See http://mockaroo.com/759da9d0
Here I use a list of 10 phone numbers with the sequential distribution to generate 200 rows of data. When you use the sequential distribution, it iterates through the list from start to finish and then repeats. Thus you'll get exactly 20 repeats of each phone number in the resulting dataset. If you chose the random distribution type, you would get roughly 20 repeats of each, but the exact number would be different each time. If you want to weight some numbers more than others, use a custom distribution.