I think he's referring to the distribution of VC-backed companies: since most will fail, the ones that win have to win really big.

If all of the winners in a fund get acquired too early (small wins instead of big wins) it could mean the fund is not actually successful because it's not enough to outweigh the losses.