... it becomes very difficult to know where accuracy is coming from, or what's preventing it...
I have pursed accuracy in shooting, machining, and designing electronics.
I thought I had a feel for how to manage my errors.
Then in 2006 I designed an amplifier that was mass produced and tested by computer.
The auto test computer sent me test data.
The Gaussian distribution of errors in output was a million times more accurate than I expected.
My seat of the pants feel for it did not allow that 100 errors cancel a lot more than they add.
That questions the accuracy rituals of shooting that I abandoned because they did not seem to make a difference.
