Discrepancy between empirical and mathematical expected probability in a Combination Problem

Question

I found an interesting discrepancy between my iterative result and the mathematical result for a combination problem.

The problem is:

I have a list of