20 is barely enough. You really want 30-40. The problem is variance. If 70% fail and only 10% produce meaningful returns, you want >>10 so you have a reasona

20 is barely enough. You really want 30-40. The problem is variance. If 70% fail and only 10% produce meaningful returns, you want >>10 so you have a reasonable chance of getting average returns. It's possible to get lucky with 5-10 startup investments but the common result is "lost everything" with "broke even" a close second. This applies to working for startups as well....