D&D 5E Using 3d6 for skill checks

[pdzoch]

To maintain the 5% crit rate, you'd want to have a crit happen on 16+ on 3d6.

I've done 3d6 for skill checks before and found it to be TOO average. If you look at the curve, it's very bulbous... meaning that the middle average results appear much more often than the ends, and that results in much less variance from the roll and much more impact from the modifier bonuses. Didn't end up working for me.

That being said... I *am* going to be using 2d10 for ability checks when I run Curse of Strahd, as the curve is much less steep. We're still going to see more rolls in the 10-12 range obviously... but we're also going to see 20% of all rolls still fall at either end (2-5 and 17-20). I'm very curious to see how it works out.

And in that regard, I also adjusted my DC chart to reflect the 2d20 roll scheme as well:

Trivial: DC 5
Easy: DC 8
Medium: DC 11
Difficult: DC 14
Hard: DC 17
Severe: DC 20

As DEFCON 1 demonstrates, changing the dice type and number for the skill checks requires an adjustment to the difficulty scale. This is too large a re-engineering of the rules for my tastes.

On the 3d6 scale, if a 16 is a critical, then those skills with a DC of 15 only have one number that equals success, because the other three numbers equate to critical, which seems excessive to me, unless you redesign the DC levels of the checks.

I think an easier technique would be to use the advantage feature already offered in the 5e rules. Role two d20s and take the higher. For easy or trivial tasks without any distraction, it would seem suitable. Or better yet, simply allow the characters to auto succeed on those tasks.

I try to limit skill checks to those events that advance the story or highlight a character's capabilities.

 

[jayoungr]

If you like your skill checks on a predictable bell curve, do it.

I don't like my skill checks on a predictable bell curve, because I like it when the dice send the story in unexpected and interesting directions.

 

[DEFCON 1]

If you like your skill checks on a predictable bell curve, do it.

I don't like my skill checks on a predictable bell curve, because I like it when the dice send the story in unexpected and interesting directions.

I got news for you... rolling a d20 for skill checks gives you just as predictable a curve. It's a completely flat curve granted... but it's still predictable.

 

[jayoungr]

I got news for you... rolling a d20 for skill checks gives you just as predictable a curve. It's a completely flat curve granted... but it's still predictable.

Okay, okay, but it's not a bell curve.

 

[Ovinomancer]

I'm actually not a fan, at all, of 3d6 in D&D. It works in Hero because the math is already set up that way. The dials turn a bit slower, such that a +1 can be a big deal for an experienced character, but it's about the same cost to go from marginally trained (8 DC) to well trained (11 DC).

The D&D math is set up to work better with linear progression. It's not really that uncommon for, say, a Fighter to have a +3 strength bonus, a +3 proficiency bonus, and a +1 magic bonus. In the standard D&D math, that's a swing of 35% chance of success. If you happen to hit the "sweet spot" on the 3d6 bell curve -- which is reasonable in a balanced adventure -- that bonus nearly doubles (+66%, back of napkin). The net result of 3d6 in D&D isn't more realism, smoother progression, or anything like that. It's increased swingyness, which almost always works against the PCs.

So, I'm not opposed to 3d6 as a resolution mechanic -- I love Hero. I just don't think it fits well with D&D (or "d20") and would be extremely hesitant to play in a game that used it. If I did play, it'd almost certainly be expressly to watch the math.

I wonder, if each player were given the choice between using d20 and 3d6, how many would choose each? If given the choice for each roll, I wonder how it'd evolve.

I'd actually say it's less swingy because it is more likely that character with bonuses succeeds. It levels out the game on the side of successes. That might not be to your taste, but the math facts of 3d6 vs d20 say that 3d6 will have less variance and so be, by definition, less swingy.

 

[Ovinomancer]

I got news for you... rolling a d20 for skill checks gives you just as predictable a curve. It's a completely flat curve granted... but it's still predictable.

d20 is actually less predictable, as it has a larger variance and SD than 3d6. True, I can tell you the statistical likelihood of rolling any number on both, but the relative chances make 3d6 a bit more predictable in the long run than d20. Statistically speaking. Of course, any given roll is unpredictable.

 

[Flamestrike]

When I was playing the RAW game, I found myself choosing not to attempt things my character wasn’t good at, e.g. kicking down a door with 8 Strength, even though my character had a reasonable chance of success and the fighter had tanked a few rolls, because I didn’t want to steal his spotlight by encroaching on his schtick.

Intresting. As a DM I wouldnt have let you roll anyways.

'If the St 18 fighter cant do it, you certainly cant.'

It annoys me when DMs make every PC attempt a survival check to navigate, or knowledge check and then going with the highest roll. If the PCs are navigating as a group, or recalling bits of lore then it's a group check at best.

Personally, when muliple players roll for something like knowledge, I sometimes take the result of the PC with the highest persusasion bonus, seeing as he's the most likely to convince the others that his directions/ information is the best.

TLR: The player that declares the attempt, rolls the dice.

I also loathe players asking around to see who has the highest bonus to a check, before that person rolls, but thats a topic for a different thread.

 

[DEFCON 1]

Okay, okay, but it's not a bell curve.

It's just a bell that's been placed under a hydraulic press.

 

[Amrûnril]

Interesting house rule. One thing I'd note is that this dramatically reduces the impact of advantage and disadvantage. If particularly high and low results are substantally lower than moderate ones, the impact of rolling twice is likely to be a lot smaller.

 

[happyhermit]

This is one of the things I was thinking about asking this forum about. On one hand, I love the simplicity of simply rolling that d20 to resolve everything, adv/dis is so easy, and new people seem to love that d20. It is exciting to roll, especially as we play with crit success/failure usually.

On the other hand, it is very swingy and although I don't think it hurts the game at all most of the time (success or failure can be just as interesting) I have looked at options and ended up messing around with 2d10.

The 2d10 seems to be much easier to fit into 5e than 3d6, it gives almost the same range of numbers (2-20), as opposed to (3-18) and while it does give a bell curve, it isn't nearly as extreme as 3d6. I found that 2d10 didn't really require any substantial rule changes to function adequately. Snake eyes works good for crit fail, and 20 works great as crit success, the chances are much smaller but they still give that 1in100 chance. Increased crit range can be kept the same numbers, albeit less likely. Obviously if used in combat this will negatively effect multi-attack and increased crit types, though not all players are concerned about that sort of thing.

However, in the end, we still end up going back to the d20

Anyways, always nice to hear other people's views on the subject.