# Has anyone tried ditching the d20?

### Pathfinder First Edition General Discussion

was going through different books looking for stuff for a campaign I am working on and came across an interesting rule varient in the 3.5 Unearthed Arcana

replace the d20 with 3d6.

I was curious if anyone had tried it and how it worked out?

I haven't tried it, but the first thing that sticks out with that is that some results that are possible with 1d20 are impossible with 3d6. 4d6-3 would be closer, giving you a possible range of 1-21.

Pathfinder Lost Omens Subscriber

so, mathematically all it does it limit the range to 3-18 and make 10 much more likely than extreme rolls(and thus making crits very rare). this would make rolls more stable and less prone to chance.

There's some nice analysis of the math over here.

TL;DR: It 'flattens' the curve of d20 rolls towards the average in a way that makes easy opponents much easier, and difficult challenges vastly more difficult. It narrows a GM's options in terms of what sort of challenges can be thrown at players as a result, but it also makes everything much more predictable, which can allow for better tuning.

JoeJ, that has an about 0,5 higher average.

You could do 4d6 with the 6 turned into 0s (0-20) to remove the math.

with 3d6 you need to find some way to determine crits so it stays fair. For the purpose of crits, 6 could be considered 7 maybe?

Pathfinder Lost Omens Subscriber

people are like

"how determine crit?"

I think 3 6s is sufficiently rare.

3 6s is 1 in 216.

Compared to 1 in 20.

It's waaaay too rare.

You could use the rule that gurps did. 17 or 18 is a crit.

That's 4/216 or 1/54. Still much rarer but not so rare. If you really wanted you could try making a 16 a crit as well.

It would make investing in an improved AC more attractive.

Excuse me while I just uh... Drop this here.

Bell Curve Rolls

Edit: For reference for everyone else, since you seem to already have UA.

Doesn't the Dragon Age system use 3d6?

I like the idea, but I would feel weird not rolling a d20.

Well... it's good to challenge assumptions in a while. It's good to be able to say "we use a d20 because these are the properties of the d20..." rather than "This is how it's always been done. 3d6 is for communists."

That said, I'm not sure I like 3d6 better. A bit of swinginess in rolls isn't bad actually. If you have to roll, it could go either way; if things are "controlled circumstances" there's Take 10.

I think this is related to the whole Fumble Deck issue. Many of us subconsciously do like the idea of swinginess. Fumbles aren't a good implementation, but I see them as a sign that we want something like that kind of uncertainty. I think the 1/20 autofail/autohit and it's relatively high frequency (10% altogether) is actually a good implementation.

I've actually been considering using 3d6 in conjunction with the Variable Modifiers rule....

But, since I'm horrible with math (for now), I'm not sure how much of an impact that would have in-game. Particularly how much modifying of enemy defenses I'd have to do. Seems fun and at face value, seems to add that 'swinginess' people love.

Artemis Moonstar wrote:

I've actually been considering using 3d6 in conjunction with the Variable Modifiers rule....

But, since I'm horrible with math (for now), I'm not sure how much of an impact that would have in-game. Particularly how much modifying of enemy defenses I'd have to do. Seems fun and at face value, seems to add that 'swinginess' people love.

Actually, as pointed out, it gets rid of swinginess. It's much harder to roll high on 3 die than 1. So in battles that you're overwhelmed, its much much harder to turn the tides and such.

 3 people marked this as a favorite.

One of my groups used a rule where you rolled 3d6 instead of taking 10. It meant that you didn't jump straight from '50% chance of failure' to 'never fail', and that you didn't lose the fun of rolling dice just because you wanted a better chance to succeed. Also, it meant that even if you did fail a roll, you would almost never fail by 5 or more, which matters a lot for things like climbing and balancing.

Lakesidefantasy wrote:
Doesn't the Dragon Age system use 3d6?

It does indeed.

A rather different feeling, considering only the basics of attacking and dealing damage - low level monsters are really low but keep their danger level for a while longer, and boss type monsters are deathly scary pretty much forever.

It's not readily comparable to a d20-switched-to-3d6 system due to its inner variants for the magic system, class level abilities, and the stunt points system.

Thomas Long 175 wrote:
Artemis Moonstar wrote:

I've actually been considering using 3d6 in conjunction with the Variable Modifiers rule....

But, since I'm horrible with math (for now), I'm not sure how much of an impact that would have in-game. Particularly how much modifying of enemy defenses I'd have to do. Seems fun and at face value, seems to add that 'swinginess' people love.

Actually, as pointed out, it gets rid of swinginess. It's much harder to roll high on 3 die than 1. So in battles that you're overwhelmed, its much much harder to turn the tides and such.

I'm already aware 3d6 gets rid of the swinginess... Or were you referring to the Variable Modifiers, which I was wondering if it added a bit more swinginess to the 3d6.

'Cause your post seems to refer only to the 3d6 part of my post.

If you are going to use it on the forum, you can use 3d7 for 3-21. With that system (either 3d6 or 3d7) you can dare to give critical and fumbles a try, because they will become very rare.

keen rapier: 3d7 ⇒ (6, 3, 3) = 12
keen rapier: 3d7 ⇒ (7, 3, 5) = 15
keen rapier: 3d7 ⇒ (2, 4, 1) = 7
keen rapier: 3d7 ⇒ (1, 6, 7) = 14
keen rapier: 3d7 ⇒ (4, 5, 6) = 15
keen rapier: 3d7 ⇒ (7, 7, 5) = 19

2d10 is also something people have tried in the past. Of course a 2 would result in an automatic failure in this case. 2d10 would also be more "swingey" than 3d6.

bodhranist wrote:
One of my groups used a rule where you rolled 3d6 instead of taking 10. It meant that you didn't jump straight from '50% chance of failure' to 'never fail', and that you didn't lose the fun of rolling dice just because you wanted a better chance to succeed. Also, it meant that even if you did fail a roll, you would almost never fail by 5 or more, which matters a lot for things like climbing and balancing.

Brilliant! It does just what you say - for those who like to roll dice - it keeps the game fun and preserves the verisimilitude on the RP side.

In Dragon Age (another 3d6 system), instead of crits you have "stunts," which let you do cool things whenever at least two dice come up the same, provided of course that you hit. This happens much less rarely than three sixes and so it works much better.

I actually came into the thread to say "base crits on rolling the same number on dies" so I'm fine with that. Quick head math says it'd be around 10% frequency for the threat. That's about where most popular weapons are.

If I recall correctly there are 216 possible combinations of a 3d6 roll; 6 of which are all the same number on the three dice, so that would be 6/216*100% or about 3%.

If criticals were to happen when merely two of the three dice come up the same number then I think the percentage increases to about 17%.

Would feel like GURPs Fantasy.

You could go crazy and try to go completely diceless (like Amber)

KestrelZ wrote:

Would feel like GURPs Fantasy.

You could go crazy and try to go completely diceless (like Amber)

Or Nobilis.

Why not just 1d6*3?

Unearthed Arcana (where I found the rule) did cover adjusting the Crit ranges. they suggest chenging them to

20 = 16-18
19-20 = 15-18
18-20 = 14-18

They also recomended that the CR for Large groups of monsters be reduced since the Bell Curve favors the PCs more than the monsters. Basically if there are 4 or more monsters reduce their CR by 1

(and yeah I really enjoyed Amber...takes a while to wrap your head around a diceless system but once you do it's fun)