A fellow on Welding Web (I think he's in witness protection so I won't name him) posted a chart of tungsten shapes, and how they relate to arc width and penetration (or lack of it). The chart is without explanatory information. I don't know if AC, DC or if it matters.
It shows that a very blunt grind produces a pinpoint arc, while a needle like grind produces a wide arc with little penetration. I've been grinding to a needle point trying to focus the arc. It seems that was the wrong thing to do.
We have discussed why, My uninformed guess has to do with the surface current phenomenon and the fact that electrons seek to continue running straight. A current on a cylindrical object will be concentrated on the surface. This higher current will tend to steer the rest of the current. As the shoulder of the grind is farther away from the work piece arc will originate farther away, if from a long pointed tungsten. From a shorter point it will be easier for current to continue along its "current" path, ionizing, then arcing across the gap.

I'm guessing here. Is there merit to my theory, or are other forces at play?