I was wondering whether you have a suggestion on how to program a specific R-program for (a lot of) d20 testing.
I have the null hypothesis that the face-up value of any dice (I'm mostly interested in d20, but the mechanic is the same) does not have any effect, positive or negative bias, on the coming roll.
The way I would like to make the experiment:
I create a table with 10 small boxes on it with walls that all can be removed at the same time. I will make a mechanic tipping of the table to simulate a human throw.
I will write down every result for each dice, input it a spreadsheet and later
scan()it into R.
Here is my problem now:
I want R to tell me which face-up value gave the highest value afterwards. This means I need to program functions with twenty variables and then somehow, probably by using
lm(), make it possible for me to analyze whether the p-value is lower/higher than 0.05.
Of course, I could force the face-up value to be the same every time, but I would rather test all face-up values at the same time if possible.