I tend to follow two ideas in my finances: (1) simplify when possible and (2) conduct my own testing when possible.
(1) In your case, the simplest idea would be to set the thermostat at 60 when you're asleep/not at home and 68 when you're up and about at home (68 is the same temp I also use btw). I would suspect that the cost difference between 55 and 60 would be minimal for such short periods. Also, the time required to warm up the house 18 degrees from 50-68 after you're home following a 12-hour shift would be longer than most folks would want to wait (and not very comfortable in the meantime). My wife and I expect the house to be up to our comfort level in about 30 minutes or less after we get home for the evening, so that only allows us to tip the dial back a few degrees while we're gone.
(2) You could start today and run your current method during January and compare the degree-days with your bill. Then do it a different way for February. This comparison would be specific to your climate, your structure type/insulation levels, your physical heating comfort level, and your method of heating.
Within a couple winter billing cycles, you'll know an exact answer specific to your situation. FWIW, this would also work for an early season test during the summer air conditioning season.