D&D 5E Surprisingly, nothing breaks when switching D&D to 2d10 instead of d20

[loverdrive]

Since I've started using fixed DC (10- fail, 11-17 partial success, 18+ full success), I was playing around with various ways to shift the numbers to generate more partial successes on average, while still allowing +0 characters to achieve full success. The most radical solution I had in my list was to switch to 2d10, which I recently tested, just as a joke.

But, to my surprise, it worked. Like, worked very well. It breaks nothing, while seriously reducing σ. The only bug I've encountered is that automatic misses just never happen, but I'm honestly fine with that.


I guess it's not something I'll be more than experimenting (it'd be hard to explain people why in the nine hells am I using 2d10 instead of d20), but that's still some food for thoughts.

 

[clearstream]

And critical hits are going to be far less common (1:100 instead of 1:20). Maybe make 2s auto-fails, and 19-20 criticals?

 

[Paul Farquhar]

It makes critical hits much less likely, and thus nerfs crit-fisher builds.

 

[clearstream]

It makes critical hits much less likely, and thus nerfs crit-fisher builds.

Like Champions. Thank goodness!

 

[John R Davis]

And it's not as satisfactory to watch

 

[NotAYakk]

2d10's main effect is to make each +1 bonus and each +1 DC larger in impact compared to 1d20. And it needs a different crit mechanic.

Ignoring the crits, the variance of 1dx is (x+1)(x-1)/12, or (x^2-1)/12. If you add dice, you add variance. So 399/12 vs 198/12 for d20 vs 2d10.

Then take the square root to get the standard deviation, getting 5.8 and 4.1. SD(d20)/SD(2d20) is 1.4, or about the square root of 2. (this is generally true: N dice that add up to the same average as 1 die has a SD of 1/sqrt(N) of the one die).

So a game using d20 where PC stats varied from 10 to 28, and proficiency went from 3 to 9, would play a lot like D&D using 2d10. Also, saves are DC(7+prof+stat), shields are +3 AC, magic items go to +5, max dex on medium is +3, plate has 21 base AC, etc.

A level 10 wizard with 20 int and 4 prof and a save DC of 17 would feel a bit like a 26 int 5 prof save DC 20 character.

A PC in plate and shield would feel like having 24 AC roughly.

 

[clearstream]

A level 10 wizard with 20 int and 4 prof and a save DC of 17 would feel like a 28 int 6 prof save DC 22 character.

A PC in plate and shield would feel like having 24 AC roughly.

Very nice approach to the analysis. And well summarised in these observations.

 

[jgsugden]

There are advantages to bell curves, but the system has to be designed for it. D&D wasn't. This essentially makes easy stuff easier, and hard stuff harder - potentially much, much harder.

Let's say that you need a 17 to hit a tough AC enemy. Under a d20 system, that is 20%. Under a 2d10 system, that is a 10% chance. It would take you roughly twice as long to take that foe down - and most enemies that have a very high AC relative to attack bonuses tend to be real nasty. Imagine spending twice as long with a Roper at 3rd level.

This flips around and makes PC AC more valuable. If you get that AC up to the low 20s early on (Warforged Fighter (Defense Style)/Forge Cleric), you really get a huge benefit - while lower AC bruisers like great weapon master barbarians are going to feel more pain. DMs with these PCs in their game will tell you that they do not need help.

 

[DEFCON 1]

I used 2d10 for Ability Checks in my Eberron campaigns and found it lovely. I had set DCs of 11, 14, 17 and 20 that I use for most actions and having players roll 2d10 worked wonderfully hand-in-hand with that.

I did not choose to use 2d10 for attack rolls and saving throws and instead kept the 1d20... first off to indeed maintain the critical on a Nat 20 mechanic... but secondly because the window of modifiers for attack rolls and save DCs was and is much narrower than skills. For skills a starting PC could have a full 8 point swing of modifiers from a -1 all the way up to like +7, so using the curve of 2d10 made the higher modifier skills have more import psychologically. Higher DC checks that resulted in a low modifier PC succeeding while the higher modifier PC didn't occurred MUCH less frequently than when using 1d20. And that made for a better feeling game experience at the table.

For attack rolls and saves, the width of the modifier window is always much narrower (almost all PCs starting with in and around +5 attack / DC 13 save DC give or take a point) meant that there was much less need to fix the perception of high modifier versus low modifier success rates. So the 1d20 for those was fine for us.

 

[DND_Reborn]

Since I've started using fixed DC (10- fail, 11-17 partial success, 18+ full success), I was playing around with various ways to shift the numbers to generate more partial successes on average, while still allowing +0 characters to achieve full success. The most radical solution I had in my list was to switch to 2d10, which I recently tested, just as a joke.

But, to my surprise, it worked. Like, worked very well. It breaks nothing, while seriously reducing σ. The only bug I've encountered is that automatic misses just never happen, but I'm honestly fine with that.


I guess it's not something I'll be more than experimenting (it'd be hard to explain people why in the nine hells am I using 2d10 instead of d20), but that's still some food for thoughts.

Yep. It works fine. A lot of tables use it for ability checks, some even for attacks and saves because they don't like the swinginess of the d20.

And it doesn't mess up anything, despite what some people might think--it changes them, but those things still work fine.