2 ms·

Sam how much randomness are you using in your selection processes? For example, are you randomly funding a percentage of teams?

• jedberg 11y ago
  Hah, I doubt they do that, but it would be interesting. Make it double blind. Pre-select a couple of teams for acceptance to YC but don't tell the interviewers until afterwards, and never tell the teams, so that the team still feels like they got in on their merit.
  • danieltillett 11y ago
    This is certainly an experiment that YC could afford to run. Going random is really the only way to know if your selection processes are extracting real value or not. The best way of doing this would be to pull the random companies out of the good, but not good enough to fund, pile. I am sure YC has little problem telling a good startup from a bad one, but I bet they have real problem picking the excellent from the good. Trying to pick the excellent out of the good is where you are most likely to run into unconscious biases.
    • ScottBurson 11y ago
      > I bet they have real problem picking the excellent from the good Yes; they've said as much.
      • danieltillett 11y ago
        Even more reason to start testing their selection processes. If you can’t do better than picking at random then don’t.