So recently there has been much discussion about the new One Dnd playtest content, with critical failure and success being one of the most highly debated topics. I thought I would give some feedback based on my experience working as a behavior analyst. I will start by saying that, I don't think having critical failures and successes is a good idea for the following reasons.
It introduces an unnecessary punishment contingency.
It disproportionately affects high-frequency behavior.
It increases uncertainty
It (might) increase counter-control behavior at the table.
In order to illustrate my points, consider driving as an example. Driving is a very complex activity involving many behaviors that all must be performed reasonably well. Good driving typically involves actively paying attention and executing moment-to-moment decisions whose outcomes could have catastrophic consequences. Because these key decisions require active behavior and could result in severe consequences, we cannot say that a driver automatically succeeds in any driving-related ability check. Having said that. think about all the decisions you have to make over the course of a simple drive to the local grocery store. A simple pedal error (rolling a 1 on a sleight of hand, or foot in this case) could result in you rear-ending the card in front of you. Failing to see a pedestrian crossing sign or noticing pedestrians (rolling a 1 on perception) might result in you hitting someone. Let us assume that a driver makes an average of 20 critical decisions that require ability checks for every mile they drive. Over the course of a simple 5-mile drive, you will encounter at least 1 critical failure if we use the rules presented by One-DnD.
Most of us have at some point been involved in a car accident or know someone who has. Think about what kind of effect that had on their driving behavior. Do you think that they might want to drive more or less after a car accident? For most of us, I think the answer is less, a car crash can be incredibly punishing and reduce our desire to drive at least for a while. Now think, how would encountering a critical failure every 5 miles affect your behavior?
This example was meant to illustrate points 1 and 2. In One-dnd every time, you roll a nat one you will be punished by the system itself. The more you roll, the more often you will encounter this punishment. Imagine a rogue trying to clear a path of traps in a dungeon. He has to succeed in perception to find the traps, and then sleight of hand to access or remove the trap without setting it off, and then a thieves tool check to disarm the trap. A nat 1 in any of these checks can result in catastrophic consequences, yet the more rolls he has to make,t the more that nat 1 becomes a statistical certainty. In One-Dnd, the rogue's proficiency and expertise on the required checks cant save him.
This ties nicely into my 3rd point. Critical successes and failures increase uncertainty in a way that is not fun for anyone. Now before you start typing your angry reply about how DnD is all about uncertainty and rolling the dice, consider the following. Behaviorally speaking, uncertainty is the enemy of appropriate and stable behavior and the progenitor of aberrant behavior. Think about it, when you know what to expect out of a given situation or context, it informs how you should behave. When situations are uncertain and you have no control over certain critical outcomes, then all bets are off, and you have nothing to guide your behavior. Fifth edition Dnd mitigates this negative effect of uncertainty by having things like Ability Score Bonuses, Skill and Tool proficiencies, Expertise, etc. There is a reason why we get excited when we increase our Wisdom to 20 and get that juicy +5 modifier. It is because we know that every single Wisdom-related spell, attack, ability check, and saving throw has now become less uncertain. By adding in critical failures and successes, they are adding uncertainty into the game that cannot be mitigated. You could have had a +13 bonus to your Persuasion roll, and still fail to convince the king to send troupes to aid your village. It doesn't matter if you delivered a rousing and moving speech. If the DM makes you roll for it, there is a 5% that you will fail in a way that feels deeply wrong. Conversely, if your 6(-2) charisma barbarian friend suddenly lets out a loud fart in the room and rolls a nat 20, the king will now agree to send troupes because the barbarian gave him a good laugh.
Finally, we have point 4, counter control. I won't dive too deeply into this point as it is the weakest and most idiosyncratic one of the bunch. All it means is that players are likely to start going out of their way to avoid rolling as much as possible in situations where failure is not an option. They might start looking for ways to break the system to make up for the control they have lost. The player starts to dread rolling a 1 and actively looks for ways to mitigate the outcomes or avoid rolling altogether.
Well, What do you think? Did my points make sense?
Your example of driving a car illustrates the main problem it seems many DMs (and players) have: too many skill checks. You set the baseline at 20 skill checks per mile, which is ridiculously high. A skill check is only supposed to be required if there is pressure (time, hazards, enemies, etc) or if it is difficult enough to cause failure on its own. Most driving does not count as that; and requiring a skill check just to drive down the street (let alone 20 per mile!) is silly.
However, say you are driving along, driving along, kids in the back seat yelling "I gotta go to the bathroom!", you turn and scream "Not now, darnit!", and when you turn back to the road...a truck tire! Suddenly you need to make a snap decision. Do you swerve? Was there a car next to you? Do you have time to check? Do you try to brake? That is a skill check, and (hopefully) not something that happens even once per mile.
You show this again with your example of the rogue. The "Roll Til You Fail" method of DMing is the problem here, not having 1's be automatic failures. If you require 3 or 4 rolls for each 30 ft stretch of dungeon corridor, of course the players will fail.
What this really relates to is not 1 being an auto fail, but rather two things:
1) How often should you have your players roll?
2) How often should your players fail at an action?
The DM controls when players roll. So even if a 1 is a failure, it only happens if you, the DM, as for a roll. You can instead say to the player, "Your skilled rogue easily detects the three traps set by the kobolds. You can either bypass them or disarm them. If you want to reconfigure them it will require a check."
The second part is why I find the distaste of failure on a 1 so weird. I think a game where you have only a 5% chance of failure boring. As you said, why even play this game if you don't ever want to roll, if you never want risk? Every time a player makes a skill check, attack roll, or saving throw, there is going to be a decent chance of failure. That should be part of the fun. You get more of a thrill when you sometimes fail.
You then make a slippery slope argument: if there is a chance to fail, you have no idea what will happen, so you might as well act randomly. But does this apply even when you fail on numbers OTHER than a 1? What if a character with +27 to persuasion attempts to convince a king to give up their kingdom. Is it wrong for the DM to set the DC at 45?
The problem, as I said, occurs if the DM makes that same character (or any character) roll a persuasion check (or any skill) every 3 minutes just because they are talking to someone (or walking, etc). But in situations that call for it, there SHOULD be a chance to fail.
Take your example of the character with +13 to Persuasion who attempts to convince a king to send troops to aid the village. The player gives a rousing and moving speech. Why is the DM still asking for a roll? I would only ask for one to see how much EXTRA support they get. Or if there was some other factor in play (the king was already dealing with an uprising/invasion/lack of funds/etc, the grand vizier who hates the PCs already convinced the king not to send aid, or any other story based reason). These vacuum examples are not how the game is actually played. It isn't a video game. You don't just blindly roll because the rules say to roll and then your hands are tied. This covers the other side of the spectrum as well. If ANY player (not just a low charisma character) said they are farting loudly in the throne room, why is the DM even asking for a roll? Would you ask for a roll in battle if I said, "My character closes their eyes, stands on one foot, and with their left hand softly lobs a dagger in a random direction. Don't I get to roll because a 20 is an automatic hit?" My answer will always be "No".
Your final point is actually part of what makes the game great, and also runs counter to the entire argument against Natural 1 fails. If it is bad for players to try to come up with ways to auto succeed, why is it good for players to want to auto succeed on rolls? You can't have it both ways. I think it is better for players to have to be creative and think up solutions rather than just say, "I made a build with bonuses so high I can't fail." The first encourages roll playing, and investment into the game. The second just encourages numbers.
I agree with the above poster, you do not have to roll for every single thing, every single time.
Simple actions probably should not require a roll unless there is some unusual things going on. in this case you should probably just use a 10+skill mod to represent normal use or maybe 6 or 8 + skill mod.
Your example of driving a car illustrates the main problem it seems many DMs (and players) have: too many skill checks. You set the baseline at 20 skill checks per mile, which is ridiculously high. A skill check is only supposed to be required if there is pressure (time, hazards, enemies, etc) or if it is difficult enough to cause failure on its own. Most driving does not count as that; and requiring a skill check just to drive down the street (let alone 20 per mile!) is silly.
However, say you are driving along, driving along, kids in the back seat yelling "I gotta go to the bathroom!", you turn and scream "Not now, darnit!", and when you turn back to the road...a truck tire! Suddenly you need to make a snap decision. Do you swerve? Was there a car next to you? Do you have time to check? Do you try to brake? That is a skill check, and (hopefully) not something that happens even once per mile.
Let me start by saying this. Driving is an active activity rather than a passive activity. Although you may sometimes feel like driving is "Effortless" or "Thoughtless" it is anything but that. We have simply learned through practice and experience. We have streamlined the process to such a degree that we no longer need to give pause for the decisions we make, but we are, nonetheless making those decisions. You can actually test this for yourself. Think back to when you first started to drive! Was everything effortless for you then? The next time you go for a drive, consider that you are actively regulating your speed, and your proximity to other vehicles, you are timing your stops, you are looking for gaps in traffic and swapping lanes. Does any of that sound like something one just automatically succeed on?
My example of 20 critical decisions per mile was actually quite an underestimation. Most places I looked at place the number way higher than 100 per mile driven.
The second part is why I find the distaste of failure on a 1 so weird. I think a game where you have only a 5% chance of failure boring. As you said, why even play this game if you don't ever want to roll, if you never want risk? Every time a player makes a skill check, attack roll, or saving throw, there is going to be a decent chance of failure. That should be part of the fun. You get more of a thrill when you sometimes fail.
You then make a slippery slope argument: if there is a chance to fail, you have no idea what will happen, so you might as well act randomly. But does this apply even when you fail on numbers OTHER than a 1? What if a character with +27 to persuasion attempts to convince a king to give up their kingdom. Is it wrong for the DM to set the DC at 45?
You missed the point I was trying to make. In Dnd, uncertainty IS THE ENEMY. Every time we get a stat increase or a weapon with a magical enchantment or a feature that gives us an advantage, we get happy because we have now made the game less uncertain. The critical failure on nat 1, invalidates all of that. Invalidates the choices you made. Oh, so you have +14 on stealth because you took proficiency and expertise as your class features? well in Dnd 5e you roll a nat 1 on stealth and that's still a 15, a pretty decent roll. However, in One Dnd, that nat 1 means you fail, regardless of your choices, bonuses, spells, items, etc.
I think the main point which I think you also missed is that this is a punishment contingency built right into the system item. I want you to imagine the doorknob on your front door. What if I told you, that from now on, every time you touch your door knob, there is a 5% that it will burn your hand to the bone. There is also nothing you can do about it. It doesn't matter how fast you touch it or whether you are wearing some sort of protection, no matter what you do, there is this ever-present unmitigable 5% chance that you will be punished.
How would this knowledge affect your "Doorknob touching behavior"? Do you think you would be more likely, or less likely to want to touch said doorknob on a regular basis? You might start to want to circumvent having to touch the doorknob at all. Maybe you start entering the house through the window instead. You will start to engage in aberrant behaviors.
Dnd 5e allows for mitigation of uncertainty. Even if you roll a nat 1 to touch the doorknob, your bonuses might still save you from the worst possible outcome.
You make my point. A novice driver, like a novice character, will have a large chance to fail. Are you saying that even a 1st level character should always succeed? And a skilled, higher level character shouldn’t be rolling for mundane tasks, just as you say an experienced driver no longer has to actively think about every move. A decision point while driving doesn’t equate to a skill check. Because if you are rolling it has to mean there is a chance of failure. Otherwise you are just wasting everyone’s time. But even the best drivers can get into a collision.
Regarding your doorknob example, of course I wouldn’t touch it! That is an example of a high risk, low reward action. Now how about this…same scenario but behind the door is a pile of treasure. You better believe I am taking the chance. Again: skill checks are for situations that call for it. A character wants to run across a room? No check. A character wants to parkour across a room and use the momentum to smash one orc into the other orcs? That is a check and the action might fail.
Finally, why are you asking the player to roll if they can’t fail?
You make my point. A novice driver, like a novice character, will have a large chance to fail. Are you saying that even a 1st level character should always succeed? And a skilled, higher level character shouldn’t be rolling for mundane tasks, just as you say an experienced driver no longer has to actively think about every move. A decision point while driving doesn’t equate to a skill check. Because if you are rolling it has to mean there is a chance of failure. Otherwise you are just wasting everyone’s time. But even the best drivers can get into a collision.
Regarding your doorknob example, of course I wouldn’t touch it! That is an example of a high risk, low reward action. Now how about this…same scenario but behind the door is a pile of treasure. You better believe I am taking the chance. Again: skill checks are for situations that call for it. A character wants to run across a room? No check. A character wants to parkour across a room and use the momentum to smash one orc into the other orcs? That is a check and the action might fail.
Finally, why are you asking the player to roll if they can’t fail?
It looks to me like, you are working from a conclusion rather than toward one. You have already decided what your stance on this issue is, and won't be convinced otherwise, regardless of how well or how much I explain my points. This is fair since I too have presented my own stance to the best of my ability.
Nevertheless, I shall try one more time. You ready? here goes.
Uncertainty is bad. Crit failure is punishing. It is invalidating your choices 5% of the time. The more your roll the more you get punished by the system, in a way that is outside of your control.
I disagree and think the rule is great. As the other posters mentioned, it is up to the DM to manage this appropriately. Attack rolls in combat already use this mechanic, and I've never heard anyone complain about it. I actually use it as a house rule for all my campaigns as well, as it lends an element of failure and challenge to the players that is much needed. If there is something that shouldn't have an auto-fail, then make an exception. But honestly, why are you even rolling for that check anyway? Most players get so many modifiers and advantage to their rolls that the auto-fail on a 1 helps balance the chances much better IMHO. I love the rule and glad they are including it in the new ruleset.
If you're over 30, roll a con saving throw to avoid a spasm. On a 1, you are paralyzed for the rest of your life.
Roll cha check to gain the will to get up. On a 1, you remain in a catatonic state for the rest of the day.
Roll dex (sleight of hand) check to put on your shirt. On a 1, you choke yourself with a t-shirt and drop to 0 hit points.
Roll dex (sleight of hand) check to put on your pants. On a 1, you fall and break a leg, ripping the pants in the process.
Roll con (athletics) check to walk to the toilet. On a 1, you fall midway in the corridor, breaking a random limb or your spine.
Roll dex check to urinate. On a 1, you piss all over your bathroom. Oh god, it's dripping from the ceiling.
Roll str check to brush your teeth. On a 1, you lose them.
Roll wis (survival) check to navigate towards the kitchen. On a 1, you become lost in your 2-room apartment.
Roll int check to remember how to cook breakfast. On a 1, you get amnesia. What was your name again?
Roll dex check to cook breakfast. On a 1, you burn down your kitchen and the entire house, while trying to make a sandwitch.
Roll con check to eat breakfast. On a 1, you choke to death.
Roll dex (sleight of hand) check to put on your shoes. On a 1, you break a leg, if you still have any unbroken legs by this point.
Roll dex (sleight of hand) check to tie shoelaces. On a 1, you break any fingers or other appendages that you used to tie shoelaces, and they are restrained by the shoelaces.
Roll dex (thieves' tools) check to unlock the door. On a 1, you break the door lock. Neighbors called the police on you for attempted break-in.
You don't need to make a check to make every step) I normally only allow players to make a check with the same task only once between long rests. A check is not one attempt, it's the sum of all your attempts until you either succeed or give up. And realistically, a check is only due in a situation where there's true uncertainty, an element of blind luck, with both chances of success and failure.
If you do a lot of rolling in your game do you, roll before every placement of your foot before you move? If so then in your game you may want to have a two tier system in which some rolls have a crit and fail option and some do not.
But in general I like and think the present rule in 5e is great (for what it does) and should not be changed.
Also how a 1 or a 20 is dealt with in some skill checks could be important in games. I have seen some GM rule a 1 is like Kamchatmonk lists actions and results above and others just say you fail a check to put on your boots with a 1 then it takes a little longer. So success means putting your cloths and shoes on in 5 units on a nat 20 you do it in 3 units and on a nat 1 you do it in 9 units. Again it does not mean you fail you walking and chewing gum roll and you die or you fail your perception roll while fishing so a fish jumps down your open mouth and you die.
Kamchatmonk and spammdc, I think you guys are misunderstanding what I was trying to say. I am full in agreement with you, that rolls are not needed for everything. In fact, one of the first forum threads I made in Dnd beyond was about using passive abilities (passive perception, insight etc) as a basis for allowing players to automatically succeed and reduce unnecessary dice rolls.
The main idea is that regardless of what you have to roll for, the more frequently you have to roll the more you contact the negative outcome.
I believe that critical failure rules will result in even more unnecessary rolling. For instance, In my brother's game, I play a paladin. I have 20 in Constitution and at level 4 I took resilient constitution to gain proficiency in con saving throws. I have 18 in charisma and a cloak of protection, giving me a total bonus to my constitution saving throws of +10. Meaning that when I am concentrating on a spell, and I take damage lower than 22, even if I were to roll a natural one on my concentration saving throw I automatically succeed. Most of the time, I don't even have to roll for concentration. Now under One-dnd rules, I would have to roll every single time, because there is always the chance I might roll a 1 and auto fail.
How does this reflex on me and my choices in the game? I chose to allocate my racial points in Con, I chose to take a Con feat and gain proficiency in con saving throws precisely because I wanted my character to be able to hold concentration without issue
I see the problem is that on any roll the chance of failure is at least 5% and there is (usually) a known outcome if failure occurs, and (as written) their are onyly two results, success or failure, though sometimes there are exceptions so if you "fail" by at least 5 you get a worse failure (it is not clear yet how one D&D will treat nat 1s in such cases)
Going back to othe driving example, I have driven about 300,000 miles in my life, lets say for a 5 mile journey we say there is one check to see if I press the correct peddle every time. By rolling a d20 with a failure on a 1, that would leed to 60,000 rolls and about 3000 failures all of which resulin an accident. In real life I might have pressed the wrong pedal 50 times and on every one I was able to correct it before an acciden occurs. If a failure is counted as an error resultin in a accident even if you only have 1 check for a 10 mile journey everyone is going to crash every 200 miles. The solution here is to only allow rolls if the probability of failure is at least (say) 2.5%, If the chance of failure is between 2.5% and 7.5% the DC is such that they can only fail on a 1 (i.e. rounding the chance of failute t o5%), if the chance of failure is less than 2.5% there is no roll and the chance of failure is rounded to 0.
It also applies for successes, using the same approach any athlete , could break the world record with an average of 20 attempts.
I will often use very "scaled" ability checks particularly for knowledge bases checks. Say the party find a document that refers to Umberlee I might ask all the players to roll religion to see if the name means anything. On a 10 they might have heard of a sailor that referred to them in a way that implied they were a diety, on a 15 they might know she is generaly considdered evil going to to knowing pretty much everything about her on a roll of (say) 28. Currently the barb with a religion modifer of -1 who rolls a nat 20, knows exactly the same as the wizard (religion +8) who rolls an 11. In one D&D does the barb know absolutely everything about Umberlee? in which case should I not allow him to roll (so we can not fnd out if he has heard the name at all)
OneD&D also encourages players to demand more ability checks (at least those where the impact of failure is neglibible). In many groups PCs will be looking for birds in the air, insects on the ground, asking if they think the barmaid if lying when she says beer is 5 cp and so on until they get a nat 20 and inspiration but that is a whole other issue.
Kamchatmonk and spammdc, I think you guys are misunderstanding what I was trying to say. I am full in agreement with you, that rolls are not needed for everything. In fact, one of the first forum threads I made in Dnd beyond was about using passive abilities (passive perception, insight etc) as a basis for allowing players to automatically succeed and reduce unnecessary dice rolls.
The main idea is that regardless of what you have to roll for, the more frequently you have to roll the more you contact the negative outcome.
I believe that critical failure rules will result in even more unnecessary rolling. For instance, In my brother's game, I play a paladin. I have 20 in Constitution and at level 4 I took resilient constitution to gain proficiency in con saving throws. I have 18 in charisma and a cloak of protection, giving me a total bonus to my constitution saving throws of +10. Meaning that when I am concentrating on a spell, and I take damage lower than 22, even if I were to roll a natural one on my concentration saving throw I automatically succeed. Most of the time, I don't even have to roll for concentration. Now under One-dnd rules, I would have to roll every single time, because there is always the chance I might roll a 1 and auto fail.
How does this reflex on me and my choices in the game? I chose to allocate my racial points in Con, I chose to take a Con feat and gain proficiency in con saving throws precisely because I wanted my character to be able to hold concentration without issue
Thanks for your further explanation of your position:
A) Yes the more you roll the more chance you have for rolling a 1.
B) You do not need to roll for everything....thus some clear rules on when you would need to roll and when not to is important. If you want some guidance then maybe say if you take damage equal to or greater then 1/2 or 3/4 your concentration check you have to make a concentration roll. A simple rule and you can have a spot on the PC sheet for the number. Note: I would also provide guidance when other factors would be involved if the GM or group wanted, ie heat, cold, tired, climbing, running, etc.
C) Feat: I agree that feats can and should be reworked to include an idea that if you roll a 1 you may get a second roll before failure. I also have talked to people who say that is what inspiration is for or spells and magic items.
D) 5e is great for its simplicity but that also makes somethings harder because of its simplicity focus. I am also not saying that more complex games would necessarily solve your problem.
E) Unfortunately I am remembering many past discussions about this same topic that I have had and read down through the years, the most recient being a couple of years ago when the 1D&D playtest doc was released for the first time and during the Pathfinder II playtest.
F) Too many rolls can be a problem just as too few rolls can be a problem. And the game should look at this as well as your group.
G) I want my PC to do this but this happens: This is a tough area to deal with, player expiations vs how the game works and can be a GM and or game problem. For example I have played in a game where a player said "we just fought a dragon and now I am going to die by not making a climbing roll". The issue was in his mind having done something that was very heroic meant his PC did not have to deal with common things anymore.
I do think we are partly on the same page but I do not think simply saying you do not need to roll or take away crits and fumbles is a good idea....instead I think it should be handled by better feat and or class abilities. Note: this is the same point I made last time the 1D&D doc was released.
Let me provide another example: I create a PC who focus is dealing lots of damage, should they do max damage all of the time? I think most people would say no, some might say they should do max damage under specific situations and some may say yes (for various reasons). But in 5e as it is now that max damage just because its my PC focus, is a dramatic change to the game.
For your example I would look at your game and see if inspiration, feat expansion or class ability expansion would work the best. Just guessing I would either have a Warcaster II feat or change Warcaster as it is worded now to provide a less chance of fumbling a concentration check.
Now under One-dnd rules, I would have to roll every single time, because there is always the chance I might roll a 1 and auto fail.
This is where things go wrong. Nowhere in the rules document does it talk about "needing to roll". People are reading "If you roll, a 1 is a failure and a 20 is a success" and translating it to "Because a 1 is a failure and 20 is a success, you HAVE to roll for every single action, no matter how mundane or absurdly impossible."
I will often use very "scaled" ability checks particularly for knowledge bases checks. Say the party find a document that refers to Umberlee I might ask all the players to roll religion to see if the name means anything. On a 10 they might have heard of a sailor that referred to them in a way that implied they were a diety, on a 15 they might know she is generaly considdered evil going to to knowing pretty much everything about her on a roll of (say) 28. Currently the barb with a religion modifer of -1 who rolls a nat 20, knows exactly the same as the wizard (religion +8) who rolls an 11. In one D&D does the barb know absolutely everything about Umberlee? in which case should I not allow him to roll (so we can not fnd out if he has heard the name at all)
OneD&D also encourages players to demand more ability checks (at least those where the impact of failure is neglibible). In many groups PCs will be looking for birds in the air, insects on the ground, asking if they think the barmaid if lying when she says beer is 5 cp and so on until they get a nat 20 and inspiration but that is a whole other issue.
I also use scaled checks (and I would guess many, if not most, DMs do. The Playtest document says nothing to invalidate that. A 20 being a success does not mean the character will "know absolutely everything", it just means the roll was a success; they know at least something. You can still set threshold numbers; the only change is that the lowest threshold is always 20. (Side issue is the view that a low Int character can never know anything about anything. Why can't the barbarian have learned about Umberlee at one point? They are a bit slow, but they aren't incapable of learning and remembering. That is exactly why the d20 is a lot more impactful than the stat bonuses. At level 1, the best character is only +6 compared to the worst character, leaving many cases where the weakest will beat the highest.
You make a similar mistake in your interpretation of the Playtest. Players NEVER get to demand a roll. This has always been the case. Players describe what they want their character to do, and the DM decides how to resolve the action. This might involve a d20 Test. A player should never say, "I want to roll a d20 Test."
Even if players try to be sneaky and describe their actions as you wrote ("I look around for birds in the air", "I look for insects on the ground", "I watch the barmaid carefully...5 cp seems like a lot for a drink...I think she might be trying to rip me off."), none of these would require a check. "You see some crows." "You see some ants." "She rolls her eyes and points at a slab of slate behind the bar. Scrawled in chalk are the words 'Drinks...5cp'."
Players trying such shenanigans will quickly realize they are wasting everyone's time and move on.
I disagree and think the rule is great. As the other posters mentioned, it is up to the DM to manage this appropriately. Attack rolls in combat already use this mechanic, and I've never heard anyone complain about it. I actually use it as a house rule for all my campaigns as well, as it lends an element of failure and challenge to the players that is much needed. If there is something that shouldn't have an auto-fail, then make an exception. But honestly, why are you even rolling for that check anyway? Most players get so many modifiers and advantage to their rolls that the auto-fail on a 1 helps balance the chances much better IMHO. I love the rule and glad they are including it in the new ruleset.
It works in combat as a 1 is almost always a miss anyways outside of slimes and zombies until absurdly high levels you miss on a 1 just because you don't hit the AC. If at level 5 you would hit 25% of the enemies on a 1 I suspect people would look at it different.
I see the problem is that on any roll the chance of failure is at least 5% and there is (usually) a known outcome if failure occurs, and (as written) their are onyly two results, success or failure, though sometimes there are exceptions so if you "fail" by at least 5 you get a worse failure (it is not clear yet how one D&D will treat nat 1s in such cases)
Going back to othe driving example, I have driven about 300,000 miles in my life, lets say for a 5 mile journey we say there is one check to see if I press the correct peddle every time. By rolling a d20 with a failure on a 1, that would leed to 60,000 rolls and about 3000 failures all of which resulin an accident. In real life I might have pressed the wrong pedal 50 times and on every one I was able to correct it before an acciden occurs. If a failure is counted as an error resultin in a accident even if you only have 1 check for a 10 mile journey everyone is going to crash every 200 miles. The solution here is to only allow rolls if the probability of failure is at least (say) 2.5%, If the chance of failure is between 2.5% and 7.5% the DC is such that they can only fail on a 1 (i.e. rounding the chance of failute t o5%), if the chance of failure is less than 2.5% there is no roll and the chance of failure is rounded to 0.
It also applies for successes, using the same approach any athlete , could break the world record with an average of 20 attempts.
I will often use very "scaled" ability checks particularly for knowledge bases checks. Say the party find a document that refers to Umberlee I might ask all the players to roll religion to see if the name means anything. On a 10 they might have heard of a sailor that referred to them in a way that implied they were a diety, on a 15 they might know she is generaly considdered evil going to to knowing pretty much everything about her on a roll of (say) 28. Currently the barb with a religion modifer of -1 who rolls a nat 20, knows exactly the same as the wizard (religion +8) who rolls an 11. In one D&D does the barb know absolutely everything about Umberlee? in which case should I not allow him to roll (so we can not fnd out if he has heard the name at all)
OneD&D also encourages players to demand more ability checks (at least those where the impact of failure is neglibible). In many groups PCs will be looking for birds in the air, insects on the ground, asking if they think the barmaid if lying when she says beer is 5 cp and so on until they get a nat 20 and inspiration but that is a whole other issue.
Yup, I also do this in my games. Everything is scaled. You might fail a check but depending on how close you came to the DC, you might get different results. I like the idea of a gradient of outcomes from good to worst and everything in between.
However, this isn't very applicable to saving throws since they are usually a very binary thing. Either your pass or you fail.
Now under One-dnd rules, I would have to roll every single time, because there is always the chance I might roll a 1 and auto fail.
This is where things go wrong. Nowhere in the rules document does it talk about "needing to roll". People are reading "If you roll, a 1 is a failure and a 20 is a success" and translating it to "Because a 1 is a failure and 20 is a success, you HAVE to roll for every single action, no matter how mundane or absurdly impossible."
Well, I highlighted a specific example in which you HAVE to roll every time ----> CONCENTRATION
Since I have a +10 bonus to my concentration saving throw (level 10 paladin) the lowest I can roll on a con save is 11, meaning that I automatically succeed in every concentration check originating from taking 22 damage or lower. However, in One-Dnd I have to roll every time because if I roll a 1 I will lose concentration no matter how high my bonus is.
I do not have a problem with the possibility of losing concentration but I do have an issue with always making concentration checks. IMHO it is a good balancing mechanism as well as other ways to roll a failed check again, inspiration, abilities, spells and magic items. Note ins general I have trouble with abilities that are essentially meta gaming, in your case how would someone who is not you know you failed your concentration check before it failed. One way it does work is if a person casts spells or uses abilities before it is needed, yes this does mean it might go unused.
Do you have an issue with a nat 1 fumble on attacks rolls?
Kamchatmonk and spammdc, I think you guys are misunderstanding what I was trying to say. I am full in agreement with you, that rolls are not needed for everything. In fact, one of the first forum threads I made in Dnd beyond was about using passive abilities (passive perception, insight etc) as a basis for allowing players to automatically succeed and reduce unnecessary dice rolls.
The main idea is that regardless of what you have to roll for, the more frequently you have to roll the more you contact the negative outcome.
I believe that critical failure rules will result in even more unnecessary rolling. For instance, In my brother's game, I play a paladin. I have 20 in Constitution and at level 4 I took resilient constitution to gain proficiency in con saving throws. I have 18 in charisma and a cloak of protection, giving me a total bonus to my constitution saving throws of +10. Meaning that when I am concentrating on a spell, and I take damage lower than 22, even if I were to roll a natural one on my concentration saving throw I automatically succeed. Most of the time, I don't even have to roll for concentration. Now under One-dnd rules, I would have to roll every single time, because there is always the chance I might roll a 1 and auto fail.
How does this reflex on me and my choices in the game? I chose to allocate my racial points in Con, I chose to take a Con feat and gain proficiency in con saving throws precisely because I wanted my character to be able to hold concentration without issue
Look at the bright side, if you receive 60 damage, you now have a chance to keep your concentration.
I think part of the problem here, as with so many issues in D&D is that we're not just dealing with the rule, but the culture and baggage surrounding the rule. The playtest packet does not have a "critical failure" rule, it has a rule that says that a roll of 1 always fails. Taking the rule itself, the only issue is that is raises the question of why anyone would be rolling if there's no chance of failure outside of the automatic failure rule. If the DM shouldn't be calling for rolls when there's no real chance of failure, why is the automatic failure rule needed in the first place? There are a handful of situations where a roll is always required (attack rolls, saving throws), so the rule could be narrowly intended to cover those situations. But there's also a long-standing tradition in D&D of DMs reflexively calling for rolls because they're stalling for time, they aren't sure what should happen, or they want to use the result as a narrative prompt. This is where the issue of induced failure that Stusano is highlighting can crop up, as the DM probably isn't stopping to gauge whether there truly is a reasonable chance of failure (this also works in reverse with the automatic success rules; the DM just wanted to gauge how far the PC gets, but they rolled a twenty and now they expect everything). This part could be addressed with a sidebar explaining what the auto success/fail rule is for and when it is meant to come up, but it's an uphill battle.
The other part of the problem is the influence of prior "critical" failure rules. RAW, rolling a 1 and automatically failing is no worse than a roll that narrowly misses the DC, but we expect it to be worse. Natural 20's mean the player gets to narrate something awesome, right? That must mean a Natural 1 should result in something awful, is how the logic goes. So where a normally failed driving roll might mean something like "you missed your turn," a Natural 1 becomes "you plow into a parked car at full speed." Your character gets put into clown shoes and looks utterly incompetent at something they're supposed to be an expert at, which is okay on very rare occasions, but a 5% chance isn't all that rare. And to reiterate, this isn't actually in the rules as written, any more that a "critical success" on a Natural 20 is in the rules, but there's a strong cultural push towards "20's are awesome and 1's are terrible," and if the 1's feel disproportionately bad (as they often do), then it will have the effect Stusano describes of discouraging players from rolling when they can get around it.
So recently there has been much discussion about the new One Dnd playtest content, with critical failure and success being one of the most highly debated topics. I thought I would give some feedback based on my experience working as a behavior analyst. I will start by saying that, I don't think having critical failures and successes is a good idea for the following reasons.
In order to illustrate my points, consider driving as an example. Driving is a very complex activity involving many behaviors that all must be performed reasonably well. Good driving typically involves actively paying attention and executing moment-to-moment decisions whose outcomes could have catastrophic consequences. Because these key decisions require active behavior and could result in severe consequences, we cannot say that a driver automatically succeeds in any driving-related ability check. Having said that. think about all the decisions you have to make over the course of a simple drive to the local grocery store. A simple pedal error (rolling a 1 on a sleight of hand, or foot in this case) could result in you rear-ending the card in front of you. Failing to see a pedestrian crossing sign or noticing pedestrians (rolling a 1 on perception) might result in you hitting someone. Let us assume that a driver makes an average of 20 critical decisions that require ability checks for every mile they drive. Over the course of a simple 5-mile drive, you will encounter at least 1 critical failure if we use the rules presented by One-DnD.
Most of us have at some point been involved in a car accident or know someone who has. Think about what kind of effect that had on their driving behavior. Do you think that they might want to drive more or less after a car accident? For most of us, I think the answer is less, a car crash can be incredibly punishing and reduce our desire to drive at least for a while. Now think, how would encountering a critical failure every 5 miles affect your behavior?
This example was meant to illustrate points 1 and 2. In One-dnd every time, you roll a nat one you will be punished by the system itself. The more you roll, the more often you will encounter this punishment. Imagine a rogue trying to clear a path of traps in a dungeon. He has to succeed in perception to find the traps, and then sleight of hand to access or remove the trap without setting it off, and then a thieves tool check to disarm the trap. A nat 1 in any of these checks can result in catastrophic consequences, yet the more rolls he has to make,t the more that nat 1 becomes a statistical certainty. In One-Dnd, the rogue's proficiency and expertise on the required checks cant save him.
This ties nicely into my 3rd point. Critical successes and failures increase uncertainty in a way that is not fun for anyone. Now before you start typing your angry reply about how DnD is all about uncertainty and rolling the dice, consider the following. Behaviorally speaking, uncertainty is the enemy of appropriate and stable behavior and the progenitor of aberrant behavior. Think about it, when you know what to expect out of a given situation or context, it informs how you should behave. When situations are uncertain and you have no control over certain critical outcomes, then all bets are off, and you have nothing to guide your behavior. Fifth edition Dnd mitigates this negative effect of uncertainty by having things like Ability Score Bonuses, Skill and Tool proficiencies, Expertise, etc. There is a reason why we get excited when we increase our Wisdom to 20 and get that juicy +5 modifier. It is because we know that every single Wisdom-related spell, attack, ability check, and saving throw has now become less uncertain. By adding in critical failures and successes, they are adding uncertainty into the game that cannot be mitigated. You could have had a +13 bonus to your Persuasion roll, and still fail to convince the king to send troupes to aid your village. It doesn't matter if you delivered a rousing and moving speech. If the DM makes you roll for it, there is a 5% that you will fail in a way that feels deeply wrong. Conversely, if your 6(-2) charisma barbarian friend suddenly lets out a loud fart in the room and rolls a nat 20, the king will now agree to send troupes because the barbarian gave him a good laugh.
Finally, we have point 4, counter control. I won't dive too deeply into this point as it is the weakest and most idiosyncratic one of the bunch. All it means is that players are likely to start going out of their way to avoid rolling as much as possible in situations where failure is not an option. They might start looking for ways to break the system to make up for the control they have lost. The player starts to dread rolling a 1 and actively looks for ways to mitigate the outcomes or avoid rolling altogether.
Well, What do you think? Did my points make sense?
Your example of driving a car illustrates the main problem it seems many DMs (and players) have: too many skill checks. You set the baseline at 20 skill checks per mile, which is ridiculously high. A skill check is only supposed to be required if there is pressure (time, hazards, enemies, etc) or if it is difficult enough to cause failure on its own. Most driving does not count as that; and requiring a skill check just to drive down the street (let alone 20 per mile!) is silly.
However, say you are driving along, driving along, kids in the back seat yelling "I gotta go to the bathroom!", you turn and scream "Not now, darnit!", and when you turn back to the road...a truck tire! Suddenly you need to make a snap decision. Do you swerve? Was there a car next to you? Do you have time to check? Do you try to brake? That is a skill check, and (hopefully) not something that happens even once per mile.
You show this again with your example of the rogue. The "Roll Til You Fail" method of DMing is the problem here, not having 1's be automatic failures. If you require 3 or 4 rolls for each 30 ft stretch of dungeon corridor, of course the players will fail.
What this really relates to is not 1 being an auto fail, but rather two things:
1) How often should you have your players roll?
2) How often should your players fail at an action?
The DM controls when players roll. So even if a 1 is a failure, it only happens if you, the DM, as for a roll. You can instead say to the player, "Your skilled rogue easily detects the three traps set by the kobolds. You can either bypass them or disarm them. If you want to reconfigure them it will require a check."
The second part is why I find the distaste of failure on a 1 so weird. I think a game where you have only a 5% chance of failure boring. As you said, why even play this game if you don't ever want to roll, if you never want risk? Every time a player makes a skill check, attack roll, or saving throw, there is going to be a decent chance of failure. That should be part of the fun. You get more of a thrill when you sometimes fail.
You then make a slippery slope argument: if there is a chance to fail, you have no idea what will happen, so you might as well act randomly. But does this apply even when you fail on numbers OTHER than a 1? What if a character with +27 to persuasion attempts to convince a king to give up their kingdom. Is it wrong for the DM to set the DC at 45?
The problem, as I said, occurs if the DM makes that same character (or any character) roll a persuasion check (or any skill) every 3 minutes just because they are talking to someone (or walking, etc). But in situations that call for it, there SHOULD be a chance to fail.
Take your example of the character with +13 to Persuasion who attempts to convince a king to send troops to aid the village. The player gives a rousing and moving speech. Why is the DM still asking for a roll? I would only ask for one to see how much EXTRA support they get. Or if there was some other factor in play (the king was already dealing with an uprising/invasion/lack of funds/etc, the grand vizier who hates the PCs already convinced the king not to send aid, or any other story based reason). These vacuum examples are not how the game is actually played. It isn't a video game. You don't just blindly roll because the rules say to roll and then your hands are tied. This covers the other side of the spectrum as well. If ANY player (not just a low charisma character) said they are farting loudly in the throne room, why is the DM even asking for a roll? Would you ask for a roll in battle if I said, "My character closes their eyes, stands on one foot, and with their left hand softly lobs a dagger in a random direction. Don't I get to roll because a 20 is an automatic hit?" My answer will always be "No".
Your final point is actually part of what makes the game great, and also runs counter to the entire argument against Natural 1 fails. If it is bad for players to try to come up with ways to auto succeed, why is it good for players to want to auto succeed on rolls? You can't have it both ways. I think it is better for players to have to be creative and think up solutions rather than just say, "I made a build with bonuses so high I can't fail." The first encourages roll playing, and investment into the game. The second just encourages numbers.
I agree with the above poster, you do not have to roll for every single thing, every single time.
Simple actions probably should not require a roll unless there is some unusual things going on. in this case you should probably just use a 10+skill mod to represent normal use or maybe 6 or 8 + skill mod.
Let me start by saying this. Driving is an active activity rather than a passive activity. Although you may sometimes feel like driving is "Effortless" or "Thoughtless" it is anything but that. We have simply learned through practice and experience. We have streamlined the process to such a degree that we no longer need to give pause for the decisions we make, but we are, nonetheless making those decisions. You can actually test this for yourself. Think back to when you first started to drive! Was everything effortless for you then? The next time you go for a drive, consider that you are actively regulating your speed, and your proximity to other vehicles, you are timing your stops, you are looking for gaps in traffic and swapping lanes. Does any of that sound like something one just automatically succeed on?
My example of 20 critical decisions per mile was actually quite an underestimation. Most places I looked at place the number way higher than 100 per mile driven.
You missed the point I was trying to make. In Dnd, uncertainty IS THE ENEMY. Every time we get a stat increase or a weapon with a magical enchantment or a feature that gives us an advantage, we get happy because we have now made the game less uncertain. The critical failure on nat 1, invalidates all of that. Invalidates the choices you made. Oh, so you have +14 on stealth because you took proficiency and expertise as your class features? well in Dnd 5e you roll a nat 1 on stealth and that's still a 15, a pretty decent roll. However, in One Dnd, that nat 1 means you fail, regardless of your choices, bonuses, spells, items, etc.
I think the main point which I think you also missed is that this is a punishment contingency built right into the system item. I want you to imagine the doorknob on your front door. What if I told you, that from now on, every time you touch your door knob, there is a 5% that it will burn your hand to the bone. There is also nothing you can do about it. It doesn't matter how fast you touch it or whether you are wearing some sort of protection, no matter what you do, there is this ever-present unmitigable 5% chance that you will be punished.
How would this knowledge affect your "Doorknob touching behavior"? Do you think you would be more likely, or less likely to want to touch said doorknob on a regular basis? You might start to want to circumvent having to touch the doorknob at all. Maybe you start entering the house through the window instead. You will start to engage in aberrant behaviors.
Dnd 5e allows for mitigation of uncertainty. Even if you roll a nat 1 to touch the doorknob, your bonuses might still save you from the worst possible outcome.
You make my point. A novice driver, like a novice character, will have a large chance to fail. Are you saying that even a 1st level character should always succeed? And a skilled, higher level character shouldn’t be rolling for mundane tasks, just as you say an experienced driver no longer has to actively think about every move. A decision point while driving doesn’t equate to a skill check. Because if you are rolling it has to mean there is a chance of failure. Otherwise you are just wasting everyone’s time. But even the best drivers can get into a collision.
Regarding your doorknob example, of course I wouldn’t touch it! That is an example of a high risk, low reward action. Now how about this…same scenario but behind the door is a pile of treasure. You better believe I am taking the chance.
Again: skill checks are for situations that call for it. A character wants to run across a room? No check. A character wants to parkour across a room and use the momentum to smash one orc into the other orcs? That is a check and the action might fail.
Finally, why are you asking the player to roll if they can’t fail?
It looks to me like, you are working from a conclusion rather than toward one. You have already decided what your stance on this issue is, and won't be convinced otherwise, regardless of how well or how much I explain my points. This is fair since I too have presented my own stance to the best of my ability.
Nevertheless, I shall try one more time. You ready? here goes.
Uncertainty is bad. Crit failure is punishing. It is invalidating your choices 5% of the time. The more your roll the more you get punished by the system, in a way that is outside of your control.
I disagree and think the rule is great. As the other posters mentioned, it is up to the DM to manage this appropriately. Attack rolls in combat already use this mechanic, and I've never heard anyone complain about it. I actually use it as a house rule for all my campaigns as well, as it lends an element of failure and challenge to the players that is much needed. If there is something that shouldn't have an auto-fail, then make an exception. But honestly, why are you even rolling for that check anyway? Most players get so many modifiers and advantage to their rolls that the auto-fail on a 1 helps balance the chances much better IMHO. I love the rule and glad they are including it in the new ruleset.
You get up in the morning.
You don't need to make a check to make every step) I normally only allow players to make a check with the same task only once between long rests. A check is not one attempt, it's the sum of all your attempts until you either succeed or give up. And realistically, a check is only due in a situation where there's true uncertainty, an element of blind luck, with both chances of success and failure.
If you do a lot of rolling in your game do you, roll before every placement of your foot before you move? If so then in your game you may want to have a two tier system in which some rolls have a crit and fail option and some do not.
But in general I like and think the present rule in 5e is great (for what it does) and should not be changed.
Also how a 1 or a 20 is dealt with in some skill checks could be important in games. I have seen some GM rule a 1 is like Kamchatmonk lists actions and results above and others just say you fail a check to put on your boots with a 1 then it takes a little longer. So success means putting your cloths and shoes on in 5 units on a nat 20 you do it in 3 units and on a nat 1 you do it in 9 units. Again it does not mean you fail you walking and chewing gum roll and you die or you fail your perception roll while fishing so a fish jumps down your open mouth and you die.
Kamchatmonk and spammdc, I think you guys are misunderstanding what I was trying to say. I am full in agreement with you, that rolls are not needed for everything. In fact, one of the first forum threads I made in Dnd beyond was about using passive abilities (passive perception, insight etc) as a basis for allowing players to automatically succeed and reduce unnecessary dice rolls.
The main idea is that regardless of what you have to roll for, the more frequently you have to roll the more you contact the negative outcome.
I believe that critical failure rules will result in even more unnecessary rolling. For instance, In my brother's game, I play a paladin. I have 20 in Constitution and at level 4 I took resilient constitution to gain proficiency in con saving throws. I have 18 in charisma and a cloak of protection, giving me a total bonus to my constitution saving throws of +10. Meaning that when I am concentrating on a spell, and I take damage lower than 22, even if I were to roll a natural one on my concentration saving throw I automatically succeed. Most of the time, I don't even have to roll for concentration. Now under One-dnd rules, I would have to roll every single time, because there is always the chance I might roll a 1 and auto fail.
How does this reflex on me and my choices in the game? I chose to allocate my racial points in Con, I chose to take a Con feat and gain proficiency in con saving throws precisely because I wanted my character to be able to hold concentration without issue
I see the problem is that on any roll the chance of failure is at least 5% and there is (usually) a known outcome if failure occurs, and (as written) their are onyly two results, success or failure, though sometimes there are exceptions so if you "fail" by at least 5 you get a worse failure (it is not clear yet how one D&D will treat nat 1s in such cases)
Going back to othe driving example, I have driven about 300,000 miles in my life, lets say for a 5 mile journey we say there is one check to see if I press the correct peddle every time. By rolling a d20 with a failure on a 1, that would leed to 60,000 rolls and about 3000 failures all of which resulin an accident. In real life I might have pressed the wrong pedal 50 times and on every one I was able to correct it before an acciden occurs. If a failure is counted as an error resultin in a accident even if you only have 1 check for a 10 mile journey everyone is going to crash every 200 miles. The solution here is to only allow rolls if the probability of failure is at least (say) 2.5%, If the chance of failure is between 2.5% and 7.5% the DC is such that they can only fail on a 1 (i.e. rounding the chance of failute t o5%), if the chance of failure is less than 2.5% there is no roll and the chance of failure is rounded to 0.
It also applies for successes, using the same approach any athlete , could break the world record with an average of 20 attempts.
I will often use very "scaled" ability checks particularly for knowledge bases checks. Say the party find a document that refers to Umberlee I might ask all the players to roll religion to see if the name means anything. On a 10 they might have heard of a sailor that referred to them in a way that implied they were a diety, on a 15 they might know she is generaly considdered evil going to to knowing pretty much everything about her on a roll of (say) 28. Currently the barb with a religion modifer of -1 who rolls a nat 20, knows exactly the same as the wizard (religion +8) who rolls an 11. In one D&D does the barb know absolutely everything about Umberlee? in which case should I not allow him to roll (so we can not fnd out if he has heard the name at all)
OneD&D also encourages players to demand more ability checks (at least those where the impact of failure is neglibible). In many groups PCs will be looking for birds in the air, insects on the ground, asking if they think the barmaid if lying when she says beer is 5 cp and so on until they get a nat 20 and inspiration but that is a whole other issue.
Thanks for your further explanation of your position:
A) Yes the more you roll the more chance you have for rolling a 1.
B) You do not need to roll for everything....thus some clear rules on when you would need to roll and when not to is important. If you want some guidance then maybe say if you take damage equal to or greater then 1/2 or 3/4 your concentration check you have to make a concentration roll. A simple rule and you can have a spot on the PC sheet for the number. Note: I would also provide guidance when other factors would be involved if the GM or group wanted, ie heat, cold, tired, climbing, running, etc.
C) Feat: I agree that feats can and should be reworked to include an idea that if you roll a 1 you may get a second roll before failure. I also have talked to people who say that is what inspiration is for or spells and magic items.
D) 5e is great for its simplicity but that also makes somethings harder because of its simplicity focus. I am also not saying that more complex games would necessarily solve your problem.
E) Unfortunately I am remembering many past discussions about this same topic that I have had and read down through the years, the most recient being a couple of years ago when the 1D&D playtest doc was released for the first time and during the Pathfinder II playtest.
F) Too many rolls can be a problem just as too few rolls can be a problem. And the game should look at this as well as your group.
G) I want my PC to do this but this happens: This is a tough area to deal with, player expiations vs how the game works and can be a GM and or game problem. For example I have played in a game where a player said "we just fought a dragon and now I am going to die by not making a climbing roll". The issue was in his mind having done something that was very heroic meant his PC did not have to deal with common things anymore.
I do think we are partly on the same page but I do not think simply saying you do not need to roll or take away crits and fumbles is a good idea....instead I think it should be handled by better feat and or class abilities. Note: this is the same point I made last time the 1D&D doc was released.
Let me provide another example: I create a PC who focus is dealing lots of damage, should they do max damage all of the time? I think most people would say no, some might say they should do max damage under specific situations and some may say yes (for various reasons). But in 5e as it is now that max damage just because its my PC focus, is a dramatic change to the game.
For your example I would look at your game and see if inspiration, feat expansion or class ability expansion would work the best. Just guessing I would either have a Warcaster II feat or change Warcaster as it is worded now to provide a less chance of fumbling a concentration check.
This is where things go wrong. Nowhere in the rules document does it talk about "needing to roll". People are reading "If you roll, a 1 is a failure and a 20 is a success" and translating it to "Because a 1 is a failure and 20 is a success, you HAVE to roll for every single action, no matter how mundane or absurdly impossible."
I also use scaled checks (and I would guess many, if not most, DMs do. The Playtest document says nothing to invalidate that. A 20 being a success does not mean the character will "know absolutely everything", it just means the roll was a success; they know at least something. You can still set threshold numbers; the only change is that the lowest threshold is always 20. (Side issue is the view that a low Int character can never know anything about anything. Why can't the barbarian have learned about Umberlee at one point? They are a bit slow, but they aren't incapable of learning and remembering. That is exactly why the d20 is a lot more impactful than the stat bonuses. At level 1, the best character is only +6 compared to the worst character, leaving many cases where the weakest will beat the highest.
You make a similar mistake in your interpretation of the Playtest. Players NEVER get to demand a roll. This has always been the case. Players describe what they want their character to do, and the DM decides how to resolve the action. This might involve a d20 Test. A player should never say, "I want to roll a d20 Test."
Even if players try to be sneaky and describe their actions as you wrote ("I look around for birds in the air", "I look for insects on the ground", "I watch the barmaid carefully...5 cp seems like a lot for a drink...I think she might be trying to rip me off."), none of these would require a check. "You see some crows." "You see some ants." "She rolls her eyes and points at a slab of slate behind the bar. Scrawled in chalk are the words 'Drinks...5cp'."
Players trying such shenanigans will quickly realize they are wasting everyone's time and move on.
It works in combat as a 1 is almost always a miss anyways outside of slimes and zombies until absurdly high levels you miss on a 1 just because you don't hit the AC. If at level 5 you would hit 25% of the enemies on a 1 I suspect people would look at it different.
Yup, I also do this in my games. Everything is scaled. You might fail a check but depending on how close you came to the DC, you might get different results. I like the idea of a gradient of outcomes from good to worst and everything in between.
However, this isn't very applicable to saving throws since they are usually a very binary thing. Either your pass or you fail.
Well, I highlighted a specific example in which you HAVE to roll every time ----> CONCENTRATION
Since I have a +10 bonus to my concentration saving throw (level 10 paladin) the lowest I can roll on a con save is 11, meaning that I automatically succeed in every concentration check originating from taking 22 damage or lower. However, in One-Dnd I have to roll every time because if I roll a 1 I will lose concentration no matter how high my bonus is.
I do not have a problem with the possibility of losing concentration but I do have an issue with always making concentration checks. IMHO it is a good balancing mechanism as well as other ways to roll a failed check again, inspiration, abilities, spells and magic items. Note ins general I have trouble with abilities that are essentially meta gaming, in your case how would someone who is not you know you failed your concentration check before it failed. One way it does work is if a person casts spells or uses abilities before it is needed, yes this does mean it might go unused.
Do you have an issue with a nat 1 fumble on attacks rolls?
Look at the bright side, if you receive 60 damage, you now have a chance to keep your concentration.
I think part of the problem here, as with so many issues in D&D is that we're not just dealing with the rule, but the culture and baggage surrounding the rule. The playtest packet does not have a "critical failure" rule, it has a rule that says that a roll of 1 always fails. Taking the rule itself, the only issue is that is raises the question of why anyone would be rolling if there's no chance of failure outside of the automatic failure rule. If the DM shouldn't be calling for rolls when there's no real chance of failure, why is the automatic failure rule needed in the first place? There are a handful of situations where a roll is always required (attack rolls, saving throws), so the rule could be narrowly intended to cover those situations. But there's also a long-standing tradition in D&D of DMs reflexively calling for rolls because they're stalling for time, they aren't sure what should happen, or they want to use the result as a narrative prompt. This is where the issue of induced failure that Stusano is highlighting can crop up, as the DM probably isn't stopping to gauge whether there truly is a reasonable chance of failure (this also works in reverse with the automatic success rules; the DM just wanted to gauge how far the PC gets, but they rolled a twenty and now they expect everything). This part could be addressed with a sidebar explaining what the auto success/fail rule is for and when it is meant to come up, but it's an uphill battle.
The other part of the problem is the influence of prior "critical" failure rules. RAW, rolling a 1 and automatically failing is no worse than a roll that narrowly misses the DC, but we expect it to be worse. Natural 20's mean the player gets to narrate something awesome, right? That must mean a Natural 1 should result in something awful, is how the logic goes. So where a normally failed driving roll might mean something like "you missed your turn," a Natural 1 becomes "you plow into a parked car at full speed." Your character gets put into clown shoes and looks utterly incompetent at something they're supposed to be an expert at, which is okay on very rare occasions, but a 5% chance isn't all that rare. And to reiterate, this isn't actually in the rules as written, any more that a "critical success" on a Natural 20 is in the rules, but there's a strong cultural push towards "20's are awesome and 1's are terrible," and if the 1's feel disproportionately bad (as they often do), then it will have the effect Stusano describes of discouraging players from rolling when they can get around it.