According to the help file you use the following to determine an ideal step size for AF:

(in focus position - out focus position) x 2 then divide this number by the data steps and then subtract 1 from that number. I don’t get the point of subtracting 1 from a large number, such as I use for step size (130). What difference would 129 make over 130?

That documentation is helping you to get the step size, Let’s say that your infocus position is 5000 and your out of focus position (approximately 3x HFR over your best HFR) is 7000 and that you want to use 7 total samples:

(7000 - 5000) * 2 / (7 - 1) = 666.66666 steps per sample, or 670 steps.

So rather than defining the step size you should calculate that based on the number of samples you want to acquire. 7 or 9 is generally a good range, but keep it odd.