D&D 5E Randomness and D&D

[DND_Reborn]

Reading this, I would posit you're giving too much weight to skills and asking/expecting them to produce a steeper power curve than is really desirable. One of 5e's best traits is that it flattened the overall power curve from what 4e and - even more so - 3e had, back to more like what 0-1-2e ran with. You seem to want to fight this trend.

Well, I feel my position is putting the appropriate amount of value on skill compared to the die size of d20.

AD&D didn't really have "contested rolls" IIRC, but it has been a while... In many respects, it handled things very differently than the d20 system.

It's not a chess newbie vs a grandmaster, it's a chess-club regular vs the top player in the club. Taking a skill in something makes you better at it, sure, but doesn't and shouldn't make you perfect.

I never said a "newbie", I said someone with base proficiency and no INT mod, so a total of +2. That is someone who has some dedicated training and practices regularly. A "newbie" would be someone without proficiency who is just learning the game, etc.

No, taking a skill doesn't make you perfect, and it does make you better. What should help you approach perfect is two things: experience and dedication. Experience is modelled by level and dedication (perhaps?) by expertise... But only with feats can you have expertise unless you are a Bard or Rogue (another huge flaw in the game). Yes, most people play with feats, but they remain optional. And before anyone harps on 5.5 I don't care about it--it isn't out and I am not playing it.

Assuming equal ability scores, +2 proficiency vs. +6 has a 30% chance of winning. So, 17+ levels of experience doesn't do much compared to having proficiency alone. Sort of pathetic IMO. Ok, let's add expertise and bump it to +12. Now the +2 proficiency still has over an 11% of winning. The maximum amount of skill and dedication is going to lose more than 1 in 10 times to someone who has only basic proficiency. Very pathetic IMO.

So, let's look at the ability (skill) check system in 5E and chances of success:

I've highlighted a couple examples:

+5 bonus vs. DC 10 "easy" task. Should someone with either vast experience or training and natural talent granting +5 bonus really have a 20% chance to fail at an "easy" task??? IMO that should a resounding "NO!!!!" It isn't even a "medium" task, it is EASY for crying out loud.

+11 bonus vs. DC 15 "medium" task. Someone whose bonus represents the maximum ability 20 and 17th level (tier 4!) experience has a 15% chance to fail at a medium task??? Ridiculous.

Here are the numbers if we examined a 3d6 alternative to the d20:

Same bonuses and tasks. Now, the first example has only a 1.9% chance of failing. Much more to my liking for an easy task by someone with experience and/or training and talent. In the second example, such an experience and skilled person would only fail about 1 in 200 times while attempting what is only medium difficulty. Failure in such a case would be a total fluke!

Such a system would support why the Variant for Automatic Success is offered in the DMG. Yet I would rather have a system that makes it happen without having to enforce additional rules about proficiency, bonus, and/or level vs. DC.

That, and one of the pure joys of D&D for me is the chance of the upset result. Take that away and it gets boring pretty fast.

Great! I am honestly glad that works for you, but it really doesn't sit well with me. I find it ridiculous when it happens.

Which is deep into "why bother?" territory.

Because it can happen. You might as well ask the millions and millions of people world-wide why they bother playing the lottery? I never have, but I know a lot of people who do regularly....

 

[GMforPowergamers]

Something Reynard said got me to thinking. When I first started playing D&D, random chance was king. What ability scores did you roll? Did you get a 17 instead of a 16? These things could make a world of difference when determining what you could play, and how effective you would be.

I remember those days, but I also remember the animosity that sometimes cropped up when a teen would roll really well and another teen rolled really bad... something that didn't start to fix until we were in our 20s... farther in then I want to admit.

If your DM rolled treasure randomly, a humble treasure chest could pay out in potions of healing or a longsword +3; the original "loot boxes", if you will.

man... I had a 2e DM that let you come into a campaign and get rolls for items... I learned the apparatus of kawalish is a reroll...

Even some magic items and spells were slot machines; what did I get from my Bag of Beans? What card did I draw from the Deck of Many Things? Did my Prismatic Spray nuke the dungeon boss, or was it a colorful dud? Even the basic game mechanics, did I hit? How much damage did I do? Did I make my save?

the deck of many campaign enders is the only one that bugged me...

Often, it proved that it was better to be lucky than good.

very often for 2e and 3e...

Over time though, D&D campaigns evolved to have ongoing plotlines and became much more than the Rogue-like and Diablo-esque games that it inspired. Some groups began to shun randomness, because it could turn an enjoyable story into a total disaster. Not only did some players reject pure randomness, but even some games did (I'll shill Amber Diceless Roleplaying here as an example).

this I did NOT experence... selling me on 2e was drama plot, a ongoing story and improve in fact that sold me on rifts and WoD too.

5e, from the very beginning, however, catered to a "less random" approach to the game.

100%

Monsters were presented with average damage totals to speed play. Players could opt to not roll hit dice, instead taking a set amount of hit points on level up. Even rolling for ability scores has changed; while still the first option presented, most groups seem to have switched to point buy, and, of course, ever since the year 2000, the difference between a 16 and a 17 has become largely academic. You no longer need certain ability scores to qualify for a class (even though you are still rewarded in other ways for higher ability scores), with the notable exception of multiclassing.

my group went to point buy or arrays (not often the standard)

Or do you feel that the game has become too predictable, and want even more chaos, like exploding dice or more d% tables to roll on?

How do you feel WotC will move the game in the future, towards one extreme or another?

I bet we will ALWAYS have SOME randomness... but I agree we will see less and less.

 

[GMforPowergamers]

If randomness bothers you, how about 5d4? You only auto fail if you roll all 1's, you only critically succeed if you roll all fours. The actual "could theoretically succeed" range is now 5-19, or 15 numbers instead of the current 18. You have an average result of 12.5 which should, given a reasonably good bonus, succeed at most tasks. Extreme results still exist, but are now incredibly rare moments, so they can be given more weight when they do occur, from the catastrophic to the miraculous.

I don't think I would do it, but that REALLY sounds interesting.

 

[GMforPowergamers]

I'd neither seen nor heard of the concept before 3e.

Then again, back around 1984 we added "body points", which gives everyone an extra 2 to 5 h.p. once at 1st level - mostly "meat" and harder to rest back or cure - which may have to some extent acxcomplished the same thing.

I know we saw an option in 2e... I think it was the reprints or the domain of dread book.

Having said that we had a big fight when someone had a really low con rolled a 1 for hp and was told they 'died' at character creation...

 

[James Gasik]

I know we saw an option in 2e... I think it was the reprints or the domain of dread book.

Having said that we had a big fight when someone had a really low con rolled a 1 for hp and was told they 'died' at character creation...

And here I thought you could only do that in Traveller or Deadlands!

 

[Lanefan]

No, taking a skill doesn't make you perfect, and it does make you better. What should help you approach perfect is ...

...nothing. IMO there should always be a chance of failure if failure is possible, even as there should always be a chance of success if success is possible. That the d20 isn't always granular enough is another issue entirely....

two things: experience and dedication. Experience is modelled by level and dedication (perhaps?) by expertise... But only with feats can you have expertise unless you are a Bard or Rogue (another huge flaw in the game). Yes, most people play with feats, but they remain optional. And before anyone harps on 5.5 I don't care about it--it isn't out and I am not playing it.

Assuming equal ability scores, +2 proficiency vs. +6 has a 30% chance of winning. So, 17+ levels of experience doesn't do much compared to having proficiency alone. Sort of pathetic IMO. Ok, let's add expertise and bump it to +12. Now the +2 proficiency still has over an 11% of winning. The maximum amount of skill and dedication is going to lose more than 1 in 10 times to someone who has only basic proficiency. Very pathetic IMO.

In most instances I'd see that as a feature rather than a bug - it's a nice flat power curve where the great can still fail and the lowly can still succeed.

So, let's look at the ability (skill) check system in 5E and chances of success:

View attachment 268617

I've highlighted a couple examples:

+5 bonus vs. DC 10 "easy" task. Should someone with either vast experience or training and natural talent granting +5 bonus really have a 20% chance to fail at an "easy" task??? IMO that should a resounding "NO!!!!" It isn't even a "medium" task, it is EASY for crying out loud.

+11 bonus vs. DC 15 "medium" task. Someone whose bonus represents the maximum ability 20 and 17th level (tier 4!) experience has a 15% chance to fail at a medium task??? Ridiculous.

Well, not ridiculous if one looks a bit sideways at their word choice.

"Easy" should be the type of thing that doesn't need a roll at all unless you're truly hopeless, thus should replace "Very Easy" above as the DC 5 descriptor. "Medium" should drop to the DC 10 point, with something like "Tricky" as the DC 15 descriptor and the rest left as is.

Put another way, what they're describing as "easy" often isn't. Further, unless the intent is that the game become something of a supers affair at very high level, there always has to be a chance for even high-level characters to fail.

Here are the numbers if we examined a 3d6 alternative to the d20:

View attachment 268618

Same bonuses and tasks. Now, the first example has only a 1.9% chance of failing. Much more to my liking for an easy task by someone with experience and/or training and talent. In the second example, such an experience and skilled person would only fail about 1 in 200 times while attempting what is only medium difficulty. Failure in such a case would be a total fluke!

The way I'd have it, and do have it for my 1e Thieving skills, is that the chance of success would never fully reach 100% - a roll of '00' would always have something funky happen even if your chance of success was a base 113%, which can happen.

I'd also want to smooth out the other end a bit, and turn some of those flat zeroes into a small but not-zero number. In fact, I'd want to smooth out the whole thing because when using a bell-curve (unless you assign uneven DCs e.g. DC 13 or DC 8) there's just too big a jump between one step and the next in the step that crosses 50% in that 3d6 table. Percentile dice for the win!

 

[Lanefan]

I know we saw an option in 2e... I think it was the reprints or the domain of dread book.

Doesn't surprise me - 2e had all kinds of funky things going on late in its run, most of which I happily ignored.

Having said that we had a big fight when someone had a really low con rolled a 1 for hp and was told they 'died' at character creation...

I'd have got that player to roll up a second character then and there to have ready as a backup, but still bring the first one into play just on the off-chance it could somehow manage to hang on.

Way long ago before we introduced body points, a character came in to one of our games with 1 h.p. She lasted long enough for b.p. to be introduced a year or so later, and min'ned out on those too.

Almost 40 years later (though there's some big gaps in there), that character is still alive, quasi-active, and in our Hall of Fame; I think she's 9th/7th level these days as a double-class.

So it can happen.

 

[overgeeked]

Doesn't surprise me - 2e had all kinds of funky things going on late in its run, most of which I happily ignored.

I'd have got that player to roll up a second character then and there to have ready as a backup, but still bring the first one into play just on the off-chance it could somehow manage to hang on.

Way long ago before we introduced body points, a character came in to one of our games with 1 h.p. She lasted long enough for b.p. to be introduced a year or so later, and min'ned out on those too.

Almost 40 years later (though there's some big gaps in there), that character is still alive, quasi-active, and in our Hall of Fame; I think she's 9th/7th level these days as a double-class.

So it can happen.

That’s one of the things that bugs me about conversations about rolled HP. Rolling low isn’t a death sentence and doesn’t mean the character’s unplayable. You don’t know until you try.

 

[Lanefan]

That’s one of the things that bugs me about conversations about rolled HP. Rolling low isn’t a death sentence and doesn’t mean the character’s unplayable. You don’t know until you try.

In fairness, the odds of survival for such a character are immensely higher if the death at -10 option is in play rather than death at 0.

 

[DND_Reborn]

...nothing. IMO there should always be a chance of failure if failure is possible, even as there should always be a chance of success if success is possible. That the d20 isn't always granular enough is another issue entirely....

Notice I said "approach" perfection, not reach it. But depending on the task, the degree of failure should simply be the minimal roll. In the case of 1 in 20, 5% is ridiculous IMO for something that is routine and very easy. At least with 3d6, the minimal roll is just about 0.5%, which is better from my point of view.

In most instances I'd see that as a feature rather than a bug - it's a nice flat power curve where the great can still fail and the lowly can still succeed.

As I said, you like flat-power curves than I'm glad d20 is the roll for you. I don't, obviously.

Well, not ridiculous if one looks a bit sideways at their word choice.

"Easy" should be the type of thing that doesn't need a roll at all unless you're truly hopeless, thus should replace "Very Easy" above as the DC 5 descriptor. "Medium" should drop to the DC 10 point, with something like "Tricky" as the DC 15 descriptor and the rest left as is.

Put another way, what they're describing as "easy" often isn't. Further, unless the intent is that the game become something of a supers affair at very high level, there always has to be a chance for even high-level characters to fail.

Oh, I agree, their word choice is silly. If something is "very easy" why should it even be in doubt unless there are other circumstances.

I would drop them all by 5, but all of them. You can keep "very easy" at 0, so failure is only possible if you have a low ability score or some other penalty. Then, a PC with +5 (standard for your "good thing" at level 1 even) would succeed on a 20. But a 5% chance for "nearly impossible" is too high for me. Using 3d6 would still be impossible unless you allowed a nat 18 to work, but you could also use 2d10, so "nearly impossible" would be 1% for that PC.

The way I'd have it, and do have it for my 1e Thieving skills, is that the chance of success would never fully reach 100% - a roll of '00' would always have something funky happen even if your chance of success was a base 113%, which can happen.

I'd also want to smooth out the other end a bit, and turn some of those flat zeroes into a small but not-zero number. In fact, I'd want to smooth out the whole thing because when using a bell-curve (unless you assign uneven DCs e.g. DC 13 or DC 8) there's just too big a jump between one step and the next in the step that crosses 50% in that 3d6 table. Percentile dice for the win!

Well, you certainly can, as I said, make a minimal roll always fail and a maximum always succeed. But the 5% of the d20 is just too high for me, so 3d6 (or even 2d10) is better and my personal preference.

Another option IIRC someone mentioned upthread which I have tried in the past but also really liked is 3d20 take middle. This makes a minimum or maximum roll just 1 in 138 (roughly), very rare certainly but quite possible! It also continues to use d20's and makes it so you don't have to do any math.

Using it produces these numbers:

And of course, you can always replace those 0's and 100's with 0.725% and 99.275% respectively.