But this is a continous rv so it also equals , but your data looks discrete.

Next you're supposed to count all the times X exceeds 2 in your sample, but one could also look at all the times X exceeds 1.

This is why it's good to use the in the normal approximation to the binomial.