While revisiting Lowenstein’s classic When Genius Failed: The Rise and Fall of Long-Term Capital Management, I found myself reading it through the lens of talent management. A perspective that revealed some surprisingly relevant insights (I’ll be sharing more on that in an upcoming article). One particular rabbit hole led me to Victor Haghani, one of LTCM’s founding partners. After LTCM, Haghani turned his attention to decision science and later co-authored a book that introduced me to the field of ergodicity economics. I wasn’t familiar with the concept at first (Disclaimer... still not an expert, far from it), but as I dug deeper, I discovered ideas that feel strikingly applicable to how we manage, develop, and invest in human talent over time.
Most talent strategies today are built around averages: average performance, average retention, and average engagement. But here's the problem... employees don't live in averages. They live in time and more specifically their experiences. And time, it turns out, it behaves very differently.
In traditional economics, it’s assumed that what works for the group will eventually work for the individual. But ergodicity economics shows that in systems where outcomes compound over time, the average result across people is not what any one person actually experiences. In other words, what’s optimal across a population isn’t always optimal over time for a single person.
Ergodicity economics, introduced by physicist and complexity theorist Ole Peters, builds on the insight that much of economic theory relies on a flawed assumption: that individuals face “ensemble” outcomes (averages across many possible parallel worlds), rather than the single path of time they experience. Careers, development and risk all unfold sequentially and often non-ergodically, where the time-average outcome can diverge drastically from the expectation value.
The term ergodic comes from statistical mechanics, where it describes systems in which time averages and ensemble averages are the same. But most of life and especially career growth is non-ergodic: a poor decision early in a career can have compounding negative effects that no amount of later “average” performance can undo.
Careers are a prime example of such systems. One missed opportunity, one bad fit, or one stagnant year can alter the arc of a career in ways that can’t simply be “averaged out.” Likewise, sustained investment, mentoring, and thoughtful role design can compound into exceptional growth.
The challenge is to stop designing programs around cross-sectional success stories and start investing in time-consistent development paths that build resilient, upward trajectories for individuals, especially in a world of uncertainty.
Because when it comes to human capital, it's not about what works “on average.” It’s about what works over time.