Yes, that's probably a better formulation of the question.
I wonder if YC already tracks this, maybe informally? It would be a good means of testing their own selection processes. Then again, they probably focus more on the "false positives" (YC startups that did poorly) than "false negatives" (the ones that got away).
Startup returns usually follow a power law. That means a missed good investment is potentially hundred times worse than a accepted bad investment. A single good investment like Heroku, Dropbox or Airbnb can easily outweight all other investments combined.
Obviously they don't want to accept just anyone, that doesn't work for various signal, scale, attention and effectiveness reasons. They do however accept (positive) outliers without traditional credentials, which is the opposite of what a traditional company does. A traditional company is looking for safety, VCs and YC are looking for maximum leverage (not the typical financial two-fat-tails kind, the only-upsides kind that Nassim Taleb et al. likes).
As for your question, I can't imagine a reality where they don't track that sort of thing.
I wonder if YC already tracks this, maybe informally? It would be a good means of testing their own selection processes. Then again, they probably focus more on the "false positives" (YC startups that did poorly) than "false negatives" (the ones that got away).