Re: What is exceptable brass weight deviation?
Sinclair recommends a variance of 1.5% as being an acceptable range, and that's one that should leave you with a workable amount of brass. This is the standard that I've held my match brass to for some time now, and I have no complaints with this tolerance.
Weigh them all, separate them by .1 grain increments and see how they fall on a bell curve. Discard those that fall outside the 1.5% range and keep those in the middle as your working brass. Should be good to go from there.