There's an argument I've seen that founders should be on a thoroughly nonstandard vesting schedule, for exactly this reason. Slower vesting, a longer cliff, an event-based cliff, or some combination thereof.

It's a decent assumption that all successful, unpurchased startups take at least 3 years to hit it big. Given that, you could reasonably put a founder cliff at 2 years (with early vesting on buyout) as a show of faith.

Of course, the other argument is that if a founder leaves with 12.5% of your company something else is terribly wrong regardless.