D&D General Normal Distribution Ability Scores

I think so, too. I might have to run some numbers to see what sort of arrays this creates.

Here are 100 arrays (always inverted to maximize total, which might not always be the choice):
9,11,14,11,8,13
13,11,8,9,14,10
17,12,10,7,11,14
9,15,16,13,8,13
10,11,12,11,18,15
11,8,15,6,10,13
11,14,16,9,8,14
11,12,8,12,10,11
12,14,16,8,6,12
5,13,10,9,10,18
16,11,11,7,12,6
13,13,14,13,12,12
17,14,9,8,13,8
15,13,15,9,9,15
12,12,13,12,8,12
10,17,14,8,10,11
15,9,8,16,11,13
7,12,14,11,10,13
17,9,10,12,9,11
9,14,12,10,7,15
12,12,9,16,13,18
11,9,11,12,9,16
12,9,8,10,14,15
13,10,9,8,12,15
13,10,7,12,13,10
16,6,15,12,7,13
14,10,13,15,9,9
14,14,4,13,9,13
13,11,17,16,14,12
14,16,7,9,11,9
12,16,15,7,10,6
11,15,10,4,11,14
11,9,15,15,14,10
9,17,7,9,16,9
10,13,6,8,16,10
8,10,11,8,10,14
10,12,17,10,8,10
12,14,9,10,9,10
10,9,11,16,8,13
9,8,11,16,15,12
14,6,15,15,15,7
12,13,3,11,17,13
12,12,6,14,9,12
8,14,10,13,13,9
10,9,10,15,13,11
12,13,10,10,7,11
12,11,10,14,9,11
15,6,12,12,10,14
8,15,9,7,16,12
5,11,10,16,8,13
8,15,15,14,8,4
12,12,9,12,15,15
15,13,11,15,9,6
9,10,10,16,9,10
8,16,13,11,10,9
12,12,8,9,12,12
11,12,9,13,10,15
12,15,13,8,10,12
14,14,8,10,8,9
15,17,10,10,12,13
14,14,14,13,8,10
14,12,10,15,9,12
13,9,11,12,13,8
10,6,12,10,13,14
9,14,10,9,14,15
14,11,12,14,11,12
8,17,7,16,9,11
10,13,9,12,9,11
16,6,13,13,8,8
14,15,8,14,17,9
12,16,15,5,12,17
16,10,13,7,10,12
11,16,15,15,9,8
12,11,14,11,10,9
8,16,10,8,12,16
15,14,15,9,9,13
8,6,16,10,14,11
11,8,6,11,11,16
11,13,12,9,8,13
15,13,13,10,8,10
12,11,14,11,9,10
10,16,11,13,13,11
12,15,14,9,6,14
11,8,12,15,8,15
14,11,9,8,13,13
11,7,17,12,12,13
10,9,14,12,11,8
16,11,10,15,12,11
13,12,9,10,8,11
7,14,12,8,7,16
11,16,10,9,11,7
11,14,10,15,14,13
11,13,9,12,11,13
15,12,6,14,14,6
10,13,12,15,7,8
8,11,10,14,7,13
11,11,8,15,10,12
14,9,7,14,15,12
9,11,13,15,12,12
16,8,12,8,14,13
 

log in or register to remove this ad



Here's a breakdown of the frequency of each score, after 10,000 iterations: (edit: that's 10,000 iterations of rolling six scores, so numbers should total 60,000. The actual math is pretty wonky because of the all-or-nothing nature of the inversion. For example, an 18 will get turned into a 3, if inverting the other five scores leads to a net numerical benefit.)

3 68
4 302
5 681
6 1553
7 2308
8 4513
9 5378
10 7492
11 7440
12 8621
13 7057
14 6007
15 3978
16 2748
17 1383
18 471
 
Last edited:

My first character I rolled randomly on the questionable "5d6 drop lowest 2" system. Unsurprisingly, I had an 18 before boosts and my lowest stat was 13. This is not a system I would recommend for people wanting challenging characters to play. But I agree that I picked which stats to put which scores into. However, there is a fallacy that I rolled a 4d6 and got 1 1 3 2 I would start over or dump into Charisma. Point buy eliminates that but also makes characters almost as repetitive as the standard point spread.
 

  1. Not being 100% equal to everyone else, but not being 50% of anyone, either. Nobody likes having a +5 primary when others are stuck with a +2.
I will say that most people try to get to 17 or 18 on their primary so we're all around +3, +4, or +5 for primary. However, I've only played for a couple years so I could be just lucky.
 

My first character I rolled randomly on the questionable "5d6 drop lowest 2" system. Unsurprisingly, I had an 18 before boosts and my lowest stat was 13.
I've been using that system (5d6 drop 2) for decades and can say with confidence the results you got for that character were unusually high.

Low of 9-10, high of 16-17 is far more typical; and the peak of the bell curve is around 13.

EDIT to add: I should note we're playing an oder version where bonuses don't kick in until 15.
This is not a system I would recommend for people wanting challenging characters to play. But I agree that I picked which stats to put which scores into. However, there is a fallacy that I rolled a 4d6 and got 1 1 3 2 I would start over or dump into Charisma. Point buy eliminates that but also makes characters almost as repetitive as the standard point spread.
1-1-3-2-1 can also occur (and has, many a time!) on 5d6. Unless it's a Cleric I'll often drop that low stat into Wisdom just for kicks. :)
 

A method I like has a push-your-luck aspect to it:

Roll 3d6 for each ability in order. I may choose either to play with what I rolled, or discard all six rolls and try again, for a maximum of six attempts.

Generally after two or three attempts I will have the scores to make a good enough character, but then have to decide whether to risk losing it to try for an even better one.
 

It would lead to even worse min maxing. A lot of people would feel that they "need" to have the highest possible score in their main stat, and they'd dump all stats not directly relevant to their class as low as possible in order to get it.

The issue with point buy is that all stats cost the same regardless of their usefulness to you. And with floating ASIs your main stats effectively cost even less than the other stats. This should be other way around, your main stats should cost more than the other stats, so that the trade-offs would be more meaningful. Like getting your wizard a two points less int so that you can get two points more of strength or even con sounds like a bad deal, but what if by getting that two point lower int would get you +2 to three or four other stats? At some point the balance is such that it is at least worth considering.

To OP. I like that you're rolling stats in order. That's the only way that randomisation makes any sense, as then you're actually randomising your concept, and not just the power of the concept like it is usually done.
I might be coming in late responding to a post from page 1 on page 6 after reading all 6 pages, but you are describing a benefit as a negative thing here. Those super low attributes are a good thing because it means every PC in the group has area(s) where they have extreme excellence and areas of large enough weakness that they they need to lean on each other's strengths instead of just expecting their PC to be good enough anything so they don't need to look to Bob when his strength in [whatever] seems important for the party. The low scores with rolling and point buy tables that have a 3-18 spread were a positive influence in play through the spread of scores represented across the party
 

I'm not a fan of point buy but I do want my PCs to have good scores. The eternal 18 bothers me less than the eternal 9 dump stat.

One idea I've considered is giving the PCs a pool of dice that they can assign to each stat. Then they roll the mini-pools and take the top 3. You have to put 3 minimum in every stat of course. That means primary stats will often be very good and there won't really be a dump stat.

An example: Take 6 stats and 4d6 for every stat on average. So you give them 24d6 to arrange how they want. You might also be generous and just say 30d6 for a high heroic fantasy game.
 

Remove ads

Top