It would hurt founders' perceptions but wouldn't actually decrease any individual founder's chances of success. YC would just accept a bunch of people who aren't likely to get funded.
What would happen if you split up the batches into two groups, YC Classic and YC Black Swan, and placed founders into the groups post-interview? People who were placed in YC Classic could continue to have the expectation of ~100% demo day success they do now, while the people placed in YC Black Swan would be told "I find your idea interesting and we'll let you in but don't expect funding after demo day." People could take the Black Swan offers as rejections if they wanted.
There are a bunch of problems with this approach, but I wouldn't be totally surprised if way more than 30% of YC Black Swan was funded.
The problem is you are applying post-hoc analysis.
While a theme was that the outsized returns can come from ideas which sound bad, it doesn't necessarily say that all outsized returns come from bad-sounding ideas.
Not only that, but I think reading the essay will show that the outsized returns are not known until several years after demo day.
Sometimes bad ideas are just bad ideas. In the Venn intersection between bad-sounding-ideas and good-ideas - the 'bad sounding and not good' is a much bigger area.
To me, the entire essay is about making sure that institutionally, the bad-sounding-but-ultimately-good ideas are not left out. Trying to further identify and silo them at application stage would be even more fraught.
It would hurt founders' perceptions but wouldn't actually decrease any individual founder's chances of success.
In a world where potential founders are making a choice between YC or doing the startup anyway without YC, sure. But the world is more complicated than that, and I'm sure there are plenty of potential founders whose decisions are informed by what they hear from YC.
[...] split up the batches into two groups, YC Classic and YC Black Swan [...] People could take the Black Swan offers as rejections if they wanted.
This sounds like a very good model, in that it would separate the "is this something YC wants to invest in?" axis from the "does Paul Graham think that I'll succeed?" axis.
I love this idea, because it addresses the two fundamental issues at play here: the social and the financial. Considering YC as a single entity, the optimal funding strategy must take into account both the power law on returns and the prestige of the program. If YC loses its place as the most prominent and well-respected startup incubator, the Dropboxes and Airbnbs of the future will either forgo the application or suffer from lack of investor interest. If YC limits itself to only those companies that have an extremely high chance of getting funded, the outliers will never find a way in.
Let's say that p_accept is the probability that YC accepts the founders of the next Dropbox. YC itself cannot optimize for p_accept because of the factors mentioned above: instead, it has to optimize for p_apply * p_accept * p_fund, the product of the chances that those golden founders will apply to the program, be accepted, and find the funding they need to grow and thrive.
With the hypothetical YC / YC Black Swan split, the original YC can optimize for p_apply * p_fund, and YC Black Swan can optimize for p_accept. Not only that, but since all Black Swan candidates would have started as applicants to YC, Black Swan's p_apply would equal that of the original YC. Numerate investors with the same sense of the power law as pg would take care of Black Swan's p_fund.
Thus, all the prestige, cultural appeal, and midsize exits would derive from the original YC, but all the power-law returns would emerge from Black Swan.
Khosla Ventures actually has two separate kinds of funds set up for such funding -- KV Seed for science experiments, and a larger fund for more classic investments.
The main difference in branding for KV's funds is really only to the LPs, though they do mention it to founders on their website. They want to cover themselves in case the seed funding both doesn't return anything and looked to be imprudent in retrospect. Luckily I think they're doing well.
Vinod is actually quite explicit about trying to find black swans in his pitches :-)
Interesting, but am not sure it'll work from a marketing/founder-motivation perspective. There may be a segregation VCs can see, but it is not clear what effects it'll have on the company/founder's future efforts/commitment to the startup business.
The group-think effects are to be considered, before adopting this model. Am obviously not predicting anything, except perhaps this move might make the Black swans even more unpredictable.i.e: if you look at the Tableau of payoffs here(http://edge.org/conversation/the-fourth-quadrant-a-map-of-th...). this move can push investments from complex payoffs to very complex payoffs easily.
Personally, i am likely to even shoot for the YC Classic group deliberately or in rare optimistic moments, the Black Swan group.
If you could pull that off without discrimination that would be an awesome idea. It may cause somewhat of a North/South Korea kind of thing but if it can be done it would be a great data collection exercise that investors, founders can really learn from.
Would be interesting to learn why the black swans have done better then the normal batch, or why the normal batch did better or why it was a 50/50 split.
You might even learn that because a startup is considered as a black swan founders in that group work harder and thus have a higher success rate. This would really put into light debates about how much intelligence vs hard work may effect success rates in the world of startups.
What would happen if you split up the batches into two groups, YC Classic and YC Black Swan, and placed founders into the groups post-interview? People who were placed in YC Classic could continue to have the expectation of ~100% demo day success they do now, while the people placed in YC Black Swan would be told "I find your idea interesting and we'll let you in but don't expect funding after demo day." People could take the Black Swan offers as rejections if they wanted.
There are a bunch of problems with this approach, but I wouldn't be totally surprised if way more than 30% of YC Black Swan was funded.