll.

If the task is beyond their capabilities, why are you allowing them to try?

[terodil]

If the task is beyond their capabilities, why are you allowing them to try?
I think the key word there was 'possibly'. I.e., the DC can [only] be beaten with good rolls, so neither success nor failure are guaranteed.

[Louro]

The "taking 20" rule for 5E is on DMG at ability checks section. When no time or consecuences, it takes 10 times.
There is no taking 20.
DMG just says you can automatically succeed on some tasks if you have time enough, and you can't ever succeed on others. It's up to the DM.

[MadBear]

If the task is beyond their capabilities, why are you allowing them to try?

My interpretation of what they're saying is essentially "The roll will determine if this is a lock the character could pick". In this way the DM and player won't know if the player could open the lock until the roll happens.

I've seen this work fine, but it does get a bit problematic when you start having other players attempting to accomplish the task.

Rogue: I rolled a 4
DM: This lock is so complicated it's beyond your understanding (DC 15)
Fighter: I try, I rolled a 16
DM: You quickly unlock it.

In this case you can always make an exception (Maybe the fighter saw this type of lock in their background, or when they learned the skill, they saw this lock in use, but the rogue never did). But the more instances of that, that happen, the less believable it gets.

[PhoenixPhyre]

There is no taking 20.
DMG just says you can automatically succeed on some tasks if you have time enough, and you can't ever succeed on others. It's up to the DM.

There are two "auto-success" and automatic failure rules.

First, the DM can decide that no check is necessary (either auto-success or auto-failure). This can be for any reason, including that there are no meaningful chance of failure and that failure is not interesting/does not change the situation at all. Conversely, your attempt could be an automatic failure based on the fiction of the scenario. You can't swing off a chandelier to attack if there are no chandeliers for miles. This isn't even "taking 20" in the sense of comparing your max roll to a fixed DC. This is not even setting a DC at all.

Second, once a DC is set and you've failed once, the DM can allow you to automatically succeed (as if you rolled a 20) but at the cost of 10x the normal time. Although, I can't seem to find the direct citation in the DMG at all. I remember reading it, but...

[Louro]

Second, once a DC is set and you've failed once, the DM can allow you to automatically succeed (as if you rolled a 20) but at the cost of 10x the normal time. Although, I can't seem to find the direct citation in the DMG at all. I remember reading it, but...
Do not agree/like that.
That means a lv1 PC can break a DC 25 (very hard) lock in just 1 minute.
EDIT: 30 seconds actually

MULTIPLE ABILITY CHECKS
Sometimes a character fails an ability check and wants to try again. In some cases , a character is free to do so; the only real cost is the time it takes. With enough attempts and enough time, a character should eventually succeed at the task. To speed things up, assume that a character spending ten times the normal amount of time needed to complete a task automatically succeeds at that task. However, no amount of repeating the check allows a character to turn an impossible task into a successful one.
-DMG p237

[Unoriginal]

Second, once a DC is set and you've failed once, the DM can allow you to automatically succeed (as if you rolled a 20) but at the cost of 10x the normal time. Although, I can't seem to find the direct citation in the DMG at all. I remember reading it, but...

Rolling a 20 isn't automatic success.

[PhoenixPhyre]

Rolling a 20 isn't automatic success.

I know. That was a poorly-worded "succeed at anything you'd be able to succeed at if you had rolled a 20 on a normal check" qualifier (so ruling out a +0 making a DC 30 check scenario). My bad for wording that confusingly.

[JNAProductions]

Do not agree/like that.
That means a lv1 PC can break a DC 25 (very hard) lock in just 1 minute.
EDIT: 30 seconds actually

MULTIPLE ABILITY CHECKS
Sometimes a character fails an ability check and wants to try again. In some cases , a character is free to do so; the only real cost is the time it takes. With enough attempts and enough time, a character should eventually succeed at the task. To speed things up, assume that a character spending ten times the normal amount of time needed to complete a task automatically succeeds at that task. However, no amount of repeating the check allows a character to turn an impossible task into a successful one.
-DMG p237


One minute. It takes an action, at least, to pick a lock, normally, and you only get one a round.

And if there’s absolutely no pressure, time or otherwise, then someone with great natural talent (16 or 17 Dex) AND training in lock picking, or good natural talent (12+ Dex) and expertise in lockpicking, then yes. You can.

[Louro]

One minute. It takes an action, at least, to pick a lock, normally, and you only get one a round.

And if there’s absolutely no pressure, time or otherwise, then someone with great natural talent (16 or 17 Dex) AND training in lock picking, or good natural talent (12+ Dex) and expertise in lockpicking, then yes. You can.

One action and one move each round, 3 secs each?
But then, lockpicking any "very hard" lock becomes just a child game.

- I take 1 minute to check for traps, total perception 25.
- And I take another minute to open the lock, total 25.
- I take 1 min to hide, stealth total 25.
- I take 1 min to track the monster, survival total 25.

I think this is wrong. Can you please tell me where is that "taking 20" thing? Because I didn't find it.

[JNAProductions]

One action and one move each round, 3 secs each?
But then, lockpicking any "very hard" lock becomes just a child game.

- I take 1 minute to check for traps, total perception 25.
- And I take another minute to open the lock, total 25.
- I take 1 min to hide, stealth total 25.
- I take 1 min to track the monster, survival total 25.

I think this is wrong. Can you please tell me where is that "taking 20" thing? Because I didn't find it.

You can't pick a lock with your movement. Movement is not an action.

Checking for traps:
Taking 20 here would mean you thoroughly explore every nook and cranny, which would certainly trigger most traps before you find them.

Opening the lock:
If it's possible, and you're NOT under pressure, then yeah, someone could open up to a DC 27 lock at level 1. (Expertise in Thieve's Tools and Dexterity mod of +3.)

Stealth:
You straight up can't take 20 here if you're moving. I would allow you to take 20 on a SINGLE LOCATION, ensuring you have the best hiding spot, but if you move and try to take 20, you just roll.

Survival:
Tracking is done over the course of several hours. No taking 20 here, since by the time you'd get to the 20, the tracks are LONG GONE.

This is all how I would rule it.

[Lunali]

One action and one move each round, 3 secs each?
But then, lockpicking any "very hard" lock becomes just a child game.

- I take 1 minute to check for traps, total perception 25.
- And I take another minute to open the lock, total 25.
- I take 1 min to hide, stealth total 25.
- I take 1 min to track the monster, survival total 25.

I think this is wrong. Can you please tell me where is that "taking 20" thing? Because I didn't find it.

Taking 20 is from 3.x, it only applies when there is no penalty for failure, though the original rules were 20x the time so 2 min. The first two are perfectly reasonable, though that means taking 2 min per 5ft of dungeon for checking for traps. For stealth, the character shouldn't know how good their stealth roll is and the penalty for failure is getting spotted. For tracking, you will be moving at 1/20th normal tracking speed, meaning you will most likely either lose ground or walk into an ambush.

[ad_hoc]

A check is a single thing. No checks should ever be 're-rolled' unless there is some sort of ability which does just that.

When an ability check is called for it should be an exciting thing. The action is interesting, the outcome is uncertain, and there are meaningful consequences for failure. If one of those is missing then no roll should be made.

Or to put it another way, don't roll when it isn't fun to do so. Just get on with it.

[Erloas]

Not that he is any more of an expert than anyone else, but he's said it well enough to not trying to recreate it:
https://theangrygm.com/five-simple-rules-for-dating-my-teenaged-skill-system/ Specifically rule 2, but it is all related.

[Dark Schneider]

I see controversy about this. Probably those (including me) who see rare to repeat until success come from another games that penalize the failure.

But looking at DMG, I think D&D is a system that favors the advance, not getting stuck by a bad check roll (like a door as unique way that you failed to open).

Probably one of the main problems is that the skill progression is low, I mean a character usually starts with +5 on their main skills, and ends with about +10 at very high level (+11 at maximum), only 5 of difference.
With the proposed DC progression, that is only 1 difficulty to auto-success (or average success at short time using passive AKA "taking 10" on 5E). Seems appropriate then to use DC like 16, 17, 18...so it determines average success or failure (when no rolling) with intermediate level characters, making difference with low level ones.

Maybe that is the intention, a character have a profession so it knows how to do the things he is intended to know, when at high level he auto-success a higher DC, and intermediate levels simply gives more probability of success under pressure, and will get less negative consecuences due to failures when applicable.

So, as example, you are a rogue, so you are supposed to know how to lock picks, even at lvl 1, then you are capable of locking picks with DC25 (very hard), at high level you will be able to lock picks with DC30. Also, while you improve your skills, you will get less failures, will be more reliable when required, and will do it faster (more success with the unique 1st roll before using the auto-20).

Not sure if this was the intention, but sounds like that for me.

Also remember that for contests is not applicable, so cannot be used for stealth and others like that.

I also suppose that another controversy is due the proficiency bonus is not high, a skilled character could auto-succeed on DC25, but a non-skilled (simply stat not negative) could auto-succeed on any DC20, that is not bad at all. Probably this is this way thinking on a group cannot have all classes, so the other characters could try, but I understand the controversy involving it.

[Louro]

Taking 20 is from 3.x
So, there is no taking 20 in 5e.
Case closed.

[Mordaedil]

Turns out it's actually in the DMG, which I don't have, but I found on another post on these forums.



Take 20 is an explicit rule in 5e, though it's not referred to by name.



Multiple Ability Checks
Sometimes a character fails an ability check and wants to try again. In some cases, a character is fre to do so; the only real cost is the time it takes. With enough attempts and enough time, a character should eventually succeed at the task. To speed things up, assume that the character spending ten times the normal amount of time needed to complete a task automatically succeeds at that task. However, no amount of repeating the check allows a character to turn an impossible task into a successful one.


So uh, yeah. That's the same as Take 20 from 3.X.

[Louro]

Turns out it's actually in the DMG, which I don't have, but I found on another post on these forums.



So uh, yeah. That's the same as Take 20 from 3.X.
I'm sorry, but the DM deciding that players can auto succeed on a task if they commit some time to it is not the same as taking 20.
DM deciding VS players deciding.

- I take 30 secs to pick the lock.
- I help him.
There goes the DC 30 impossible lock.
(Action is half of the round, movement the other 3 secs.)

[Pelle]

So uh, yeah. That's the same as Take 20 from 3.X.

There's some nuance, though. It is not supposed to be declared by the player, but adjudication by the DM instead. And a DM might allow automatic success on a DC 25 task to character with only +0, depending on how you read/rule it.



Do not agree/like that.
That means a lv1 PC can break a DC 25 (very hard) lock in just 1 minute.


That assumes that the DC 25 is for picking it with 1 action, in 6 seconds. The DM could just as well have ruled it that a single check represents 10 min of trying instead, and that picking it with 1 action is DC 30 or automatic failure.

[Randomthom]

I'd allow a re-roll where the following are true;
Time is not a factor
Failure has no consequences
The character knows they failed and has a reason to believe that trying again could result in a success

That last one has particular implications for things like perception and intelligence (nature/arcane/history etc.)

[Unoriginal]

I'm sorry, but the DM deciding that players can auto succeed on a task if they commit some time to it is not the same as taking 20.
DM deciding VS players deciding.

- I take 30 secs to pick the lock.
- I help him.
There goes the DC 30 impossible lock.
(Action is half of the round, movement the other 3 secs.)

Just to say, DC 30 isn't impossible, just "nearly" impossible.

An impossible-to-open-lock cannot be opened, no check needed.

[Louro]

Just to say, DC 30 isn't impossible, just "nearly" impossible.

An impossible-to-open-lock cannot be opened, no check needed.

But you get the point? Don't you?
"Nearly impossible" task 100% success if you have 1 minute and a friend.
Does that falls within the "nearly impossible" meaning for you?

EDIT: It seems to me that some people have forgotten the first rule when you transition from 3.x to 5e (forget everything about 3.x).

[PhoenixPhyre]

But you get the point? Don't you?
"Nearly impossible" task 100% success if you have 1 minute and a friend.
Does that falls within the "nearly impossible" meaning for you?

EDIT: It seems to me that some people have forgotten the first rule when you transition from 3.x to 5e (forget everything about 3.x).

To even potentially pick a DC 30 lock even with the auto-success rule, you need a +10 modifier. That's being level 5+ with expertise and +4 DEX (meaning you've chosen to be really really good at picking locks, way above the norm) or level 17+ with proficiency and a +4 DEX (meaning you're one of the best darn people in the world already).

I see no problem with either highly dedicated/specialized lock-picker or a very high-level PC being able to quickly open any mundane lock.

Edit: and the labels (easy, medium, etc) are referenced off of beginning, unspecialized adventurers, specifically those with a +4 total mod. That gives the following probabilities of failure:

DC Failure
5 0%
10 25%
15 50%
20 75%
25 100%
30 100%

[Unoriginal]

People seem to forget that the task need to be possible for the one who attempt it for it to have a chance at auto-success.

If breaking down a DC 20 door is possible-but-unlikely for someone with +1 in STR, they may eventually succeed. Someone with 8 in STR won't.

[Louro]

Mmmm...
Lv 1 Rogue with 16 DEX with the 3.5 "taking 20" rule:
20 + 3 (DEX) + 2 (prof) + 5 (help/advantage) = 30
Same for the DC 30 almost impossible to break door. Character with STR 16 and his buddy expend one minute and the door is gone.

Am I the only one seeing a problem here?
When I set a DC on something I already know players can fail. If I want them to succeed them there would be no roll required, which is what DMG says.
Taking 20 makes no sense in 5e.

[Dark Schneider]

The +5 for help is only applied to passive checks, so it allows to use 15 instead 10. For auto-20 I have not seen that. See that in this case the max roll is 20 in any case. So we could tell that the max is limited by the most skilled one.

But sounds reasonable if want to apply also to auto-20. But in DMG is not mentioned. But I think in the case of auto-20 is better to short the time used, because rolling with advantage you roll 2d20 instead 1d20, so you would get the 20 earlier.

Could be good using 20 times for auto-20, or 10 times with help? Using the core rule, would be 10 times for auto-20 and 5 times with help (because advantage).

Rules also mention about failing by more than 5 (or was by 5 or more?, not sure) to apply some penalty. In this case as the 1st check is always a roll, then apply something if failed by much. Maybe increasing DC by 1, breaking the tools...

[Bacon Elemental]

Wait, why do you get +5 for having help? I thought Advantage was always just a flat "reroll, pick highest"?

And since you are considered to have rolled a 20 already I dont see how it would help...

[Louro]

The +5 for help is only applied to passive roll, so it allows to use 15 instead 10. For auto-20 I have not seen that. See that in this case the max roll is 20 in any case. So we could tell that the max is limited by the most skilled one.

But sounds reasonable if want to apply also to auto-20. But in DMG is not mentioned.
DMG doesn't even mention auto-20.
Do you really find reasonable a 100% succeed rate on "nealy imposible" tasks?
Can't argue more. Not in the books and source of nonsense.

[Dark Schneider]

DMG doesn't even mention auto-20.
Do you really find reasonable a 100% succeed rate on "nealy imposible" tasks?
Can't argue more. Not in the books and source of nonsense.
Is not mentioned directly, but with other words. It is on DMG at Using Ability Checks (or something like that, I have the translated book, sorry). It says something like (sorry translated):

"If have enough time, he can try until success. For make faster the process, if have 10 times the time required to achieve the task, it succeed automatically. Anyway it does not allow a character to succeed in an impossible check."

It is supposed this is to avoid players throwing the dices so many times that could be annoying, as they eventually will get a 20. So instead if auto or not auto a 20, I think is more related to apply some penalty at failure, allowing a new check roll after some time instead each round, and so on.

I'l try to put some examples based on other games with the "partial success" conception. In our case, let's take failing by 5 or less as partial success, and it means character can try again the next round, if failed by more than 5, is a failure:
- Lock pick: if failure, you have to wait 1 hour (alternative, a short rest is required) to try again.
- Detect hidden door: on failure you have to wait 1 hour (or a short rest) to try again.
- Bash door: on failure lose some HP.

I would not apply the time penalties if time is critical, like picking lock to exit a room while a trap is activated.

[Pelle]

People seem to forget that the task need to be possible for the one who attempt it for it to have a chance at auto-success.

If breaking down a DC 20 door is possible-but-unlikely for someone with +1 in STR, they may eventually succeed. Someone with 8 in STR won't.

I'm not sure you need to limit if something is possible for the one attempting it purely by the DC.

I don't usually define the task as "breaking down the door - DC 20". I'd rather do "break down the door in 1 min is DC 20". If given enough time, even a Str 8 can do it, I then just define "breaking down the door in 10 min is DC 15", thus 'taking 20' is possible.

If there's a threshold strength that needs to be there in order to break down the door, and work over time don't matter, then a simple Strength criteria is a better way of adjudicating it. "To break this door/lift this rock/bend this steel bar, you need Str 18". If you are rolling a check for these kind of situations, then you are essentially doing a state-of-the-world check. Your result determines how heavy the rock is, not if the character is able to lift the 50 kg rock.

[Mordaedil]

There's some nuance, though. It is not supposed to be declared by the player, but adjudication by the DM instead. And a DM might allow automatic success on a DC 25 task to character with only +0, depending on how you read/rule it.

Right. That is not different from a take 20. A player in 3.x still has to ask and the DM still has to rule whether they can or not. Heck, that goes for nearly everything in older editions of D&D. But that is a conversation or semantics at this point.

[Pelle]

Right. That is not different from a take 20. A player in 3.x still has to ask and the DM still has to rule whether they can or not. Heck, that goes for nearly everything in older editions of D&D. But that is a conversation or semantics at this point.

Hehe, all I can say is this nuance is what some of the most vocal 5e detractors rail about.

[PhoenixPhyre]

Right. That is not different from a take 20. A player in 3.x still has to ask and the DM still has to rule whether they can or not. Heck, that goes for nearly everything in older editions of D&D. But that is a conversation or semantics at this point.

But take 20 has hard rules around when it's possible and feats that modify when it's possible. Combined with the basic philosophy of 3e that the text is paramount, you get very little room for DM discretion once the facts are established. "Does my character take 10/20" is a player decision made before the roll happens.

Compare to 5e, where the very possibility of a check is left in the hands of the DM, and the "auto-success" rule, like everything in the DMG, is explicitly an option that DMs can employ to speed things up if they wish. Or ignore if they choose. It also happens once they've failed once on the check, not preemptively. It's not a player entitlement, not something they can plan on. What defines "no meaningful consequences for failure?" When is the only concern time? Those are entirely up to the DM's vision of what the situation requires.

[Icecaster]

My general rule is that if a task is reasonably within a character's capability, I let them complete the task within a couple of minutes for a task like opening a door without pressure. I would skip all of the rolling until you succeed in favor of time and fun. Otherwise I'll implement costs or punishments for failing a task, such as breaking lockpicks or other tools.

[Dark Schneider]

One action and one move each round, 3 secs each?
But then, lockpicking any "very hard" lock becomes just a child game.

- I take 1 minute to check for traps, total perception 25.
- And I take another minute to open the lock, total 25.
- I take 1 min to hide, stealth total 25.
- I take 1 min to track the monster, survival total 25.

I think this is wrong. Can you please tell me where is that "taking 20" thing? Because I didn't find it.
Haha didn't noticed that post. The total skill compendium.

But in some cases like contest or how Survival is used (DM requires a Survival roll every X time, so it cannot be reattempted like that), it can't be applied.

As some people mentioned, one thing is

reasonably within a character's capability
that is the passive skill, the average, in other words, how good the character is. And another thing is that over that, and even more, a 20, is a brilliant moment for the character, really a good execution moment. So I think is not bad to distinguish them.
For DC below passive skill, you should determine rolling below is something that can happen, but with time that character should be able to accomplish the task without problems, so don't let a character to fail a task for DC less or equal its passive skill. If failed on check roll, simply it will take some longer for auto-success.
For greater DC than passive, you could require to success on check roll, and if failed apply some penalty.

[PhoenixPhyre]

Haha didn't noticed that post. The total skill compendium.

But in some cases like contest or how Survival is used (DM requires a Survival roll every X time, so it cannot be reattempted like that), it can't be applied.

Or any case where there are meaningful consequences for failure. Which is 90% of the actual checks made, because anything where there isn't a meaningful consequence for failure is a no-check-needed success or failure because it's not interesting.

Jump to the moon? Fail.
Climb that tree (with no one around, in narrative time)? Success.

5e's rules are clear. Before a check is called for, the DM needs to ask if the results of the roll matter. If not, anything reasonable just plain happens. Anything unreasonable just plain doesn't happen.

Locks and traps, by themselves, are boring and aren't worth rolling for (as long as someone has the requisite tools and proficiency). Locks and traps when you're under pressure, when every minute counts or when there's an alarm or another interesting failure condition, or when they're not so easily bypassed (like a puzzle lock or an obvious trap with the working parts out of reach), then they're potentially interesting.

[Louro]

Haha didn't noticed that post. The total skill compendium.

But in some cases like contest or how Survival is used (DM requires a Survival roll every X time, so it cannot be reattempted like that), it can't be applied.
Yeah, sometimes you can't. But what if just wanna get info on a skirmish scenario? One check would allow you to guess how many, what kind of creatures, what direction they came and went... Same for investigation checks. Same for spotting at junctions or doors... Smart players would take a minute to check in deep a notable door.

Here we tend to use the "you fail, you failed" approach. There is no reroll unless something has changed (no short rest reroll either). A fail means that particular task is beyond your capabilities/experience.

[Dark Schneider]

Yeah, sometimes you can't. But what if just wanna get info on a skirmish scenario? One check would allow you to guess how many, what kind of creatures, what direction they came and went... Same for investigation checks. Same for spotting at junctions or doors... Smart players would take a minute to check in deep a notable door.

Here we tend to use the "you fail, you failed" approach. There is no reroll unless something has changed (no short rest reroll either). A fail means that particular task is beyond your capabilities/experience.
Well but think that the roll is 1-20 and skills usually 0-16, so much relies on luck. Then some kind of reroll based on skill (like using passive or some difference of failure) should be applied to allow reroll, if not we rely too much IMHO onto luck, and skill itself should be important.

[PhoenixPhyre]

Well but think that the roll is 1-20 and skills usually 0-16, so much relies on luck. Then some kind of reroll based on skill (like using passive or some difference of failure) should be applied to allow reroll, if not we rely too much IMHO onto luck, and skill itself should be important.

The whole point of rolling in 5e is to include luck (chance, more precisely). When things are certain, don't bother rolling. Even when things are nearly certain, it's often better not to roll. I'd say most DMs roll too many checks for things that just plain don't matter. Especially knowledge-type checks. IME, it's better to give them information so they can make meaningful choices. This means that simply having information can't be the key--choices have to have weight on both sides instead of having an obvious answer.

Another thing is that rolls aren't player-initiated. While this seems obvious, it means that the player must describe what the character does that triggers that resolution mechanic. You can't simply say "I diplomacy it!" and roll some dice. You have to (at minimum) describe your approach and intent in enough detail for the DM to decide the best resolution. How much detail is required will vary from situation to situation and DM to DM, as it should.

[Louro]

That's true.
What I tend to use on dramatic scenes is this sort of collective challenge, accounting for successes.
Let's say they find an attacker corpse and wanna find what happened.
I'll have players roll for medicine, survival and nature. With 2 successes they will get a good hint about what monster did it.

Inside some crumbling ruins.
Dex saves, int check to remember shortest path, str checks for stamina, int check to avoid weak structural paths...

[Slipperychicken]

Then, can you rolling-until-sucess? Then opening locked doors is not so hard if no trap is involved.


Think about this. The character is attempting a routine task for which he has all the necessary skills, resources, and time. That means there isn't any reason to involve random variables. Skip the roll and say that he completes the task after an appropriate amount of time.

A random variable like a dice roll is only needed when outcomes are not certain. When you can look at a situation from a lore perspective and are certain that there is only one plausible outcome, then you can skip the random variable, declare the result, and move on. It will save you many headaches and a lot of time.

[PhoenixPhyre]

That's true.
What I tend to use on dramatic scenes is this sort of collective challenge, accounting for successes.
Let's say they find an attacker corpse and wanna find what happened.
I'll have players roll for medicine, survival and nature. With 2 successes they will get a good hint about what monster did it.

Inside some crumbling ruins.
Dex saves, int check to remember shortest path, str checks for stamina, int check to avoid weak structural paths...

I don't even require that much.

For knowledge-recall situations that aren't under extreme time pressure (AKA combat or other such things), I either background/proficiency-gate the information (so someone with the appropriate background, including where their character is from and what they've said they focused on, will be the one who receives the information without a roll) or do degrees-of-success. The proficiency they choose to apply (ie the approach, set by what questions they ask) distinguishes what kind of information they gather from it. And no one can re-attempt those--they're basically a "collective understanding of the group" check.

This way they almost always get something they can act on. What information exactly they get and how it's phrased depends on who does the asking. The urchin necromancer wizard and the outlander druid might get completely different views on the same information, both colored by their experience and their outlook.

So for inspecting a corpse (assuming it's fresh enough to tell, older might incur disadvantage or might make success impossible):

WIS (Medicine): What killed him?
5+ = injury (rough type/manufactured vs natural) vs something else, very rough time of death
10+ = type of weapon or presence of poison/normal disease/obvious magic, narrower time window
15+ = signs of struggle, less obvious causes of death, narrow time window
20+ = subtle things like "the wound was made to look like it came from a claw, but was actually a dagger."

INT (Nature): "What kind of animal could have killed him?"
5+ = existence of major, common predators
10+ = narrowing it down based on habitat, ruling some things out (it can't be a dire foo, those always eat the entrails)
15+ = You're pretty sure it was a <creature type> or maybe a <different creature>
20+ = It was a Dire Baz, female. Facts about Dire Baz...

WIS (Survival): "Was he alone? Where did he come from? Where did his attackers go?"
5+ = Seems he walked in from the West, here's a matching footprint.
10+ = Looks like there were two others with him, one bigger than the other.
15+ = Attacked by 3 creatures that dropped down from the trees and then went thattaway.
20+ = His companions were a X and a Y, <details about the fight>

[terodil]

I agree that repeating the same roll multiple times can take the excitement out of the game. I also find it too unrealistic to just assume that one ability can be repeated endlessly, even without pressure. A lock might jam. You might run out of ideas where to look for a hidden item. You might simply have completely forgotten that scrap of obscure lore.

I seem to remember reading somewhere (in 5e) about adjudicating ability checks based on a degree of failure and success, but unfortunately I seem unable to find it. Does anybody else recall this? But even without that, there's enough room in the rules for a GM to make stuff more exciting:


If the total equals or exceeds the DC, the ability check is a success [...] Otherwise, it's a failure which means the character or monster makes no progress toward the objective or makes progress combined with a setback determined by the GM.


You might decide that the initial failure makes those checks more difficult to pull off.

So, instead of making rolls completely irrelevant via auto-success in the case of non-combat and non-timepressure, I could see something like this maintain some degree of excitement:

Consider the difference d = DC - ability check when adjudicating the result. E.g., if d > 5, increase the DC. If d > 10, jam the lock. If d > 15, (chance to) break the thieves' tools.
Alternatively, if the check fails, roll a failure dice and pick options like above. I.e., 5% to break the tools. 10% to jam the lock, etc.

It goes without saying that such modifications should be disregarded if they somehow got in the way of keeping the game going.

Edit: Kinda phoenixed. I took far too long to write this post :( -- I personally like the difference approach better for adjudicating than setting separate DCs to represent degrees of success, though.

[PhoenixPhyre]

I seem to remember reading somewhere (in 5e) about adjudicating ability checks based on a degree of failure and success, but unfortunately I seem unable to find it. Does anybody else recall this?


DMG 242

Section headings "Success at a cost", "degrees of Failure", and "Critical success or failure".

Degree of success isn't listed, but it's a simple modification that makes a lot of sense, basically the inverse of degrees of failure.

[Tanarii]

Then, can you rolling-until-sucess? Then opening locked doors is not so hard if no trap is involved.


But after reading the DMG, I have this now. So confused. Seems that the core rules allow to roll as-many-as-want in a rolling-until-sucess style.

Late to the show, but yes, this is something DMs often miss. A bit less so if they're used to the idea of taking 20, but many DMs never grokked that idea.

Of course, you're not supposed to ask a player to roll until success, you're just not supposed to call for a check in the first place. After warning them it will take ten times as long to do it that way.

This is, as you noted, a huge impact on exploration, especially in the "finding things" task. Searching is automatically successful if you can take ten times as long. I even interpret that as PCs can move along at 1/10th the "normal" speed and automatically find traps without passive perception. They can take ten minutes to throughly check out a part of a room instead of a minute, and automatically find anything that can be found. They can take a minute to search all of a fallen enemies pockets instead of six seconds, and no checks are needed. Etc.

As noted, this can also apply to straight forward tasks like picking a lock in a few minutes instead of a few rounds, or breaking down a door or object. (Edit: Although in this case I find many time players choose just one attempt, because its important to their plans to surprise someone on the other side.)

It can also apply to more complicated tasks like recalling some complicated instructions you were given a week ago, taking ten minutes to call them precisely to mind instead of a minute.

What's the cost in these cases? Time is the cost. And if time isn't important in your adventures and campaign, then it might be a problem. Because in that case, it's no longer an important decision for the player to weigh.

[PhoenixPhyre]

What's the cost in these cases? Time is the cost. And if time isn't important in your adventures and campaign, then it might be a problem. Because in that case, it's no longer an important decision for the player to weigh.

I find that time is important when it's important. Mostly it's not, but when it's important it's critical. Which is why I'm not bugged by it. I only track time in certain increments:

* Seconds for combat or other tense situations

* Minutes (could be 1, could be 10, usually more the latter) for small-scale exploration/actions--searching that room is going to be on this scale by default unless they're giving it a quick once-over. And this usually gets brushed over and folded into the hour scale.

* Hours or days for travel.

Basically a strategic (hours, days) scale and a tactical (rounds/turns) scale.

Things happen either very fast or over an indeterminate, but much longer, period of time. Of course, my "plots" tend to be long-running things that aren't in any particular hurry until the party forces a change by getting involved. The status quo will remain such until the party starts kicking over beehives, and then it changes quite quickly.

[Tanarii]

Time is important to my players whenever they explore an adventuring site. Because I use wandering monsters / events. So the decision to take ten minutes to properly search a room vs one minute with passive checks is meaningful. Ditto for the choice to move at 1/10 normal speed vs normal speed with passive checks.

Edit: that's not even counting when they stir up the denizens and have to deal with reactions to their incursion.

[Dark Schneider]

Many failures that seems not to be important are. If you fail to pick lock so can't try again until something changes, you have to find another way, that is important. You should require the wizard to sacrifice a 2nd spell to prepare Knock (if it has, if not, go to copy it to the spellbook), if you have a sorcerer instead, sacrifice a known spell by Knock...
Not beign able to directly open the prison door, you have to search the key, probably getting it from the guard or commander (sleeping at that moment)...
Putting auto-20 on the table is more like a board game, instead role-playing, where failures open new options to advance, and the players must find them.

Then, each check roll should have its tension, and also be balanced to the characters levels and skills. This auto-20 probably shock at begin to users that come from another games where the failure and partial success exist (like our case, because that I created this topic).

I think putting a 20 on the table is excesive, in any case, think that a roll 20 is the most brilliant moment of that character in that skill, so can anyone bright so much each minute by default?.

I think I'll get the way in the middle, as I don't want also to reduce it so much to simply luck, bypassing the characters skills. Then, I think the passive is the key, is the average character skill, and it should be involved in many things concerning skill checks, instead the auto-20.

I think the "degrees" section on DMG fits better in our game than auto-20. This plus actively using the passive as character skill reference. Currently working on it, but could be something like:

- If failed by less than 5 (like DMG says) you can retry.
- If failed by more than 5, cannot retry until something changes. Also could involve consecuences.
- If failed by 10 or more, fatal failure, consecuences.
Passive skill is reference, so:
- Player should declare they want to do something "with care", in this case roll and use the greater one the roll or passive. This should take more time, like 1 minute, or maybe 1d4 minutes. I.e. is not the same to make a tie in 1 round in a hurry, than checking it is well done with your average skill, that could also get a better result than average, but not worse. The time used will depend on the skill and have to determine, i.e. perception could be simply move at half speed, as we could notice that passive perception is normally used when your are not actively looking, at normal pace, but in this case you cannot roll, so allowing to roll and use greater roll or passive requires some sacrifice (you usually want to do this when looking for traps).
- If the DC is not 5 greater than passive, you can always retry, because your average allows to not get a failure (remember is failing by 5 or more). This could help to differentiate the intermediate values from +6 to +9 compared to +5 on the skill, you can accomplish greater DC checks, but it can get some time (multiple checks).
- This, of course, on skills you should be able to retry.

[Erloas]

I think a big part of the nature of "take 20" is more, as the DM, to think about what makes this "thing" an encounter and something worth worrying about. Why bother putting a lock on a door out in the middle of a forest when there is no one inside, no one outside, and nothing about the situation is going to change if you take 10 seconds to unlock the door or 4 hours.

If "take 20" is a legitimate option, there is no time factor, there is no hidden trap that could hurt you, there is no monster on the other side that could hear you and prepare, etc. etc. then why bother putting that there at all? If your players spend 15 minutes of game time getting past the door, get inside, and find absolutely nothing of value, then ... why? why bother. If there is instead a +4 flaming sword behind that door, and the only thing stopping them from getting it is one good or bad roll on a DC check... that seems very poor design.


Now if you're sneaking into a castle, you've got a limited amount of time before you're likely to be noticed, and you're picking every single door you come across, not knowing if one has nothing behind it or that +4 flaming sword, you're doing so knowing that every check you're making is increasing the likelihood that you'll be discovered, that someone will hear you or find a room you've already been in. That is a time when taking 20 doesn't make sense, but it is also a time where the choice to keep trying or moving on is a legitimate question.

For lock picking in particular, saying taking 20 only takes a minute, I don't think is so much an issue with the idea of taking 20 as it is that trying to claim someone can pick a lock in 6 seconds is realistic. It gets defined as 6 seconds based on combat actions (and in combat you can't take 20 anyway), but in most cases skills like that don't really fit into a 6 second combat situation, it is just there because that is the base of the system. One could also say that picking a lock in 6 seconds is significantly harder than picking it in a "normal" amount of time. A stress free attempt at picking the lock might be DC10, but trying to pick that same lock in combat, in 6 seconds could be DC20.
AKA you fail to pick it while trying to escape and end up fighting and killing the guard rather than getting away. Now that the combat is over and you've got a bit less stress picking that same lock could be trivial.

[Tanarii]

Many failures that seems not to be important are. If you fail to pick lock so can't try again until something changes, you have to find another way, that is important.Why though? I mean, that's not a rule anywhere in 5e. It explicitly goes against the 5e rule set. If there's not a explicitly a consequence for failure other than time, and it's possible to succeed, you can take ten times as long and automatically succeed.

"Because I, the DM, say you can only try once" isn't a consequences for failure. At least, not a good one, since the advice in the DMG explicitly tells you as a DM not to run the game that way.

It may hark back to old-school Theives and the Pick Locks ability, where is was the rule and no reason given. But this is a game where they wrote a logical thing into the book. Surprising. That if you aren't stopped from trying again by something changing or it being impossible to accomplish, you can.

Now if there's an explanation for your DM whim, that's fine. Failing to disable the trap will set it off. Failing to unlock the lock on the first try will break the glass in the tubler, locking it shut permanently. Failure to break the door in one blow will alert anyone on the other side automatically. Etc.
(Note in the last case the players can sill choose to have their PCs take ten times as long an automatically succeed, since it doesn't prevent them. It just changes the consequences due to their choice.)

[Knaight]

For lock picking in particular, saying taking 20 only takes a minute, I don't think is so much an issue with the idea of taking 20 as it is that trying to claim someone can pick a lock in 6 seconds is realistic. It gets defined as 6 seconds based on combat actions (and in combat you can't take 20 anyway), but in most cases skills like that don't really fit into a 6 second combat situation, it is just there because that is the base of the system.

It's not even the base of the system so much as something assumed - nowhere does it say that every skill check takes one round, especially when not even in combat. Lock picking taking six seconds by default isn't even the most egregious example of this, something like an intelligence check to research or a charisma check to gather information are even more hilariously off in time* (the skills are explicitly bonuses to attribute checks if using them in particular ways, they're not supposed to restrict what attribute checks can be called for).

[Dark Schneider]

Why though? I mean, that's not a rule anywhere in 5e. It explicitly goes against the 5e rule set. If there's not a explicitly a consequence for failure other than time, and it's possible to succeed, you can take ten times as long and automatically succeed.

"Because I, the DM, say you can only try once" isn't a consequences for failure. At least, not a good one, since the advice in the DMG explicitly tells you as a DM not to run the game that way.

It may hark back to old-school Theives and the Pick Locks ability, where is was the rule and no reason given. But this is a game where they wrote a logical thing into the book. Surprising. That if you aren't stopped from trying again by something changing or it being impossible to accomplish, you can.

Now if there's an explanation for your DM whim, that's fine. Failing to disable the trap will set it off. Failing to unlock the lock on the first try will break the glass in the tubler, locking it shut permanently. Failure to break the door in one blow will alert anyone on the other side automatically. Etc.
(Note in the last case the players can sill choose to have their PCs take ten times as long an automatically succeed, since it doesn't prevent them. It just changes the consequences due to their choice.)
It is not like "I say you can only try one". Notice that I put the option for reroll. But if failed by a large ammount (5 or greater at this moment), then that maneuver is beyond your current capabilities, being this all involved, i.e. for picking locks, it would be your pick lock skill, the knowledge of the lock, your performance at that moment (that is the roll). In fact, is not usual even when we are plenty of time to reach our limits in any task we try in real life, we usually perform our average.

We come from a game (Rolemaster) where the failure exists, and I can ensure that ANY check is interesting if things can go wrong. A simple door?, "it is not interesting", well, put the failure option on the table, if they fail, you will notice how it become interesting, as players will need to find another way to pass.

I put an example: you have to rescue a prisoner, you sneak (rolls involved, fine) up to the prison door, there is a big difference between "you open the door (with your auto-20) because a door is not interesting", to:
- Roll (results):
1) ... 19!, Ah, I did it!
2) ...3!?, WTF!? What can we do now? Let's rest and think... Get the knocked guard suit and time to Disguise and Deception.
Also added the option to allow reroll when near success, to add the time factor.

Why must be something interesting ONLY when there can be passing creatures? Is beign caught the only interesting option? What about the consecuences of not beign able to achieve the task itself? In fact, the last is even much more interesting than the 1st one. The "passing creatures" one usually only bring to a combat, while the "not able to achieve the task", so much more interesting what goes after :)

Even if is a door in the middle of the forest, no guards, no interesting...why? If you fail to open, it adds the fact that you cannot enter to explore, players don't know if there is a powerful item inside, they will try to enter by other ways, and that is interesting by its own.

As said, we come from a game where the failure is an option, I recommend you to try add it to your game, and you will notice all the hilarious situations it will bring to your games. And the tension in the air with each check.

In fact, I notice it is usual to maximize all characters for combat, the "build" concept (like if this was a Diablo game, instead a role-playing), and I think that is precisely because that idea that all the things out of the combat always goes fine, so why to worry about other skills. Because when something goes "wrong", it also usually reachs to a combat anyway.

And is not out of the rules, the good is that in 5E there is no "taking X", in DMG it gives the DM the option of applying something similar, but is not strictly a rule, but at the same time it introduces the failure options (failing by 5 or more, or by 10 or more), and near success (failing by 1 or 2), which fit better in our games, as also mentioned.

[terodil]

DMG 242, Section headings "Success at a cost", "degrees of Failure", and "Critical success or failure".

Degree of success isn't listed, but it's a simple modification that makes a lot of sense, basically the inverse of degrees of failure.

Right, thank you, Phoenix. Must have failed my spot check.

To be honest, I couldn't imagine playing without some variant of degrees. Would seem really dull to me.

[Tanarii]

It is not like "I say you can only try one". Notice that I put the option for reroll.

But it is still very much DM whim. There's no logic or reason behind the determination that a graded roll is necessary in the particular case. You appear to be advocating just applying it globally to all declared actions and cases, then inventing reasons for unsuccessful rolls causing an undesirable outcome after the fact.

As opposed to the clear intent for 5e, per DMG Chapter 8 discussion on resolution*, which is to determine if there is a possibility of an undesirable or positive outcome, or if it's automatic, and then if variable choose what kind of roll is appropriate.

Also, I strongly recommend this article:
https://theangrygm.com/adjudicate-actions-like-a-boss/

---------
*DMG pg 237:
When deciding whether to use a roll, ask yourself two questions:
Is a task so easy and so free of conflict and stress that there should be no chance of failure?
Is a task so inappropriate or impossible- such as hitting the moon with an arrow-that it can't work?
If the answer to both of these questions is no, some kind of roll is appropriate.

This should really have been Chapter 1, the first thing in the DMG. Not how to create your own worlds.

[KorvinStarmast]

This should really have been Chapter 1, the first thing in the DMG. Not how to create your own worlds.yep.

Apparently, a significant number of people either don't have the DMG, or don't bother to read it. Makes having a fruitful discussion difficult sometimes.

[Dark Schneider]

But it is still very much DM whim. There's no logic or reason behind the determination that a graded roll is necessary in the particular case. You appear to be advocating just applying it globally to all declared actions and cases, then inventing reasons for unsuccessful rolls causing an undesirable outcome after the fact.

As opposed to the clear intent for 5e, per DMG Chapter 8 discussion on resolution*, which is to determine if there is a possibility of an undesirable or positive outcome, or if it's automatic, and then if variable choose what kind of roll is appropriate.

Also, I strongly recommend this article:
https://theangrygm.com/adjudicate-actions-like-a-boss/

---------
*DMG pg 237:
When deciding whether to use a roll, ask yourself two questions:
Is a task so easy and so free of conflict and stress that there should be no chance of failure?
Is a task so inappropriate or impossible- such as hitting the moon with an arrow-that it can't work?
If the answer to both of these questions is no, some kind of roll is appropriate.

This should really have been Chapter 1, the first thing in the DMG. Not how to create your own worlds.
Don't worry if I require a roll is because it has relevance. The skills that have auto-success under circumstances are preserved (like long jump your STR, and check if want to jump longer only).

But, a lock is something relevant on its own. If it wasn't, then simply don't lock it. Not all the doors are locked, if you lock one, it is usually because behind there should be something relevant, characters, items, or simply because you want to get the players attention (that is relevant by itself). Then the players probably will try to open by any way, and if they failed the "good" way (picking lock), maybe they use magic or bash the door, making noise.

Currently I have simplified the rule we will use, based on DMG guidelines to keep under game mechanics in any case:

- We have 2 types of checks, short (usually takes 1 round), and long (takes from about 1 hour or longer). Examples of short are pick lock or make a tie, examples of long are identify an item at short rest or research at downtime.
- Short: you can attemp quickly in 1 round, roll and get the result; or you can attemp carefully in 1 minute, roll and get the greater value the roll or your passive. Alternatively the time required is 1d10 rounds, throw both dice d20 and d10 at the same time for check result and time used.
- Long: you decide if roll or use your passive, as it is a task that repeats itself over time, you could use your average directly. So you can try to do something within your capabilities (passive), or try something beyond but with probability of failure. If using passive there is no retry with passive obviously, you should retry with a roll.
- Moving: this is a special kind of check, like looking for traps, in this case you can move at half-speed (that added to the half of the search would be 1/4 speed) to do it carefully.
- Results: if failed by less than 5 can reroll, if failed by 2 with +2 and by 1 with +5. These bonuses was taken from cover (core rules) so it is supposed to be balanced in the system. Dependeing the case, allow to succeed with consecuences (like indicated in DMG). If failed by 5 or more, cannot retry unless something changes (usually a new source of info or something like that). In the case of short tasks, the player decides how to retry, if quickly or carefully on each reroll; for long tasks, add time, i.e. for 1 hour 1-10 minutes per retry, or for 1 month 1-2 days, until getting a result (succeed or failed).

It may seem a bit long to write, but is easy to remember, only need to get in mind the difference of 5, the cover bonuses, and the use of the passive.

Some notes:
- The average determines the character base capabilities, like in real life, that is how all we work.
- Bonuses +6-+9 are used, they allow to not fail greater DC checks. I.e. with the initial +5 you could fail DC20 even if trying carefully, with +6 you can always retry DC20, but taking some time (until rolling 14), that IMHO is realistic to take a few minutes to accomplish a task like that. Notice that with a bonus of +8-+9 you would get directly a +2-+5 to the next retry, so a new difference from a +5, and even from +6-+7 bonuses, of course if doing the task carefully putting your average on it. At top of your skills (+11), you cannot even fail DC25 checks if doing them carefully.
- For long tasks, notice that could use the passive 1st to get the failed by 1-2 bonus, and then retry rolling. It represents well how a more skilled character (its average) could take longer to use its greater skill to achieve something harder.

[Mordaedil]

But take 20 has hard rules around when it's possible and feats that modify when it's possible. Combined with the basic philosophy of 3e that the text is paramount, you get very little room for DM discretion once the facts are established. "Does my character take 10/20" is a player decision made before the roll happens.

Compare to 5e, where the very possibility of a check is left in the hands of the DM, and the "auto-success" rule, like everything in the DMG, is explicitly an option that DMs can employ to speed things up if they wish. Or ignore if they choose. It also happens once they've failed once on the check, not preemptively. It's not a player entitlement, not something they can plan on. What defines "no meaningful consequences for failure?" When is the only concern time? Those are entirely up to the DM's vision of what the situation requires.
I thought DM ruling was one of the things 3.x had as a core component, given so many people were really mad that 4e gave players powers that dictated to the DM what was gold or not and 5e returning to form a bit in this manner.

At least I recall most of the rulings in my sessions being firmly in the hands of my DM, with our players asking "may I?" a whole lot. But I can see the other perspective, especially if you take these forums as the norm, given the "Reading RAW threads" and the constant arguing over RAW. I think at most tables RAI is what rules, people mostly refer to RAW as a framework and recognize it doesn't quite hold as it should, but I've been told I'm wrong before.

Anyway, sorry about that tangent, I think this wraps it about up.

[PhoenixPhyre]

I thought DM ruling was one of the things 3.x had as a core component, given so many people were really mad that 4e gave players powers that dictated to the DM what was gold or not and 5e returning to form a bit in this manner.

At least I recall most of the rulings in my sessions being firmly in the hands of my DM, with our players asking "may I?" a whole lot. But I can see the other perspective, especially if you take these forums as the norm, given the "Reading RAW threads" and the constant arguing over RAW. I think at most tables RAI is what rules, people mostly refer to RAW as a framework and recognize it doesn't quite hold as it should, but I've been told I'm wrong before.

Anyway, sorry about that tangent, I think this wraps it about up.

3.x was designed (according to several developers who worked on that) to limit DM agency as a reaction to a perceived glut of "tyrant" DMs in 2e. They tried (in good faith!) to codify as much as possible so that fewer rulings would have to be made. That, plus internet and a few other developments led, as I understand it, to a rise in seeing a particular variant of the "rules" as being binding on everybody. Each forum has "RAW" interpretations, and they don't always agree. 4e took that to its logical conclusion, streamlining the DM's position to basically a content-provider and rules engine, but everything was codified as much as humanly possible. Made it very simple to DM and to build stuff for, since there was a full-fat framework on which to build. But at the same time, it sometimes felt a bit stifling if you wanted to go off-books. At least to me.

5e is a return to 2e and before, with lessons learned from later editions (including how to write non-labyrinthine rules :smallwink:). It's basically (and intentionally!) a "D&D, best-parts edition". Even in the marketing they call back to "classic D&D experiences/monsters/characters/etc."