D&D 5E Common sense isn't so common and the need for tolerance

[Lanefan]

That's a possibility. Or it could have 'easy' checks that are actually 'easy' rather than 'the best possible character at this check still fails 10% of the time with no confounding factors'.

Succeeding 90% of the time at something where success is not certain sounds mostly like 'easy' from here.

Keep in mind that in theory neither the characters nor the players should have any idea what their odds are of success, particularly in a role-play interaction. They've no way of knowing whether the situation is easy for them, or hard, or nigh-impossible...which is also why such rolls really ought to be made in secret by the DM.

Right, but that just ends up with characters walking into a room and failing what is described as an easy task because the DM decided a roll was necessary because of 'consequences' and assigned an 'easy-but-actually-fairly-risky-if-you-think-about-it' DC, then rolled against it and failed. You've just hidden the final numbers, not the result.

Yeah, that's on the DM - maybe not for forcing the roll (there's nothing wrong with forcing a roll even when one isn't needed, just to throw the red herrings in) but for making the consequences of failure too severe. Or, perhaps, for not giving one or more subsequent rolls or opportunities to mitigate the failure.

Right. But even previous editions of D&D did a better job at making DCs conform to expectations, through the devious mechanic of simply making a starting trained character have a high enough modifier to automatically succeed at an easy DC.

Except that wherever possible success - or failure, for that matter - should not be automatic if an alternative exists. And even where the DM knows the success/fail is a foregone conclusion she should still roll anyway, just to keep player knowledge and character knowledge in synch (the characters don't know success/fail is automatic thus neither should the players).

Lanefan

 

[Celebrim]

Succeeding 90% of the time at something where success is not certain sounds mostly like 'easy' from here.

Definition of "easy" is very important to an RPG. There are numerous things were there exists a non-zero rate of failure, yet a 10% failure rate would be rightly perceived as ridiculous. An example might be taking a sip from a glass without spilling water, or walking quickly from place to place without stumbling or falling. There are numerous random small accidents of life that occur in actions we do hundreds of times without failing. And even among things that are more difficult, without stressors that make them more difficult, there are plenty of people who can do things 99.9% of the time without failure that untrained people could not perform even half the time.

Keep in mind that in theory neither the characters nor the players should have any idea what their odds are of success, particularly in a role-play interaction. They've no way of knowing whether the situation is easy for them, or hard, or nigh-impossible...which is also why such rolls really ought to be made in secret by the DM.

This directly relates to the truth that rulings are simply rules that the player doesn't know before they experience them. There is probably a good discussion to be had about how much knowledge a player ought to be given about their chances of success before or when they make propositions, and equally importantly how a DM ought to narrate the outcome to convey the odds of success to the player in a clear way so that in the future the player will have some indication of how likely they are to succeed in similar situations or on retries.

For example, if a player purposes to jump a chasm, how clearly should the GM indicate to the player the likelihood of the character's success before he jumps and why, and likewise should the GM indicate how close the character came to succeeding on the basis of the actual likelihood of success or on what the GM believes makes for a dramatic or cool story?

I think based on what you've written that I'm much more of a champion of automatic success than you are. In fact, I'd go so far as to say a skill system is completely broken if there does not exist the possibility of automatic success (or at least 99.999%) once a certain level of skill is reached.

 

[Tony Vargas]

Definition of "easy" is very important to an RPG. There are numerous things were there exists a non-zero rate of failure, yet a 10% failure rate would be rightly perceived as ridiculous.

Yep, d20 doesn't have the resolution - nor players and DMs the patience - to model something like that in a robust way. Mostly those things will be trivial, the DM will narrate them happening routinely...
...when he does call for a check on something like that the 5 or 10% chance of failing that instance can, I suppose, be put down to a gamblers-fallacy 'building up' of the odds against you from all those previous successes.

Or, the DM could go all the way an narrate success 99 or however many times, then pick a fun time to narrate an unexpected failure....

...with or without the cover of placebo dice rolling behind the screen.

 

[Yardiff]

There are automatic successes in 5e, its called 'automatic success', anything else requires a roll.

 

[Satyrn]

A rogue or bard with expertise is still sitting at a 10% failure chance. Which is better, to be sure, but bear in mind we're now talking about only two character classes, in their focussed skills with the best available stats in place, still failing at an 'easy' task.

Thank you.

This has triggered a thought that might help in my adjudication.

I was calling DC 10 checks "easy," but you're right that it really doesn't mesh with expectations of what easy is. So, I'm gonna start calling DC 10 "Tricky" - to describe that it truly has a chance of failure but an experienced, skilled person ought to succeed most of the time.

I was calling DC 15 hard, but I think I'll go with "Challenging," and DC 20 is now gonna be "Improbable"

 

[Celebrim]

Yep, d20 doesn't have the resolution - nor players and DMs the patience - to model something like that in a robust way. Mostly those things will be trivial, the DM will narrate them happening routinely...

Agreed, but my point is that D20 has tended to err on the side of "If there is a very small chance of failure, there should be a chance of failure." rather than "If there is a very small chance of failure, then there should be no chance of failure." Back when D&D was less hidebound to a single resolution mechanic, you had D% chances and so forth to model very low chances of failure (or success) when needed, but if all you have is a hammer then everything starts looking like a nail.

Some published rules systems do give a 5% chance of failure to everything.

"Automatic success" or hand waving a roll misses the point though. The game may realize its limitations enough to define certain trivial things as an automatic success, but typically a game will not make that automatic depend on the skill of the character so that - in so far as the actual rules are concerned - walking across the floor is just as likely to succeed whether you are an arthritic octogenarian or an Olympic gymnast. The level of 'automatic success' available to you doesn't increase as you level up in skill.

 

[Celebrim]

I was calling DC 10 checks "easy," but you're right that it really doesn't mesh with expectations of what easy is

I haven't played 5e, but in 3e I frequently have skill checks with DC of less than 10, and even DC 5 isn't uncommon. DC 0 is "easy" - people of average skill always succeed. DC 5 is "routine" in that people of average skill expect to usually succeed, and actually experienced people almost always do so.

A good example of why that matter is charging across a typical uneven cavern floor. A DC 5 or even DC 3 "balance" check to run or charge in such an environment sounds trivial - in 3e only a failure of 5 or more would mean you actually trip and fall over - but if you are trying to do this heavily burdened then suddenly even a "routine" check becomes an interesting constraint on your movement.

 

[Satyrn]

I haven't played 5e, but in 3e I frequently have skill checks with DC of less than 10, and even DC 5 isn't uncommon. DC 0 is "easy" - people of average skill always succeed. DC 5 is "routine" in that people of average skill expect to usually succeed, and actually experienced people almost always do so.

A good example of why that matter is charging across a typical uneven cavern floor. A DC 5 or even DC 3 "balance" check to run or charge in such an environment sounds trivial - in 3e only a failure of 5 or more would mean you actually trip and fall over - but if you are trying to do this heavily burdened then suddenly even a "routine" check becomes an interesting constraint on your movement.

Aye.

It's funny, I've rather forgotten what 5e specifically says about skill checks and DCs (and hiding and yadda yadda) because I discovered a much more comfortable way to adjudicate it all ad hoc. It of course requires earning the trust of my players. I simply call for a check based on the player's action, then set the DC at either Easy (10), Hard (15), or near impossible (20). If I would set the DC lower than 10, that's auto success, if I'd set it higher than 20 it fails.

Success, of course, means the player did what he set out to do.

I've been calling the DCs easy or hard just because one is easier than the other. I'm thinking by calling it "tricky" instead, I'll reduce the number of trivial rolls I call for, which I'm hoping will encourage the players to do even more crazy things.

 

[Tony Vargas]

Agreed, but my point is that D20 has tended to err on the side of "If there is a very small chance of failure, there should be a chance of failure." rather than "If there is a very small chance of failure, then there should be no chance of failure."

In 5e, that's left entirely up to the DM's judgement. Narrate success/failure, or call for a check. If you don't want there to be a failure, don't call for a check. A little pat, I suppose...

I haven't played 5e, but in 3e I ...

You should give it a try sometime! It is a lot like 3e in a few ways, and sheds many of that edition's worst problems (well, or at least gives the DM more latitude to deal with them). I find it more fun & rewarding to run than to play, but you might like either or both...

 

[Lanefan]

Agreed, but my point is that D20 has tended to err on the side of "If there is a very small chance of failure, there should be a chance of failure." rather than "If there is a very small chance of failure, then there should be no chance of failure." Back when D&D was less hidebound to a single resolution mechanic, you had D% chances and so forth to model very low chances of failure (or success) when needed, but if all you have is a hammer then everything starts looking like a nail.

This is a very good point - sometimes d20 just isn't granular enough.

I find this hammer-nail approach annoying - a (group of) designer(s) will come up with a fine new game mechanic that works really well for some things, and then try to shoehorn much of the rest of the game into it whether it is the best game mechanic for the job or not. In 3e (and since) it was the d20. In 5e it's advantage/disadvantage.

Lanefan