The line about the experiment made me consider if "no ideas" is a symptom or a cause.
Borrowing from epidemiology, I think "no ideas" might be an effect modifier. That is, it is just an intermediate step when going from the exposing factor to the outcome (the outcome, we know, is failing start-ups). It would be interesting to know what the real proximal cause is, or at least a really good proxy for it.
Maybe YCombinator could benefit from the input of a biostatistician or similar statistical expert to design some cohort studies on start ups.
Borrowing from epidemiology, I think "no ideas" might be an effect modifier. That is, it is just an intermediate step when going from the exposing factor to the outcome (the outcome, we know, is failing start-ups). It would be interesting to know what the real proximal cause is, or at least a really good proxy for it.
Maybe YCombinator could benefit from the input of a biostatistician or similar statistical expert to design some cohort studies on start ups.