Rolling IQ

Hey guys, I am trying to figure out the best way to roll for IQ (intelligence quotient) on a character. IQ has a mean of 100 and a standard deviation of 15 (so, a variance of 225) and I'm trying to find the most efficient way to roll standard dice (d2, d4, d6, d8, d10, d12, d20) to get this probability distribution.

First I looked at the factors of 100 (1, 2, 4, 5, 10, 20, 25, 50, 100) and used that many number of rolls to find the variance where they all had a mean of 100. The two closest were 10d19 (var 300) and 20d9 (var 133.3~). Using normal dice, 9d20 (mean 94.5 var 299.25) was as close as I could get.

So, I'd need to roll multiple kinds of dice. So far, the closes I have gotten is 5d20 + 6d10 + 2d6 + 3d4, which gives a mean of 100 and a variance of 225.3~ standard deviation of about 15.011. Obscenely close... but not perfect.

Has anyone gotten closer, or even figured out ab exact combination of rolls to get this? (And, am I on the right board?)

Other urls found in this thread:

anydice.com/program/d4eb
twitter.com/SFWRedditVideos

Right board but why?

Have you considered EQ/Emotional Intelligence as well, or are you just doing straight up IQ?

straight irl IQ

Just use FATAL.
It doesn't have direct IQ, but stats are on range where 100 is average, so you could just take the different sub-intelligence and use their average as IQ.

4d100 / 2 - 1 eh? The variance is too high; 9d20 +5 would be better and simpler than that

>4d100 / 2 - 1

Its (10d100/5)-1 , your roll was just a old mistake at 1d4 that someone started to use at other places

>at 1d4
1d4chan the wiki

Ah, I had looked up the FATAL book and that's what it had.

(14d100/7)-1
has a standart deviation of 15.43
anydice.com/program/d4eb

1d100
If you're not man enough to play a blithering retard you don't deserve to know IQ

>Ah, I had looked up the FATAL book and that's what it had.
I have some version released at 9/18/2004 (latest version I could find, the book is not complete, its just part of the book, since he released partial versions of the book after some amount of time) and the roll is
(10d100/5)-1

PS: I also have the version from 04/05/2004 this one is the entire book, and also has (10d100/5)-1

Why not just 3d100, take middle? Standard deviation is around 22, but it's probably still fine for practical application

>on a character. IQ has a mean of 100 and a standard deviation of 15 (so, a variance of 225)


The average is not 100, 100 is the average.

WUT?


WTF???


They set 100 to be the average IQ, if everyone excluding mentally retarded people and people that do iq tests and stuff survive, the average will still be 100, the average iq will be corrected to be 100.

>take middle
Because 100 is the average, not 50

>?
Messed up, the mean is actually 50. Ignore this post.

Though, on a related note, what would be the standard deviation of rolling 3d10 twice, take the middle each time, and using one set for the tens place and one for the ones place?

3d6 times 10.

Have you considered just making a d100 table that maps to the bellcurve?

Also the standard deviation is only 15 for men, it's about half that for women

>trying to find the most effecient way
>rolls non weighted dice
you need to roll something that fits the bell curve.

EQ is non-empirical and therefor not scientific

Mapping to the bellcurve is what I am trying to do with these rolls, user.

And yet it has strong correlations to life success and a stronger inverse correlation to likelihood of being jailed.

IQ isnt even used as it was intended, the inventor made it to help identify struggling and advanced children, not to be broadly expanded beyond its mandate to be "the intelligence score"

No, I mean a manual map. Like a 1 is iq 65, a 2 is iq 67, 50 is iq 100, 100 is iq 135 and such

>man invents thing for niche purpose
>others discover it has much broader use
Yeah, that's never happened before

Ah, I see. I could do that but, I'll keep on this with this idea...

You lose the extreme ends with that.

X=36d6

E(X)=36*3.5=126
Var(X)=105

Y=A+B*X
E(Y)=A+B*126=100
Var(Y)=B^2*105=225

Solve for A and B

It's a bitch because 225's common factors (1, 3, 5, 9, 15, 25, 45, 75, 225) don't really overlap with 100's.

B=1.463850...

A=-84.44511...

3d6 x10 ya git.
Will average slightly more than average and stray from extremes - perfect for adventurers.

Or just do it in Excel and stop being an autistic dice retard

Do you really want to roll an IQ of -125 or of 225? You'd want a range more in the 50-150 range, so just roll 2d10% and add 50.

Just use the same dice roll they use in real life

You misunderstand, I am just talking about statistical variance. The distribution I described in the OP is the one that they use for IQ.

Confer pic related:
The center (mean) is 100.
The standard deviation (distance between the bars) is 15.
The variance is just the SD squared: 15^2=225

3d6 times 10 actually works very, very well for the numbers you want. The mean comes to 105, but also the standard deviation ends up being 17.2, which is extremely close to the value you want

>3d6 times 10 actually works very, very well for the numbers you want. The mean comes to 105, but also the standard deviation ends up being 17.2, which is extremely close to the value you want

((14d100 )/7)-1 is better

But that would require rolling 28 dice and dividing (probably requiring a calculator), versus just 3 with simple addition

It would be, but OP is clearly not looking for reasonable solutions, he's searching for maximum autism.

He may have found it.

Wait, so then you want a multi-dice roll (because that produces a bellcurve naturally) but you really want a variance from 70 - 130, or a 60 point range.

So basically 10d6 + 65.

unironically 3d6, with 10.5 normalized to 100 and each point of INT worth 5. the curve matches almost exactly.

if you really want to model geniuses and retards, add rolls for 18/xx and 2/xx on a natural 18 or 3.

Roll a 1d4. 1-2 subtract from 100, 3-4 increase it. Anytime you roll max or roll minimum, you roll again to alter it further and increase the dice size.