Hi Folks,
Some of you know me, some don't, but I'll opinionate fer a spell here.
Sunnydaze is dead on correct that you must align with the cantilever, very few are in perfect alignment with the cartridge body. I don't agree with the pin hole as I think that can skew the cantilever's apparent alignment, I prefer a very smooth surface with incredibly tiny down-force, just enough to make the diamond touch the surface, but that's personal preference.
Wayne, for the technically inclined you simply must, immediately, I mean now, download Conrad Hoffman's two point arc protractor alignment generator over at that other vinyl site that will go forever unmentioned in my posts. It's so much fun it's ridiculous. Even if you don't use the protractor you generate, just to see the different null points and offsets instantly generated for IEC, DIN and your own personally chosen inner run out points, and Loefgren, Baerwald, and Stevenson alignments all at your fingertips is invaluable to the terminally tweaky.
Which leads to my last point (thankfully I'm sure) that the null points of Loefgren, Baerwald, or Stevenson are completely relative to the inner and outer lead in and run out points chosen. These are listed in the usual calculations, but are no means static, fixed or predetermined. There are standards for DIN (Japanese arms usually), IEC which is what is usually quoted, or you can decide on your own, the last of which makes the most sense to me. Modern record run out distances are way different than classical IEC standard 60.3mm, much closer to 70 or even 75mm. You can achieve dramatically lower distortion levels across the record surface by altering the IEC standard in your calculations.