you say character 1 and 3 feel like crap, I think you mean player and sorry, you are playing a game based on dice rolls, sometimes the dice roll against you and my experiance every time it happens is the players laugh about it and love the unexpected story beat it throws up.
It is based on dice rolls and modifiers. If the modifiers are enough to succeed on a nat 1, which takes investment, then the player should be allowed to succeed on a nat 1, Save or Ability Check.
If someone built a character to get modifiers high enough to do so, chances are that they are not going to find that 5% auto fail negating their investment to be fun, at least from my experience.
Current 5E RAW and RAI let you succeed on a nat 1 if your modifiers are high enough and it works well.
you say character 1 and 3 feel like crap, I think you mean player and sorry, you are playing a game based on dice rolls, sometimes the dice roll against you and my experiance every time it happens is the players laugh about it and love the unexpected story beat it throws up.
It is based on dice rolls and modifiers. If the modifiers are enough to succeed on a nat 1, which takes investment, then the player should be allowed to succeed on a nat 1, Save or Ability Check.
If someone built a character to get modifiers high enough to do so, chances are that they are not going to find that 5% auto fail negating their investment to be fun, at least from my experience.
Current 5E RAW and RAI let you succeed on a nat 1 if your modifiers are high enough and it works well.
I have been running nat 1 auto fail for years and my tables know it, they like that there is a chance that they might fail at something they are good at because that is realistic. I don't insist on a dice rolls all the time but now and again I will and, when that 1 turns up, the party laugh and love it, I don't run that 20's auto succeed but I do let 20's usually do something cool for an ability check.
When you get to the higher levels of the game you need an element of that chance of failure still there, otherwise the game loses a lot of it's challenge. I can't just keep increasing lock pick DC scores, the locks of the world get harder as you level up.
That failure chance is called challenges hard enough to actually be challenging, not a random 5% chance on everything they do. If I put enough effort into a skill to beat DC 15 checks even on a 1 that means I want to be good enough at that specific thing to succeed on that check even on a 1. I don't want a constant 5% chance of failure just because. That's the point of putting effort into skills. Now if there are some people who are willing to put up with it being otherwise that's fine, but that should definitely be a houserule and nothing official.
As a DM I won't be using this rule ever, I'll be sticking with how it is now and deciding if something special happens on a nat 20 depending on the check and the circumstances.
1 being an auto fail is IMO the worst part of this. You have expertise? A +15 modifier and spent years (maybe hundreds) training this skill. Yeah sorry, now you're going to fail 5% of the time at something you deliberately built your character to be good at. No thanks.
The biggest problem I see with this is contested skill checks (Of various kinds that you may want to do for various reasons) Here's an example of something that could happen at the table, I've labelled them Chars but they could be Monsters/PCS/NPCs.
Char1 tells a lie, they roll 1 on deception but have a +12 modifier. Do the others even need to roll? Well yes, because they could also get a 1 and auto fail at their insight. Char2 rolls insight and rolls a 10 with no plus, they know Char 1 is lying despite Char1 having the higher total score. Char3 rolls insight too, they are an expert at it and have a plus 12 but also roll a 1 so they don't know Char1 was lying despite meeting their total of 13.
Char1 feels like crap, they were caught in a lie by Char2 that Char2 wouldn't previously have been able to tell. Char3 feels like crap, they were unable to tell that Char1 was lying even though their character is an expert and they met the total that was required - to make this worse they'll feel bad that Char2 detected the lie despite them not being particularly good at this usually and having no investment at all into the skill. Char2 is probably the only one feeling ok about this situation - But they aren't going to feel great because they did something great, they didn't really succeed, Char1 failed, it's not the same thing. So... Yay?
Completely agree , i currently play at tables that do use Nat 1 and nat 20 on skill checks being auto fail/success and it is probably my most hated house rule. As someone who does play characters with expertise a lot it always super unsatisfying to fail 5% of the time on stuff that should be easy. Also having players with lower score get better results just because you rolled a nat 1 is so frustrating.
It completely breaks immersion and makes it harder to roleplay the character as someone who is an expert at something if they have such a high chance of failure.
Being an expert doesn't mean you're infallible and 'never' make mistakes. Though I personally wouldn't make someone in my games roll for something where they would have a high enough modifier to hit the dc on a nat 1 anyway. That doesn't seem to me like a good point to call for a roll.
True but 5% is a very high chance of failure for an expert. That's more novice with brief training levels.
I think it becomes more relevant and problematic for contested rolls , or group check type stuff. Where you can't just state that the dc is low enough that a certain character should just be able to do it and not have them roll.
But it is the smallest chance that the game's fundamental d20 system can account for. The system doesn't really support a 2% chance etc without doing something clunky like introducing a d100 roll to skill checks etc.
Personally, I have always played with critical successes and failures on skill checks. It’s… not actually a problem and I have never had a player throw a temper tantrum because the barbarian managed to pick a lock - it’s a simple matter of narrating things in a way that brings out the amusement of an “even a blind hog finds an acorn every now and then.”
It is a 1 in 20 chance to create a lucky situation (that barbarian is unlikely to get advantage on a check they know nothing about) offset by a 1 in 20 chance of the player critically failing (even experts have a bad day sometimes) and possibly making the situation worse by jamming the lock. Those are relatively balanced odds and reflect the simple reality that sometimes things just work out… or don’t.
And, sure, it makes rouges with main character syndrome feel bad when someone else crits on one of the rouge’s dozen or so trained skills, but, if that player is so desperate to always be the one succeeding on checks that 1 in 20 odds upset them, they probably are not fun to play with in the first place.
I think this rule would feel a lot better to people if the don't roll suggestion/rule was something like characters shouldn't roll if the DC is 3-5 higher or lower than their best or worst possible roll. So you have a 5% chance to get a smidgen past your normal max, or normal min. But 5/30 just seems really off a 5% chance a klutz can do what a high level character trained in acrobatics can barely do seems weird a master chef failing at something a 4 year old would make for their mom seems weird. The d20 especially in 5e removes skill a lot as is, I'm not sure they need to double down on it.
Personally, I have always played with critical successes and failures on skill checks. It’s… not actually a problem and I have never had a player throw a temper tantrum because the barbarian managed to pick a lock - it’s a simple matter of narrating things in a way that brings out the amusement of an “even a blind hog finds an acorn every now and then.”
It is a 1 in 20 chance to create a lucky situation (that barbarian is unlikely to get advantage on a check they know nothing about) offset by a 1 in 20 chance of the player critically failing (even experts have a bad day sometimes) and possibly making the situation worse by jamming the lock. Those are relatively balanced odds and reflect the simple reality that sometimes things just work out… or don’t.
And, sure, it makes rouges with main character syndrome feel bad when someone else crits on one of the rouge’s dozen or so trained skills, but, if that player is so desperate to always be the one succeeding on checks that 1 in 20 odds upset them, they probably are not fun to play with in the first place.
Honestly when I was playing a rogue, when I finally got reliable talent I found it to be bitter sweet. When rolling stealth for any dc under 23 was a formality it kind of made things less fun overall. Stripping out that chance for an unexpected setback also inadvertently takes out a fun element of the game. A lot of the more memorable moments in D&D for me aren't just rolling really well, but when some bad rolls create a situation we have to think on our feet to work around or get ourselves out of.
But it is the smallest chance that the game's fundamental d20 system can account for. The system doesn't really support a 2% chance etc without doing something clunky like introducing a d100 roll to skill checks etc.
If the smallest chance is way too high I'd rather them just remove that chance altogether. If someone specifically makes a character who is supernaturally good at doing something then just let them be supernaturally good at doing something.
As a DM I won't be using this rule ever, I'll be sticking with how it is now and deciding if something special happens on a nat 20 depending on the check and the circumstances.
1 being an auto fail is IMO the worst part of this. You have expertise? A +15 modifier and spent years (maybe hundreds) training this skill. Yeah sorry, now you're going to fail 5% of the time at something you deliberately built your character to be good at. No thanks.
The biggest problem I see with this is contested skill checks (Of various kinds that you may want to do for various reasons) Here's an example of something that could happen at the table, I've labelled them Chars but they could be Monsters/PCS/NPCs.
Char1 tells a lie, they roll 1 on deception but have a +12 modifier. Do the others even need to roll? Well yes, because they could also get a 1 and auto fail at their insight. Char2 rolls insight and rolls a 10 with no plus, they know Char 1 is lying despite Char1 having the higher total score. Char3 rolls insight too, they are an expert at it and have a plus 12 but also roll a 1 so they don't know Char1 was lying despite meeting their total of 13.
Char1 feels like crap, they were caught in a lie by Char2 that Char2 wouldn't previously have been able to tell. Char3 feels like crap, they were unable to tell that Char1 was lying even though their character is an expert and they met the total that was required - to make this worse they'll feel bad that Char2 detected the lie despite them not being particularly good at this usually and having no investment at all into the skill. Char2 is probably the only one feeling ok about this situation - But they aren't going to feel great because they did something great, they didn't really succeed, Char1 failed, it's not the same thing. So... Yay?
I feel like this shows you have clearly never played in a game with this rule. Because the players feeling like crap isn't what actually happens. What actually happens is laughter erupts at the table over the bad luck. Char1 Is down hardened at first "ok you guys got me." ... "wow wow wow, we could still roll 1's". Second player 10 "ah ha." Third player Nat 1. Whole table erupts in laughter. "NAT 1 BABY LETS GO!" "I can't believe we both rolled nat 1's." This is what legitimately happens at the table and they tell stories about it with their other DnD friends.
I run a Fantasy Flight Games Star Wars game. The dice have a triumph side and a Despair side. Now Triumph does not mean auto success nor does Despair mean auto fail on a check, but Triumph guarantees something good happens with the check, and Despairs guarantees something bad happens with the check. The despairs are as much fun as the triumphs. Sometimes because they are dreaded, sometimes because they are funny. Always because they add drama to the roll.
The reason that system works so well is because there is actually a success/failure gradient. There isn't just Triumph and Despair, there are also advantages and setbacks, which account for side benefits/penalties. And to go even further, Triumphs/Despairs aren't the main deciders of whether the action succeeds and fails in the first place. So you can get a lot of Triumphs and fail anyways, the same way you can get a lot of Despairs even if you succeed on what you were trying to do (all of this along with however many advantages/setbacks you may have gotten).
Simple to say, this definitely isn't an accurate comparison. What would be a more accurate system for the new edition would be if 1's and 20's didn't determine whether you succeeded, and only the total compared to the target DC did, but instead adding a positive/negative dramatic twist on top of whatever it is you rolled. And even that wouldn't be super accurate because chances are if you've rolled a 1 or a 20 the math of this game works out in a way that you probably are going to succeed/fail anyways just off of the number differences.
As a DM I won't be using this rule ever, I'll be sticking with how it is now and deciding if something special happens on a nat 20 depending on the check and the circumstances.
1 being an auto fail is IMO the worst part of this. You have expertise? A +15 modifier and spent years (maybe hundreds) training this skill. Yeah sorry, now you're going to fail 5% of the time at something you deliberately built your character to be good at. No thanks.
The biggest problem I see with this is contested skill checks (Of various kinds that you may want to do for various reasons) Here's an example of something that could happen at the table, I've labelled them Chars but they could be Monsters/PCS/NPCs.
Char1 tells a lie, they roll 1 on deception but have a +12 modifier. Do the others even need to roll? Well yes, because they could also get a 1 and auto fail at their insight. Char2 rolls insight and rolls a 10 with no plus, they know Char 1 is lying despite Char1 having the higher total score. Char3 rolls insight too, they are an expert at it and have a plus 12 but also roll a 1 so they don't know Char1 was lying despite meeting their total of 13.
Char1 feels like crap, they were caught in a lie by Char2 that Char2 wouldn't previously have been able to tell. Char3 feels like crap, they were unable to tell that Char1 was lying even though their character is an expert and they met the total that was required - to make this worse they'll feel bad that Char2 detected the lie despite them not being particularly good at this usually and having no investment at all into the skill. Char2 is probably the only one feeling ok about this situation - But they aren't going to feel great because they did something great, they didn't really succeed, Char1 failed, it's not the same thing. So... Yay?
I feel like this shows you have clearly never played in a game with this rule. Because the players feeling like crap isn't what actually happens. What actually happens is laughter erupts at the table over the bad luck. Char1 Is down hardened at first "ok you guys got me." ... "wow wow wow, we could still roll 1's". Second player 10 "ah ha." Third player Nat 1. Whole table erupts in laughter. "NAT 1 BABY LETS GO!" "I can't believe we both rolled nat 1's." This is what legitimately happens at the table and they tell stories about it with their other DnD friends.
I run a Fantasy Flight Games Star Wars game. The dice have a triumph side and a Despair side. Now Triumph does not mean auto success nor does Despair mean auto fail on a check, but Triumph guarantees something good happens with the check, and Despairs guarantees something bad happens with the check. The despairs are as much fun as the triumphs. Sometimes because they are dreaded, sometimes because they are funny. Always because they add drama to the roll.
I have only played in games that used this house rule and i absolutely hate it. It is awful to fail a stealth check on a 24 just because you rolled a nat 1 . Or have somone know you're lying on a 10 against your 16. Its totally jarring and immersion breaking and generally makes role playing more difficult if you cant rely on your specialists to actually be able to do anything without advantage,
The reason that system works so well is because there is actually a success/failure gradient. There isn't just Triumph and Despair, there are also advantages and setbacks, which account for side benefits/penalties. And to go even further, Triumphs/Despairs aren't the main deciders of whether the action succeeds and fails in the first place. So you can get a lot of Triumphs and fail anyways, the same way you can get a lot of Despairs even if you succeed on what you were trying to do (all of this along with however many advantages/setbacks you may have gotten).
Simple to say, this definitely isn't an accurate comparison. What would be a more accurate system for the new edition would be if 1's and 20's didn't determine whether you succeeded, and only the total compared to the target DC did, but instead adding a positive/negative dramatic twist on top of whatever it is you rolled. And even that wouldn't be super accurate because chances are if you've rolled a 1 or a 20 the math of this game works out in a way that you probably are going to succeed/fail anyways just off of the number differences.
Correct ya, I was just using a comparison of another game in addition to the moments I have had at the table with the nat 1 and nat 20. I don't typically run the DnD game, but am a player and the DM has used nat 1 fail, nat 20 succeed thing forever, and what I described has definitely happened. But still you are correct that it is much more granular in Star Wars. I do love that system though it is a lot of fun. Not for everyone, or for every game type as I clearly like 5e as well, just it works for the right game.
Our gaming circle has been talking it over and we decided that we would just play skills with the current 5e rules. It works fine with Attacks and Saves, we just don't like the proposed rules for Ability(skill) Checks. That is what we plan to say on the Survey as well.
I guess I am different. I have no problem with my character failing. If you're just going to be the best at everything and never have it go wrong then where is the fun. My hig dex Rogue with expertise in Acrtobatics gets a +10. Occasionally he is going to fall flat on his face doing a simple backflip hge has done 99 times (but that 100th). He will pick himself up and say he meant to do it, If I get a bad luck streak and roll a lot of 1's he will say what the heck is going on and ask his sister to talk to her god and see if he annoyed someone.
But it is the smallest chance that the game's fundamental d20 system can account for. The system doesn't really support a 2% chance etc without doing something clunky like introducing a d100 roll to skill checks etc.
If the smallest chance is way too high I'd rather them just remove that chance altogether. If someone specifically makes a character who is supernaturally good at doing something then just let them be supernaturally good at doing something.
They are though. THey just aren't infallible. A rogue with a +13 to stealth is pretty much supernaturally good at stealth. They just aren't perfect.
If the DM thinks the rogue is good enough that they couldn't feasibly fail then a roll should not be called for.
But it is the smallest chance that the game's fundamental d20 system can account for. The system doesn't really support a 2% chance etc without doing something clunky like introducing a d100 roll to skill checks etc.
If the smallest chance is way too high I'd rather them just remove that chance altogether. If someone specifically makes a character who is supernaturally good at doing something then just let them be supernaturally good at doing something.
They are though. THey just aren't infallible. A rogue with a +13 to stealth is pretty much supernaturally good at stealth. They just aren't perfect.
If the DM thinks the rogue is good enough that they couldn't feasibly fail then a roll should not be called for.
If the rogues bonus is higher than the check they need to make then they definitely shouldn't need to make a roll for it, we can both agree with that. But assuming the DM hasn't memorized everyones bonus and calls for everyone to make a roll, the supernaturally stealthy rogue shouldn't fail to be stealthy if his roll total beats the DC of the check, regardless of what he rolled on die. Sure, if the total is lower than what is needed to beat the check then they should definitely fail the same as anyone else, but if they're that good they should be allowed to be that good.
A key thing to remember about skill checks is: if there's no meaningful cost to trying, and no penalty for failing, there's really no reason to roll in the first place; just tell people they succeed (or that it's impossible). "I'm going to waste (important resource)" is already plenty to discourage people from keeping rolling hoping for a 20.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
It is based on dice rolls and modifiers. If the modifiers are enough to succeed on a nat 1, which takes investment, then the player should be allowed to succeed on a nat 1, Save or Ability Check.
If someone built a character to get modifiers high enough to do so, chances are that they are not going to find that 5% auto fail negating their investment to be fun, at least from my experience.
Current 5E RAW and RAI let you succeed on a nat 1 if your modifiers are high enough and it works well.
I have been running nat 1 auto fail for years and my tables know it, they like that there is a chance that they might fail at something they are good at because that is realistic. I don't insist on a dice rolls all the time but now and again I will and, when that 1 turns up, the party laugh and love it, I don't run that 20's auto succeed but I do let 20's usually do something cool for an ability check.
When you get to the higher levels of the game you need an element of that chance of failure still there, otherwise the game loses a lot of it's challenge. I can't just keep increasing lock pick DC scores, the locks of the world get harder as you level up.
That failure chance is called challenges hard enough to actually be challenging, not a random 5% chance on everything they do. If I put enough effort into a skill to beat DC 15 checks even on a 1 that means I want to be good enough at that specific thing to succeed on that check even on a 1. I don't want a constant 5% chance of failure just because. That's the point of putting effort into skills. Now if there are some people who are willing to put up with it being otherwise that's fine, but that should definitely be a houserule and nothing official.
Completely agree , i currently play at tables that do use Nat 1 and nat 20 on skill checks being auto fail/success and it is probably my most hated house rule. As someone who does play characters with expertise a lot it always super unsatisfying to fail 5% of the time on stuff that should be easy. Also having players with lower score get better results just because you rolled a nat 1 is so frustrating.
It completely breaks immersion and makes it harder to roleplay the character as someone who is an expert at something if they have such a high chance of failure.
Being an expert doesn't mean you're infallible and 'never' make mistakes. Though I personally wouldn't make someone in my games roll for something where they would have a high enough modifier to hit the dc on a nat 1 anyway. That doesn't seem to me like a good point to call for a roll.
True but 5% is a very high chance of failure for an expert. That's more novice with brief training levels.
I think it becomes more relevant and problematic for contested rolls , or group check type stuff. Where you can't just state that the dc is low enough that a certain character should just be able to do it and not have them roll.
But it is the smallest chance that the game's fundamental d20 system can account for. The system doesn't really support a 2% chance etc without doing something clunky like introducing a d100 roll to skill checks etc.
Personally, I have always played with critical successes and failures on skill checks. It’s… not actually a problem and I have never had a player throw a temper tantrum because the barbarian managed to pick a lock - it’s a simple matter of narrating things in a way that brings out the amusement of an “even a blind hog finds an acorn every now and then.”
It is a 1 in 20 chance to create a lucky situation (that barbarian is unlikely to get advantage on a check they know nothing about) offset by a 1 in 20 chance of the player critically failing (even experts have a bad day sometimes) and possibly making the situation worse by jamming the lock. Those are relatively balanced odds and reflect the simple reality that sometimes things just work out… or don’t.
And, sure, it makes rouges with main character syndrome feel bad when someone else crits on one of the rouge’s dozen or so trained skills, but, if that player is so desperate to always be the one succeeding on checks that 1 in 20 odds upset them, they probably are not fun to play with in the first place.
I think this rule would feel a lot better to people if the don't roll suggestion/rule was something like characters shouldn't roll if the DC is 3-5 higher or lower than their best or worst possible roll. So you have a 5% chance to get a smidgen past your normal max, or normal min. But 5/30 just seems really off a 5% chance a klutz can do what a high level character trained in acrobatics can barely do seems weird a master chef failing at something a 4 year old would make for their mom seems weird. The d20 especially in 5e removes skill a lot as is, I'm not sure they need to double down on it.
Honestly when I was playing a rogue, when I finally got reliable talent I found it to be bitter sweet. When rolling stealth for any dc under 23 was a formality it kind of made things less fun overall. Stripping out that chance for an unexpected setback also inadvertently takes out a fun element of the game. A lot of the more memorable moments in D&D for me aren't just rolling really well, but when some bad rolls create a situation we have to think on our feet to work around or get ourselves out of.
If the smallest chance is way too high I'd rather them just remove that chance altogether. If someone specifically makes a character who is supernaturally good at doing something then just let them be supernaturally good at doing something.
I feel like this shows you have clearly never played in a game with this rule. Because the players feeling like crap isn't what actually happens. What actually happens is laughter erupts at the table over the bad luck. Char1 Is down hardened at first "ok you guys got me." ... "wow wow wow, we could still roll 1's". Second player 10 "ah ha." Third player Nat 1. Whole table erupts in laughter. "NAT 1 BABY LETS GO!" "I can't believe we both rolled nat 1's." This is what legitimately happens at the table and they tell stories about it with their other DnD friends.
I run a Fantasy Flight Games Star Wars game. The dice have a triumph side and a Despair side. Now Triumph does not mean auto success nor does Despair mean auto fail on a check, but Triumph guarantees something good happens with the check, and Despairs guarantees something bad happens with the check. The despairs are as much fun as the triumphs. Sometimes because they are dreaded, sometimes because they are funny. Always because they add drama to the roll.
The reason that system works so well is because there is actually a success/failure gradient. There isn't just Triumph and Despair, there are also advantages and setbacks, which account for side benefits/penalties. And to go even further, Triumphs/Despairs aren't the main deciders of whether the action succeeds and fails in the first place. So you can get a lot of Triumphs and fail anyways, the same way you can get a lot of Despairs even if you succeed on what you were trying to do (all of this along with however many advantages/setbacks you may have gotten).
Simple to say, this definitely isn't an accurate comparison. What would be a more accurate system for the new edition would be if 1's and 20's didn't determine whether you succeeded, and only the total compared to the target DC did, but instead adding a positive/negative dramatic twist on top of whatever it is you rolled. And even that wouldn't be super accurate because chances are if you've rolled a 1 or a 20 the math of this game works out in a way that you probably are going to succeed/fail anyways just off of the number differences.
I have only played in games that used this house rule and i absolutely hate it. It is awful to fail a stealth check on a 24 just because you rolled a nat 1 . Or have somone know you're lying on a 10 against your 16. Its totally jarring and immersion breaking and generally makes role playing more difficult if you cant rely on your specialists to actually be able to do anything without advantage,
Correct ya, I was just using a comparison of another game in addition to the moments I have had at the table with the nat 1 and nat 20. I don't typically run the DnD game, but am a player and the DM has used nat 1 fail, nat 20 succeed thing forever, and what I described has definitely happened. But still you are correct that it is much more granular in Star Wars. I do love that system though it is a lot of fun. Not for everyone, or for every game type as I clearly like 5e as well, just it works for the right game.
Our gaming circle has been talking it over and we decided that we would just play skills with the current 5e rules. It works fine with Attacks and Saves, we just don't like the proposed rules for Ability(skill) Checks. That is what we plan to say on the Survey as well.
She/Her Player and Dungeon Master
I guess I am different. I have no problem with my character failing. If you're just going to be the best at everything and never have it go wrong then where is the fun. My hig dex Rogue with expertise in Acrtobatics gets a +10. Occasionally he is going to fall flat on his face doing a simple backflip hge has done 99 times (but that 100th). He will pick himself up and say he meant to do it,
If I get a bad luck streak and roll a lot of 1's he will say what the heck is going on and ask his sister to talk to her god and see if he annoyed someone.
They are though. THey just aren't infallible. A rogue with a +13 to stealth is pretty much supernaturally good at stealth. They just aren't perfect.
If the DM thinks the rogue is good enough that they couldn't feasibly fail then a roll should not be called for.
If the rogues bonus is higher than the check they need to make then they definitely shouldn't need to make a roll for it, we can both agree with that. But assuming the DM hasn't memorized everyones bonus and calls for everyone to make a roll, the supernaturally stealthy rogue shouldn't fail to be stealthy if his roll total beats the DC of the check, regardless of what he rolled on die. Sure, if the total is lower than what is needed to beat the check then they should definitely fail the same as anyone else, but if they're that good they should be allowed to be that good.
A key thing to remember about skill checks is: if there's no meaningful cost to trying, and no penalty for failing, there's really no reason to roll in the first place; just tell people they succeed (or that it's impossible). "I'm going to waste (important resource)" is already plenty to discourage people from keeping rolling hoping for a 20.