Mentions of the "80/20 principle" make me wince, because both 80 and 20 are stated with 2 significant figures too many. But the idea behind it--"there's a lot of variation in the effectiveness of activities you classify as 'work'"--is correct. If Mark does just the most promising tasks for each startup, he'll be doing pretty damn good.
Your comment is a good example of a HN user suggesting a simple heuristic for answering an important question, which I find annoying. Another example would be Paul Graham's advice to make a product that scratches your own itch. Your domain expertise should definitely be a factor when choosing between two startup ideas, but I could probably think of three other factors that are at least as important.
I guess I could live with the simple heuristics if they were accompanied by detailed justifications explaining why they're better than most conceivable complex heuristics, but they never seem to be. Simple heuristics are sometimes correct, but there are so many people spouting simple heuristics just because they like simplicity that simple heuristics have a large burden of proof to overcome, in my mind at least.
- Simpler principles make better mantras.
- Occam's razor--all else equal, you should assume the world is simple.
I can't think of anything else.
Both of those considerations seem weak compared to other factors.