This demonstrates a lack of understanding of venture returns. Actually Khosla, Andreessen, Benchmark and the rest of the top tier account for the vast majority of the outsize returns. The "average" VC you haven't heard of, and those are the ones who can't even get introduced to the startups that account for the majority of the return. The power law distribution for startup returns is widely acknowledged, and the top VC phenomenon is that power law at work.
"Adjusted for risk." Actually this particular part of what you say is particularly untrue. When people talk about risk, they also talk about reward. Typical public companies may only yield a few percentage points increase, or 2X at extreme best over a year, but startups can yield 100X to 1000X return. So even after adjusting for risk, VC at its best (given a: good deal flow and b: good ability to help / pick companies) absolutely can outperform the S&P500 / index funds. The asset class as a whole is terrible (understandably professional LPs have difficulty picking the right general partners to back) but the best VCs do just fine.
Professional VC's aren't throwing darts at a board. The best ones see great founders and great new businesses far in advance of the rest of Wall Street or mainstream business press.