"[2] When I say the summer 2012 batch was broken, I mean it felt to the partners that something was wrong. Things weren't yet so broken that the startups had a

"[2] When I say the summer 2012 batch was broken, I mean it felt to the partners that something was wrong. Things weren't yet so broken that the startups had a worse experience. In fact that batch did unusually well." When something unusual happens (every partner needs to keep track of more startups) and the result is unexpectedly more success instead of less, doesn't that suggest that the partners were counterintuitively wrong about feeling wrong? An experiment might be in order.