> Thankfully almost no one is the best in their field :) You probably can't afford the best, you probably don't need the best, you probably can't keep the best happy, you definitely don't have problems that need the best.

You don't need the single best, you just should aspire as high a concentration of n * x for n>1 (2x, 10x, ...) as your talent to raiase funding permits.

But don't forget it's not all about "best coding"; you want people best in different things (PM, UX, architecture, security, databases, algorithms, compilers, documentation, networking etc. - also depending on what the product is about) - a diverse team of rock stars.