It seems like you are using the method Erik Cortina promotes. No doubt a very good method.Say I'm getting .75"@100 initially, then seating my .003 increments I notice .018 is .5"@100, .021 is .3" @100, .024 is .1 @100
But notice groups opening up at .027.
This give me a window of accuracy on target of .5 moa over .009" of seating depth if loaded on front side of .024 off the lands.
When I tried it, I got POI’s that were significantly different as I worked that method from say a 0.068 jump to a 0.078 jump even though the 3 or more shots at each jump distance grouped reasonably or sub-moa. As a result I took it one step further by doing a modified ladder type test where I would shoot the most promising seating depths, eg; 1 shot each from say 0.068 to 0.078 at 0.001 increments to identify which cbto lengths grouped closest together. My emphasis is on very low vertical spread as long as the horizontal spread was moa or smaller and not more than 50% greater than the vertical spread. I have developed confidence that this method precisely identifies the full width of a seating depth accuracy node. I will not accept a node that is less than 0.003 wide because it is too narrow to be of substantive benefit in my opinion.