The Unrealistically Realistic

Paths come in two kinds: the realistically unrealistic or the unrealistically realistic. The most realistic-looking paths are often the most unrealistic, while the most unrealistic-looking paths are often the most realistic.

The realistically unrealistic path is the one that appears safe, attainable, and socially approved. They are realistic in theory, but unrealistic in practice because everyone else wants them too. We assume that the more realistic a path is, the more attainable it must be. But this is only true in a world without competition.

In theory, competition should be self-correcting. When a path becomes overcrowded and harder to enter, fewer people should pursue it. But in reality, the opposite often happens. Competition does not reduce demand but further increases it. People’s perception of what’s realistic and attainable remains almost independent of the actual odds.

Roughly 54,000 people applied to Harvard’s Class of 2028. About 1,900 got in. You don’t apply to something you think is truly impossible. Every applicant thought they had some chance, however small. Over 52,000 of them were wrong.

For realistic paths, success is often disguised as merit but is over-determined by luck. In highly competitive environments like Ivy League college admissions, differences in capability become increasingly difficult to discern. The perfect test score and outstanding set of extracurriculars begin to matter less, while personal circumstances and luck begin to matter more.

Competition breeds competition. It is no surprise that 52% of the Harvard class flows into the same three familiar paths: finance, consulting, and technology. The path is extremely legible. Your career is planned from the beginning, and uncertainty is minimized from every angle. This is the safe, low-variance path.

The irony is that the people inside these tracks are some of the most statistically fluent in the world. Finance teaches them to think in terms of distributions, variance, and risk-reward optimizations. And they do exactly that. A prestigious, high-paying career with defined promotion tracks and little left-tail variance has an unmatched Sharpe ratio.

But you cannot reach the far right tail by minimizing variance and optimizing for risk/reward. The realistic path protects you from the disastrous outcome, but it also excludes you from the exceptional one. The hidden opportunity cost is not what could have gone wrong, it was what could have gone unimaginably right.

But truly unrealistic paths exist outside the domain of statistics. When the population size is one (N=1), statistical metrics no longer make sense. There is no distribution to compare against, and no reference class to calculate probability from. Instead, we call these paths impossible: only a crazy person would attempt them, and only the truly lucky can succeed. The unrealistic paths are mistaken for impossible, not because they violate the laws of physics, but because they violate social precedent.

To achieve the impossible, you cannot simply do what has been done before, because by definition, it has not been done before. There is no one to copy. The problem must be broken down and rebuilt from first principles. Chance gives way to causality. The question is no longer how likely it is, but what would have to be true for it to happen. History is shaped by events with a sample size of one.

From the outside, the unrealistic paths look like luck. From the inside, they actually require the most skill. Some of the most unrealistic paths are, in some sense, the most realistic because they are the least dependent on luck. Success no longer depends on whether the odds are favorable, but on sheer force of will.

The solution exists somewhere in the set of all possible actions. You just need to find it.

Pick the unrealistic path.