Short Method III Study

[Lemunda's Brother]

Method III for character generation in the DMG is rolling 3d6 six times for each attribute, you may recall. I tried this a long time ago and generated a Bard, Herulind, a half elf with one attribute 18, two 17s, a 16 and two 15s. I concluded that this was much too good and wouldn't use this method again, nor allow my players to use it if I were DMing. I recently discovered the useful dice rolling site anydice.com and tried it out. The code to put in for Method III is:
output [highest of 3d6 and [highest of 3d6 and [highest of 3d6 and [highest of 3d6 and [highest of 3d6 and 3d6]]]]] named "Method III: 3d6 Six Times"
On average, you will roll 14 or 15 for an attribute. Low attributes are very unlikely, with only a 3 in ten thousand chance of getting an 8, but 18s are still rare, with only a 2.75% chance of getting one (0.4% with a straight 3d6) and so roughly a 15% chance that one of your attributes will be an 18. This seems roughly in line with EGGs suggestion on character power in the DMG: my view was skewed because it turns out I'd been exceptionally lucky with my only character created using this method...

[Chainsaw]

Interesting, thanks.

[francisca]

This seems roughly in line with EGGs suggestion on character power in the DMG: my view was skewed because it turns out I'd been exceptionally lucky with my only character created using this method...

ARE YOU NUTS?

This is the interwwwebz, where opinion and anecdotal evidence is held in higher regard than actual fact!

[prespos]

These articles in Dragon may be of interest:

Odds "Let the Good Dice Roll" Scott David Gray 132(38) D&D1
"What Are the Odds?" William Keely 39(39) D&D1
"What Are the Odds?" Arthur J. Hedge III 117(14)

Thanks to http://www.aeolia.net/dragondex/

[thedungeondelver]

Here's a direct link in case anyone would like to try it.

http://anydice.com/program/4c41

[bobjester]

I must have really bad luck or fawlty dice. I've used Method III to generate a lot of potential NPCs for city or human encounters, and out of 36 sets of 6 ability scores rolled, I think I rolled an 18 once or twice, but at least twice as many 17's, and lots of 13 through 16, but also a lot more lower than 10 scores, and an inordinate amount of 3's.

The dice are even worse when I roll 4d6-L. Statistically, I roll 3d6 with higher average results than with 4d6-L.

[Lemunda's Brother]

I must have really bad luck or fawlty dice. I've used Method III to generate a lot of potential NPCs for city or human encounters, and out of 36 sets of 6 ability scores rolled, I think I rolled an 18 once or twice, but at least twice as many 17's, and lots of 13 through 16, but also a lot more lower than 10 scores, and an inordinate amount of 3's.

Either you're using the method incorrectly, or you must have angered Zeus or something: the odds of getting a 3 in any attribute are the same as rolling a regular 3d6 character with a score of 3 in every attribute. The odds of this happening are just over one hundred trillion to one...

[Chainsaw]

The odds of this happening are just over one hundred trillion to one...

[bobjester]

I must have really bad luck or fawlty dice. I've used Method III to generate a lot of potential NPCs for city or human encounters, and out of 36 sets of 6 ability scores rolled, I think I rolled an 18 once or twice, but at least twice as many 17's, and lots of 13 through 16, but also a lot more lower than 10 scores, and an inordinate amount of 3's.

Either you're using the method incorrectly, or you must have angered Zeus or something: the odds of getting a 3 in any attribute are the same as rolling a regular 3d6 character with a score of 3 in every attribute. The odds of this happening are just over one hundred trillion to one...

You should see the rotten scores when I use 4d6-L. They're worse!

The odds of this happening are just over one hundred trillion to one...

Now, that right there is funny, I don't care where you're from!