I hate to say this here, but msuster is the one VC out there who really knows how things work and talks straight -- and the one I'd really want to go to bat for

I hate to say this here, but msuster is the one VC out there who really knows how things work and talks straight -- and the one I'd really want to go to bat for me. There is no sugar coating or extra idealism, but he's not limited to just software stuff either. In fact, I'd argue that makes him a better investor since he understands other industries -- and many of the dynamics in software are no different than other places. To the specifics: I think when you are hacking on something large you need external motivation. Since the YCombinator model is to start with little upfront investment until you are "ramen profitable" you basically end up with a couple of people in a room hacking on something until it is presumably worth something (of course, valuations are notoriously difficult in software -- esp. social software). Since people don't get a lot done on their own, the purpose of a co-founder is for the motivational factor. If you were an uber-hacker (e.g. Drew Houston comes to mind) probably you don't need anyone else. You can build a functional prototype, get investment, hire, etc. There is also the problem of the "idea." Although there are lots of loosy goosy "idea" people, there are also a fair number of talented hackers without any practical ideas -- and you really need to have a practical idea to have a startup. While personally I think you are usually better focusing on a problem domain where you have experience and there is a clear need, there are certainly cases where a hacker needs a co-founder simply to provide that interface layer between code and real life. As an example, anyone who has been in the industry for awhile will note just how terrible at design many otherwise very talented programmers are. Anyways, I found this extraordinarily inspiring as a single founder YC reject who has been doing reasonably well w/o any YC support.