In "A", rows 1-1000, I generate random numbers using RAND(). I want to pick numbers between say 1-100.

In "B" for 1000 rows I have `=RAND()*(100-1)+1`

In "C" for 1000 rows I use the numbers that I generated in "A", so `=A1*(100-1)+1`

The averages in B and C are the same, as expected.

I only want to use the values for 20% of the time, so for each of the 1000 rows:

For B, I use `=IF(A1<0.2,B1,0)`

For C, I use `=IF(A1<0.2,C1,0)`

If I do this, the averages for each of the two columns are vastly different.

**Can anybody explain why?**

Both approaches are using random numbers. The first approach uses two different random numbers (from column "A" and the one used in the "B" column calculation).