He can't answer this without dissing the remaining groups he did not choose.

Exactly.

I thought that was exactly why blored was asking...

It is. The problem is that what pg says will have a real impact on y combinator companies.

Even if we disregard its effect on cliques, and even if you disregard the morale issues that could result from the mentor saying "you're not as fun as these other guys," relationships with investors and end-users could be harmed. I gather that's not the sort of thing you should carelessly mess around with.

Here's my bets (maybe I'll corner pg one day and find out if I'm right):

1. Loopt (this is the only one I'm sure is on the list)

2. reddit (news.yc is rather overwhelming evidence)

I'm on to pure speculation at this point...but I'll take a stab:

3. Zenter (the only thing I think makes this possibly not as likely is that they sold so quickly...I think pg might like to shoot for something bigger and more world-changing this time around, since he already has plenty of money...maybe Weebly or Buxfer would be number three in that case, while Parakey probably gets disqualified for the similarly early exit)

But, I obviously can't speak for pg, and I'm just making stuff up based on limited knowledge. Take it as you will.

This list is more or less identical to YC's most successful companies.

How about another question. Which companies were furthest from your expectations, e.g. failed when you expected them to succeed. I suppose you just don't fund companies you don't expect to succeed, so every failure disappoints.

I say: - Loopt - Zenter - One of this session's companies (probably those English guys that were looking to hire at the beginning of the session)