Outliers matter most

Many important objectives in life are achieved by repeatedly sampling from a partly unknown probability distribution over possible outcomes. We want to find an apartment, so we go to many different inspections. We do our research and develop some feeling for which ones are promising and which ones aren't, but the distribution evolvoes over time and the uncertainty is large. So we try repeatedly until we find an apartment that we actually like. The same goes for finding a co-founder or a great deal for a second hand car and so on.

Average results are easy to come by. But real success comes from hitting statistical outliers, the rare long-tail opportunities. My strategy has always been to maximise the chance of encountering those outliers.

If you want to roll a six, the key is to roll as often as possible. However, we usually have an additional constraint: limited time. So it comes down to focusing on three things. Maximising the time we have available to sample, minimising the cost for sampling and sampling as fast as possible. That's it.