last day (14 days later) » 

The reason I put it in the comment and not as an answer was because my knowledge of programming is minimal at best. I wouldn't know how to approach the actual writing the function, but I do know how programming structure such as storing variables works.
Also, the comment for posterity:
Your best bet at modeling this in anydice is to define the rolled dice from C as variables and then write a function to remove those variables from the A and B sets to find your end result. To find the non-eliminated dice from Pool B use the defined variables from C and store any non-removed dice based on that information to output when the function ends. — Sandwich 18 mins ago
I created this room because I figured it'd create a single space where people can discuss the solution to this conundrum if they're interested. Brainstormy comments on that question don't seem like they'd be too bad really, but there are limitations to comments that'd probably prove a nuisance after a while.
Yeah, but anydice is weird - the variables are dice with their values and probabilities all mixed up together - removing specific values is problematic.
I agree, I don't really know how to put my thoughts into actual code
But I can sort of shape them into concepts of how the code would work
It'd probably have to use variable defining, and a function with a few if/then/else statements
 
2 hours later…
05:10
FWIW, this is what it would look like in Ruby: gist.github.com/scgruber/e4d69a81ae0810a0e001
But that's just to perform the roll.
05:27
I think the expected value for bCount is just SUM[1<=i<=6] { (b/6) * P[i not in C] }
P[i not in C] equals P[i in C] which is SUM[1<=k<=c] { (5/6)^(k-1) * (1/6) }
05:43
WolframAlpha claims that this shakes out to b * (5/6)^c.
Plotted for 0 < b < 10 and 0 < c < 10, where each color band corresponds to an integer result.
 
1 hour later…
07:00
room topic changed to Technoir dice calculations: Hashing out these weird things and anydicing them. For the question: rpg.stackexchange.com/q/66991 [anydice] [technoir]
 
2 hours later…
09:25
I have hacked something together in python that returns the statistics for a lot of rolls. Complete with a rudimentary bar chart!
(written in python3, but python2 might work as well)
If anyone comes up with a neat solution, rather than a brute-force one, this can serve as a reference.

last day (14 days later) »