Quote Originally Posted by tsbrownie View Post
Marv, You're a smart guy, it scares me when you say the first method is loaded with error sources. Assuming a round hole and round ball, what can go wrong?
My favored medium is metal, not wood, so I tend to evaluate measurement techniques with a machinist's eye.

First, the method depends on the existing hole being the same diameter throughout, no chamfering. Then there's the problem of accurately marking a string and finding and marking the midpoint of the marks with a scale and transferring it to the ball. It would be impossible to maintain a tolerance of 0.050, totally unsuitable in the machinist's world.

But you're working in wood where such a tolerance is totally acceptable. I should have noted that in my response. I apologize for not doing so.

Your second method is much better. The close-fitting shaft averages the hole location, avoiding the problems of edge finding and marking in the first method. Rolling on a marking medium auto-locates the equator without the need to fuss with making measurements with/on a string. The second ball has to be the same size as the workpiece but, again, that's an error source with which only a machinist would concern himself.

In fact, your second method comes close to one of the methods I would use if it was a precision metal sphere and I needed to minimize errors in locating the hole.