May need a little more info on your setup. If you are counting on an electrical path through your spindle bearings, you could get flaky results. I recently got to inspect a bearing on a 600hp motor. Voltage was building up in the rotor, and the only path to ground was through the bearings. If there was contact all the time, the voltage would never build up. However, it would build to the point that it would arc through the lube, and create a pit on the race. High velocity edm machining!
Back to my point, try to remove as much of the unknowns from your electrical path by connecting a conductor from the probe to the table.