Is there a way to calculate this?:

I've got 3 twelve sided dice that i want to roll against another 3 of the same dice, the dice are then arranged in pairs (1 form each opponent) from highest to lowest roll, and i need to find the probability of the each player getting a better roll in 1, 2 or 3 of the dice, then find if ties goes to player one whats the probability of him getting better rolls in 1, 2 or 3 of the dices.

Attached: sxsci.jpg (360x540, 16K)

Other urls found in this thread:

frinklang.org/
twitter.com/NSFWRedditGif

Do your own homework kid.

this sounds a lot like homework

actually i'm trying to develop a game i'm cant find a way to figure this probabilities

i'll do this if you give me a tl;dr with less words

Allright lets try:
Whats is the chance of a player rolling 3 twelve sided dice and getting 1, 2 or 3 better results against anoter set of 3 rolls, higher rolls should be paired together.

Yes. If you can't do this, don't bother trying

unless i'm missing something this problemes is more complex than it looks, the fact that higher rolls are paired completely changes the table of probabilities, if anyone can at least point me a direction i can use to calculate this i would be gratefull.

>for (amount of roles going down) if (role 1 = role 2) {then this}
Veery complex

not that simple, example: a player A that rolls (11, 4, 8) loses the 3 rolls agains a player B that rolls (6, 9, 12), because the pairs of higher number would then be (12b,11a), (9b,8a) and (6b,4a).

you would do it for each element in the array/vector. Imagine being this retarded

so instead of doing math you solution is just to create a program to crunch numbers...

What do you want from me user. Just do a function to sort it from least to greatest and compare. Or just do if(element == greatest then look in other vector and repeat until no numbers are left)

Virgin math fags btfo'd by chad CS degree!

0.5^n, where n is the number of dice you compare

You asked this question on Jow Forums. Most people here are just going to want to sim it because it's easy and proves itself correct. If you want answers from people who took higher order stats classes go to /sci/. Besides you said you wanted this for a game where you'd end up implementing the Sim anyways. Sounds like you really are doing homework and you're upset people are doing it in a way your professor won't give you credit for.

Independent sampling from discrete uniform distributions
Let a, b , c .. be the sides of each dice A, B, C, ...
Let r_a, ...be the result of throwing dice A, ... , etc.
p(r_a, r_b, ...) = 1/a * 1/b * ....

Use Frink: frinklang.org/

Permutations

crunched numbers

57.5% for the highest pair
55% for the medium pair
57.5 for the lowest pair

If you don't know the math just write a short program to run 100000 simulated games and use the empiric probability.
You very likely do not need more than three significant figures.