Re: Case sorting - how much variance do you allow?
I see our posts crossed, and I did not mean to infer any criticism. Rather, my intent is to introduce a smidge of caution about accepting any strong belief that what a shooter/handloader does can result in anything even nearing perfection.
It's a big world. I made my career out of reconciling computer programs with cold hard fact, and deriving some relatively reasonable generalized predictive outcomes.
On Wall Street, and that's not making the task and easier. For example, one of my programs looked at the overal trends of business at close of day, and made a prediction about how many additional personnel would need to be called in to work early to handle massively inflated numbers of Margin Calls.
Computer algorithms require data points in order to model reality. But the number of relevant data points usually far exceeds the amount of accurate data which can be gathered and/or can be inputted in time to render a useful output. The more accurate a computation, the more its result relates to history.
There are (at least) two basic types of computers; Analaogue and Digital.
Analogue computers are uniquely specific to their task (like a thermostat) and gather their own data on a contemporary basis; and can render contemporary results.
Digital computers are generic and versatile (like a calculator), but require some sort of additional interface/intervention between the gathering and the computation of data; and this both delays results and limits the practical aspects of relating those results to actual, comprehensive facts.
We are left with estimates. Some are good, none are actual reality.
In this particular context, what I'm really trying to say is that there are considerable limits to how much exquisite handlading diligence can equate to realistic marksmanship imprevements.
Greg
