You've (Likely) Been Playing The Game Of Life Wrong

The video introduces the concept of normal distributions, where data clusters around an average, like human height or IQ. In contrast, it highlights power laws, where extreme events are far more likely than in a normal distribution, significantly skewing averages and making systems unpredictable. Pareto's discovery of power laws in income distribution in the late 1800s serves as a foundational example.

Three casino games illustrate the difference between normal distribution, log-normal distribution, and power laws. The first game (additive winnings) exemplifies a normal distribution, where consistent small wins lead to a predictable average. The second (multiplicative returns) shows a log-normal distribution, generating greater inequality and a 'long tail' of large events. The third game (St. Petersburg paradox) demonstrates a true power law, where the expected value is theoretically infinite, and extreme payouts are possible, though rare, making the average meaningless.

Power laws are characterized by an 'infinite standard deviation,' meaning the average constantly increases with more measurements due to the significant impact of extreme outliers. This concept is visualized with the analogy of Bill Gates in a room, where his wealth skews the average. This property highlights the unpredictable nature of systems governed by power laws.

The video explains that power laws often arise from the interplay of two underlying exponentials. The example of earthquakes is used: the frequency of earthquakes decreases exponentially with magnitude, while the destructive energy released increases exponentially with magnitude. When these two exponentials combine, they result in a power law distribution of energy release.

Power laws are intrinsically linked to fractal-like patterns and indicate a system in a 'critical state.' This is demonstrated by the behavior of a magnet at its Curie temperature, where magnetic domains exhibit self-similar patterns across all scales, indicating no inherent scale to the system. This critical state is characterized by maximal instability and unpredictability.

Some systems naturally evolve into a critical state, a phenomenon called self-organized criticality. This is exemplified by forest fires and earthquakes. In these systems, small, identical causes (e.g., a lightning strike or a small rock crumble) can lead to wildly different outcomes, from minor events to catastrophic ones, all following a power law distribution. The Yellowstone fires of 1988 are presented as a real-world example, highlighting how fire suppression policies can inadvertently lead to larger, more destructive events by preventing the system from releasing stress through smaller fires.

The sandpile model, a simple thought experiment, further illustrates self-organized criticality. Dropping grains of sand onto a pile eventually leads to avalanches of various sizes, with their distribution following a power law similar to earthquakes. This model demonstrates 'universality,' where diverse systems in a critical state exhibit similar macroscopic behavior, regardless of their microscopic details.

Power laws are prevalent in human systems, influencing city populations, stock prices, scientific citations, and even the number of deaths in wars. This implies that many real-world phenomena are characterized by a few extreme outcomes dominating the aggregate. Industries like venture capital and book publishing thrive on this principle, as a small number of massively successful ventures compensate for many failures.

The video concludes by emphasizing that understanding whether a domain is governed by a normal distribution or a power law is crucial for effective decision-making. In normal distribution environments (like running a restaurant), consistency and average performance are key. In power law environments (like venture capital), taking calculated risks and making many intelligent bets, even if most fail, can lead to disproportionately large successes. The internet's growth and the emergence of 'super-connectors' also follow a power law due to preferential attachment, where new entities are more likely to link to already popular ones.