Originally Posted by Alan Mendelson View Post
To everyone who thinks I'm wrong: please get yourselves two dice, and a cup, shake rattle and roll, and then figure what the odds are for the second die to also be a two.
Alan, your suggestion here is quite revealing. You are telling us to do something that you have not done yourself. Because if you did you would be quite surprised by the results.

All that's needed is a spreadsheet with two columns. After every roll where "at least one of the dice is a two" make a mark in the 1st column. Then take note of whether the other die is a two. If it is then make a mark in the 2nd column. Keep collecting those stats. You will start to see a pattern developing. Through about 500 decisions you will find that the frequency of the other die being a 2 is a hell of a lot closer to 1 in 11 than 1 in 6.

My prediction, you heard it here first, is that through 500 decisions the other die will be a 2 somewhere between 40 and 50 times. If the frequency is 1 in 11 then the average would be 45 times.

By your logic the other die would be a two on average 83 times though 500 decisions.

I'm used to collecting emperical evidence like this. I do it on video line games all the time where I don't have information on the frequency of line pays. So this is what I'm going to do. I'm going to take your suggestion. I just need to find a store that sells dice cups and dice. I will make this emperical study as free time permits. And I will publish my results here.

Alan, why don't you do the same and publish your results.