“Long streaks are, and must be, a matter of extraordinary luck imposed on great skill.” — Stephen Jay Gould

Download the Adobe pdf report.

The big point? Consider this from Michael Mauboussin:

The central message is that across domains, long streaks typically indicate skill. And since humans have a hard time relating to all but the easiest probabilities, we often fail to see the significance of streaks.

Michael Mauboussin offers clarity on perception, probability, and skill. Download white paper above.

The Gould and Mauboussin observations are two sides of the same truth. Gould says long streaks require both extraordinary luck and great skill — luck alone cannot sustain a streak, and skill alone cannot guarantee one. Mauboussin adds that because human beings handle probability poorly, we systematically misread what long streaks mean. We attribute them to luck when they demonstrate skill, and we attribute mediocre results to bad luck when they reflect an absence of skill.

For traders, the probability failure is acute. The human brain is designed to detect patterns and assign meaning to sequences of events. A few consecutive winning trades feels like confirmation of a superior method. A few consecutive losing trades feels like evidence that the method has stopped working. Neither inference is statistically justified from a small sample. The correct inference requires knowing the probability distribution the system is expected to produce and evaluating whether the observed sequence is consistent with that distribution. Almost no retail traders perform this calculation. Most systematic traders know it is essential.

Mauboussin’s streak research applies directly to evaluating trading track records. A manager who has produced above-market returns for 10 consecutive years has demonstrated something that luck alone cannot explain. The probability of achieving that outcome by random chance is calculable and small. That streak is almost certainly evidence of genuine systematic edge. A manager who outperformed last year is telling you almost nothing about whether the performance reflects skill or luck, because a single good year is consistent with a wide range of underlying skill levels including zero skill with good fortune.

The Turtle experiment illustrates both points. Richard Dennis produced extraordinary long-run returns over a career that spanned decades and multiple market environments. That streak, by Gould’s formulation, required both great skill and extraordinary luck in the sense that the markets presented the conditions that allowed the skill to express itself. But the skill was real and documented. It was teachable. When Dennis taught the same systematic approach to 23 novices, a significant fraction of them produced the same streak characteristics, which is the statistical evidence that the underlying skill, not luck alone, was driving the results. When you can transfer the streak, you have identified the skill component.

The probability framework also explains why trend following’s win rate of 35-40% is not evidence of poor skill. The win rate is a single dimension of a two-dimensional probability distribution. The other dimension is the payoff ratio — the average size of wins relative to the average size of losses. A system with a 35% win rate and a 3:1 payoff ratio has a higher expected value than a system with a 60% win rate and a 1:1 payoff ratio. Human probability intuition focuses on win rate because it is visible and emotionally salient. Expected value, which combines both dimensions, is what actually determines whether a system is profitable. Great trend following traders know the odds because they understand expected value rather than just win rate.

Frequently Asked Questions

What does Gould mean by “extraordinary luck imposed on great skill”?

He means that long performance streaks require both elements simultaneously. Skill alone cannot guarantee a long streak because random variation will eventually produce adverse outcomes regardless of skill level. Luck alone cannot sustain a streak because without the underlying skill to maintain positive expected value, luck will eventually revert. The combination of genuine skill and favorable random variation is what produces the extended streaks that mark the great performers in any domain.

Why do humans misread streaks according to Mauboussin?

Because human probability intuition is calibrated for simple, direct probability assessments and fails for complex sequential probability calculations. Most people cannot intuitively calculate the probability of a 10-year outperformance streak occurring by chance. They therefore cannot correctly assess whether a given streak is evidence of skill or luck. This systematic probability failure causes investors to attribute skilled performance to luck and lucky performance to skill, producing systematic errors in manager selection and strategy evaluation.

How does understanding probability improve trading decisions?

By allowing the trader to evaluate whether observed outcomes are consistent with the system’s expected probability distribution rather than reacting emotionally to each individual result. A system with a 35% win rate will produce losing streaks of five, six, or seven trades in a row with calculable frequency. A trader who knows this will not abandon the system during a normal losing streak. A trader who does not know this will interpret the losing streak as evidence the system has stopped working, abandon it at the worst possible moment, and miss the winning streak that follows.

Trend Following Systems
Want to learn more and start trading trend following systems? Start here.