D&D General Normal Distribution Ability Scores

[ezo]

Nah, PCs are special. This was more an exercise in seeing what a more normal distribution could be for, and how to switch rolling to d4s

Well, if you really enjoy rolling, the best method I have found for creating a normal distribution is 15d2-12.

For PCs, roll 15d2-10 (AnyDice). This shifts the PCs to an average modifier of exactly +1, with a range of -3 to +5.

As you can see, for PCs you have (basically) the follow chances for the listed ability modifiers.

Edit: of course, you can still roll 18 or higher as a PC, but the chance of a +4 modifier is about 1 in 270, and +5 is less than 1 in 32000.

 

[Peter BOSCO'S]

Point buy with 15 points, not 27 or a "Weak Standard Array" of 8, 9, 10, 11, 12, and 13 (then add bonuses from Background) will get you something suitable for NPC's with average stats.

 

[Lanefan]

If the system in use doesn't give me the chance (not guarantee, just chance) of starting with a 17 in one stat and a 6 in another, or to roll up one character with a net +9 modifier while the next has a net +3, I'm almost certainly not interested.

 

[jmartkdr2]

Another method I’ve heard of:

You pick stats in tiers before rolling. Say I’m thinking of playing a druid. I put my stats in priority order: wis, then con, dex, int, str, cha. I actually only need three tiers.

The top two stats I roll 5d6 and keep the best 3. So probably really good numbers. My druid will be Wise and tough. But when I roll for Wisdom, that’s it. That’s my starting Wisdom.

The next two are 4d6k3. (The usual method). Probably good but I might get a surprise.

The last two are just 3d6. Maybe I’m strong, maybe I’m a wet noodle. We won’t know until rolling day.

This preserves my base concept since I can be sure my “main” stats will be good, but leaves in some variety since my non-class-dependent stats could be anything.

 

[ezo]

Another method I’ve heard of:

You pick stats in tiers before rolling. Say I’m thinking of playing a druid. I put my stats in priority order: wis, then con, dex, int, str, cha. I actually only need three tiers.

The top two stats I roll 5d6 and keep the best 3. So probably really good numbers. My druid will be Wise and tough. But when I roll for Wisdom, that’s it. That’s my starting Wisdom.

The next two are 4d6k3. (The usual method). Probably good but I might get a surprise.

The last two are just 3d6. Maybe I’m strong, maybe I’m a wet noodle. We won’t know until rolling day.

This preserves my base concept since I can be sure my “main” stats will be good, but leaves in some variety since my non-class-dependent stats could be anything.

We've done this before. If you have an idea of what you want your PC to look like, it works well IME.

With the 5d6k3 you average about 13.4, and over 50% of the rolls will be 14 or better (16+ is still uncommon at under 25%).

 

[Xeviat]

Another method I’ve heard of:

You pick stats in tiers before rolling. Say I’m thinking of playing a druid. I put my stats in priority order: wis, then con, dex, int, str, cha. I actually only need three tiers.

The top two stats I roll 5d6 and keep the best 3. So probably really good numbers. My druid will be Wise and tough. But when I roll for Wisdom, that’s it. That’s my starting Wisdom.

The next two are 4d6k3. (The usual method). Probably good but I might get a surprise.

The last two are just 3d6. Maybe I’m strong, maybe I’m a wet noodle. We won’t know until rolling day.

This preserves my base concept since I can be sure my “main” stats will be good, but leaves in some variety since my non-class-dependent stats could be anything.

That's interesting! Now you've got me thinking of a character building mini-game.

 

[Horwath]

Well, if you really enjoy rolling, the best method I have found for creating a normal distribution is 15d2-12.

View attachment 380514
For PCs, roll 15d2-10 (AnyDice). This shifts the PCs to an average modifier of exactly +1, with a range of -3 to +5.

View attachment 380515

As you can see, for PCs you have (basically) the follow chances for the listed ability modifiers.
View attachment 380517

Edit: of course, you can still roll 18 or higher as a PC, but the chance of a +4 modifier is about 1 in 270, and +5 is less than 1 in 32000.

15 dice? little overboard maybe?

If rolling is ever on the table(and in 99% of cases it's not, thankfully) then I prefer dropping +2/+1 ASIs and roll;

(4d3D1)×2, this gives spread of 6-18 with 100% chance of avoiding next to useless odd scores.

 

[DEFCON 1]

If the system in use doesn't give me the chance (not guarantee, just chance) of starting with a 17 in one stat and a 6 in another, or to roll up one character with a net +9 modifier while the next has a net +3, I'm almost certainly not interested.

Heh heh... yeah, but your tables go through PCs like they are paper cups, so it's not like any of them are going to survive long enough for the score disparities to ever make much of an impression.

 

[TwoSix]

Method III forever.

Roll 3d6 six times, per stat, take the highest, in order.

 

[DEFCON 1]

The one campaign I had players do randomized "roll in order" ability scores, I used the playing card method and it worked really well for me.

Deck of 16 cards: 4, 4, 5, 5, 6, 6, 7, 7, 7, 8

Shuffle and deal six piles of two cards each. Add each pair of cards and assign to abilities in order. Produces mainly numbers in the 9-14 range, but with a slight chance of getting a 15 (which could become a 17 with score bump from species or background) and an extremely slight chance of an 8.

And obviously you can adjust the numbers you get by swapping cards in and out. More 5s and 6s bring numbers closer to the middle, more 3s, 4s, and 8s give more chances for high and low scores. But all in all the arrays end up in a good parity even though they will all have slight total modifier differences.