Searching for Outliers

The gap between how people search for jobs, partners, and ideas, and how they should search for them, is one of the most consistently underexplored practical questions in popular writing. Most advice tells you to be patient, to improve your resume, to network — all of which are optimisation moves on individual samples. Kuhn's essay argues for something harder to hear: in a heavy-tailed world, the thing that matters most is not making each sample better but drawing more of them.nnThe first key argument is the statistical framing. A heavy-tailed distribution is one where extreme outliers are far more common, relative to the mean, than a normal bell curve would predict. Pareto's 80-20 rule is the classic version. In a world where the top 1% of outcomes (romantic partners, startup ideas, employees) are not just twice as good but ten or a hundred times better than the median, it becomes irrational to stop searching after finding something 'probably good'. You are far more likely to be at the 90th percentile and within reach of the 99th than to have already found the 99th. The practical implication is that most people's intuitions about 'good enough' are calibrated for normal distributions that don't apply.nnThe second move is the filtering insight, which Kuhn frames in terms of Y Combinator's counterintuitive approach to startup evaluation. YC mostly ignores idea quality because the best startup ideas historically sounded terrible — Airbnb, Dropbox, Reddit. Filtering aggressively for 'this idea sounds good' actively selects against the kind of outlier that produces the 15% of portfolio value that Airbnb represents. Applied to jobs and relationships: if your filters are designed to minimise the chance of a bad outcome, they may simultaneously be eliminating most of your chances at an exceptional one.nnWe picked this because it reframes one of the most common sources of adult stagnation — settling, in any domain, because the search is exhausting — as a structural problem with a structural solution. The question worth carrying is whether the thing you most care about finding is being searched for like a fruit at the supermarket when it should be searched for like a startup.