3d6, re-roll 1's

[waysoftheearth]

I guess by "reasonably equal" I meant more that everyone has "something to work with" rather than everyone is exactly equal.

Yep, rolling is more fun

[aher]

Here's another possible solution the conundrum of rolling "sucky numbers at the bottom" of the 3d6 ability score distribution...

In AD&D 1e, the DMG (p. 15) states that abilities are "composites". Each subsumes 3-6 characteristics. A PC with a low ability score could be exceptional in one of these characteristics, but very poor in the others. A PC with an average ability score could be average in all characteristics; or (alternatively) better than average in some characteristics and below average in the rest. A PC with a high ability score could be really good in say all but one characteristic.

Furthermore, individual characteristics can range below 3 and above 18, but together they "average out" to the relevant ability score.

Example: The CHA ability comprises 3 characteristics: beauty, persuasiveness, and magnetism. Caesar, Napolean, and Hitler have CHA 18, but clearly their beauty << 18. So we could imagine Hitler has magnetism = 24, persuasiveness = 24, beauty = 6. The average is CHA = (beauty + persuasiveness + magnetism)/3 = (24 + 24 + 6)/3 = 54/3 = 18 = CHA. So, while he is ugly (beauty=6), Hitler unfortunately can draw a very large crowd (magnetism=24) and convince them what to do (persuasiveness=24).

Another example: Parker the Thief is an expert cat-burglar and safe-cracker. She is socially awkward (persuasiveness=0) and relates poorly to most people (magnetism=0) but is extremely physically attractive (beauty=18). So her CHA = (0+0+18)/3 = 6.

In sum, we can't deduce anything about the subsumed characteristics on the basis of a low ability score. So a low ability score isn't necessarily "bad", any more than a high score is "good." There's a lot of flexibility in that we can finagle a high characteristic from a low ability score.

The DMG p. 15 lists the following characteristics subsumed by each ability:

STR
physical power
endurance
stamina

INT
mnemonic
reasoning
learning ability skills in additional areas outside the written word

WIS
willpower
judgment
wile
enlightenment
intuitiveness

DEX
hand-eye coordination
agility
reflex speed
precision
balance
speed/movement

CON
physique
health
resistance
fitness

CHA
physical appearance
persuasiveness
personal magnetism

[Finarvyn]

aher, this sounds a lot like the direction later 2E rulebooks took with the sub-attributes. Not a bad idea, but leaves open the potential for tweaking the numbers to min/max a character. It would be interesting to have an equal number of sub-attributes for each stat and then have players roll for each sub-attribute with 3d6 (or whatever).

[aher]

This link rolls 6 random characters using the distribution in the OP (3d6, re-roll 1's), using the online Troll dice roller.

While this link computes the probabilities of the distrib in the OP.

Note that this distribution is exactly the same as that of 3d5+3, shown here.

Which means if you happen to have three GameScience d5s, then you can generate ability scores using this distribution without ever rerolling.

[giantbat]

Exalt for aher. I dislike rerolling. Another procedure which would achieve the same range of values, without rerolls, using more commonly available dice, but with a stronger tendency to average values: 4d4+2

[oldkat]

I am currently analyzing the Supplement VII section on character generation and find it intriguing.

[runequester]

I rather like it, though the guys I game with tend to expect "3D6 in order!" when its old school D&D

Someone upthread mentioned making a 7th roll and using that for a trade, which we had done in D&D games in past. I rather liked it.

[rms]

Back in the day, before we read about rolling 4d6 drop the lowest, we used to use the following method:

Roll 3d6 for each stat in order. However, for each roll you could pick up and reroll a single die. The catch is that you had to live with the result.

It still allows the entire range from 3-18. It just makes low rolls less likely, while not increasing the likelihood of high stats by much.

[aher]

Dec 19, 2012 11:33:12 GMT -6 oldkat said:

I am currently analyzing the Supplement VII section on character generation and find it intriguing.

I had a peek at Supplement VII, Lost Lore by RC Pinnell, pp 3-4, ABILITY SCORES.

If I'm reading this ability score generation method correctly, then one gets a very interesting, W-shaped probability distribution with the three modes at 6, 12, and 18; and P(X=6)=P(X=12)=P(X=18)=13.889%.

As with 3d6, reroll 1's, Supplement VII's range is 6-18 and the mean is also 12. The two distributions have extremely different standard deviations: 3d6, reroll 1's has σ = 2.45 while Supplement VII has σ = 4.41.

Click here if you want to see a plot of Supplement VII's probability distribution using the Troll dice roller.

And click here if you want to roll-up a half-dozen PCs using Supplement VII's distribution.

[waysoftheearth]

Dec 20, 2012 2:30:34 GMT -6 @aher said:

And click here if you want to roll-up a half-dozen PCs using Supplement VII's distribution.

I clicked it, and the first PC to appear before my eyes was this:

Str 12
Dex 18
Con 16
Int 18
Wis 18
Chr 18

[aher]

Dec 8, 2012 6:33:22 GMT -6 Finarvyn said:

...sucky numbers at the bottom...

Crazy idea: Send your players to the Bizzaro World, where all their abilities are inverted according to the ability score inverting permutation π = (18 3)(17 4)(16 5)(15 6)(14 7)(13 8)(12 9)(11 10). Then see how much they like all those 18s on their character sheets.

"Us do opposite of all Earthly things! Us hate beauty! Us love ugliness! Is big crime to make anything perfect on Bizarro World!"
The Bizarro Code

[oldkat]

I just rolled up 6 characters, and it didn't give me anything I havent seen generated by loads of other methods--with the exception that no score was below 6.

Here are the results,
S D C I W CH
7 14 7 17 17 12
7 17 12 14 6 17
6 12 6 18 17 7
18 10 13 6 12 15
9 6 12 8 7 15
12 7 6 9 12 7

Some good, some not so good. But none ubermunchy, or hopeless.

[Deleted]

I had a thought about this the other day. What if you altered the number that allowed a reroll based on the character's race instead of giving racial ability score adjustments? For example, a hobbit might not get any reroll for Str or Wis but would reroll 1s and 2s for Dex and Con. Dwarves wouldn't get rerolls for low CHA but might get to reroll 1s, 2s, and 3s for CON (or some such combinations). This would allow demi-humans to have the equivalent of a bonus without them ever going over 18.

[Deleted]

I think it's an excellent idea. I'm using it from now on.

[blackbarn]

I wouldn't mess with the usual method, but I would have no problem if a player wished to use this one. Personally, I don't like the feel of it, and love the simplicity of 3d6 down the line. It's elegant, even if it doesn't produce superheroes.

[Deleted]

waysoftheearth: I *like* that!

Let everybody roll up characters using strict 3d6 and just as the whining about "crappy characters" reaches a crescendo, tell all the whiners to subtract each of their rolls from 21 then shut up and play! <patented DM's evil grin>

Lance ==)----------------

[bestialwarlust]

My group is really doing good with this I have one player that has a fighter with a str 6 and dex of 18. Everyone rolled 3d6!

[spacelem]

I'll hopefully be running OD&D in a few weeks, and I've decided to go with roll 3d6 in order, then sum all ability mods. If the total is negative, then invert all 6 stats (new = 21 - old).

This keeps the stats at 3-18, but ensures no one ends up with an overall penalty, and you can't be awful at everything. It might also be a nice surprise ("these are awful stats... No wait, they're actually okay ")

I'm going to crunch the numbers tomorrow to see what the final distribution of scores looks like after inserting negative sets (my intuition is that a sum of +0 will still be most common, but we'll see).

I might also offer swap any two stats as well (or instead). Not sure yet.