I understand why a part to be knurled must be a diameter that meshes with the pitch of the knurling wheels. And I've done plenty of successful knurling. But I have found that the "right" diameter is often NOT the one the math recommends, and I've ended up finding it more by trial and error. I don't understand why. 1) I have measured my knurling wheels by rolling them on a piece of paper across, say, 10" - counting the impressions and dividing by 10. I admit that this procedure produces slightly different numbers each time I try it so i do it a few times and average. These never seem to be round numbers - I've always (perhaps mistakenly) assumed the wheels were some metric pitch. Is there a better way? 2) I use this formula in Excel: =ROUND(INT(A1*B1*PI())/B1/PI(),3) where A is the "goal" diameter and B is the LPI of the wheels. My most recent frustration arose trying to use an 11 LPI wheel set to knurl a piece of 1.250". The math says I should turn to 1.244 before knurling. Tried it, results were awful. 1.238 worked much better. What's the deal? Is my LPI measuring technique to blame? Is the formula wrong? Is there a better way?