Geo Hahn said:
Did you mean 'points gap'?
Nope, plug gap. Ideally it doesn't jump the point gap at all (which is the main function of the condenser).
The theory at least is that it's easier to pry electrons from the center electrode of the plug (which is a special alloy that releases electrons easily) than from the (usually iron) ground electrode. Electrons are of course negative, so making the plug terminal negative means electrons jump from the center electrode to the ground electrode.
Some sources claim as much as 40% more voltage is required to fire the plug in the 'wrong' direction. Personally I think the difference is much smaller than that, but I've never actually tried to measure it so I don't really know.
But it's worth noting that many modern distributorless engines fire half of their plugs backwards all the time. I had an old Honda motorcycle that worked that way too, and it seemed to run OK. By using a single coil to fire 2 plugs in series (one on each side of the high tension winding), these "waste spark" systems get by with half as many coils as would be otherwise required. Eg, that 2-cylinder Honda had only one ignition coil, with two high tension terminals.