Yiannis1970: Cups and grids are always the same. It's the angle maybe that gives a faulty impression but they are always the same. Note: The grids are square shaped eventhough if they seem rectangular.
But, the grids aren't always the same size... at least not in the NV.
The way I figure breaks in the NV is by measuring the distance with the transparent circle at the bottom of the pointer. That circle is normally the size of the hole, and if you divide that circle in half, there are 12 half-circles from the center of the hole to the edge of the grid. So if I have a putt where I need to move the pointer 44%, that equals 5.3 of those half-circles measuring out from the hole toward the grid line.
If the angle (like you say) is right, the transparent circle is the same size as the hole and the measurement is precise and there's no problem. But sometimes the hole is about 1/8 larger than the transparent circle and the edge of the grid line is reduced to about 10.5 half-circles from the center of the hole using the transparent half-circle. If I use the size of the hole (one half at a time) to measure, rather than the transparent circle, the edge of the grid should still be 12 of those half-holes from the center of the cup, but instead it's 10 of the larger half-holes, not 12.
There's also an angle where the hole is smaller than the transparent circle, so that throws things off, too.
When the transparent circle is the same size as the hole, and moving the marker 6 half-circles places the grid line in the center of the hole, and 12 half-circles places it at the edge of the grid, the measurement is 100% accurate and putts drop every time (assuming my dot count is right). But I'm struggling to figure out the measurement when the hole/grid (or angle) is different.
I've tried ignoring the size of the holes and trusting my measurements using the standard half-circles, but that isn't accurate. I've tried measuring using the new size of the hole instead of the circle, but that isn't always accurate. Even going by the % of break isn't as accurate.
Any ideas of what else to try?