What Are Alternative Histories? (Nassim Taleb's Concept)
Alternative histories is Nassim Taleb's term for the full tree of outcomes that could have resulted from a decision — not just the one that did. Any realized outcome is one draw from a distribution of possible outcomes. To evaluate the decision correctly, you need to understand the distribution, not just the leaf you ended up on.
The Russian Roulette Case
The clearest illustration: someone plays Russian roulette with a $10 million reward, one bullet in six chambers, and survives. The visible outcome is good. By headline measures, the decision "worked."
But five of the six alternative histories end with a bullet. The decision drew from a distribution dominated by catastrophe. Evaluating the decision by the one visible outcome ignores the five outcomes that didn't happen but were equally possible — and far more probable.
The problem: the one visible outcome feels real; the five alternatives feel hypothetical. Emotionally, the good outcome swamps the alternatives. Analytically, the alternatives are what define whether the decision was sound.
The Practical Application
For any decision you're evaluating — yours or someone else's — Taleb asks you to mentally reconstruct the distribution at the moment of decision (before the outcome was known):
- What were the realistic possible outcomes?
- What was the probability distribution across those outcomes?
- Where does the observed outcome sit in that distribution?
A decision that looks brilliant because it produced a great outcome, but whose distribution was dominated by bad outcomes, was a bad decision. A decision that looks unremarkable because it produced a mediocre outcome, but whose distribution was centered on good outcomes, was a sound decision.
This evaluation refuses to let good outcomes validate bad processes. That's uncomfortable. It's also the only evaluation that generates useful information for future decisions.
The alternative histories concept is the foundation of the probabilistic worldview in Fooled by Randomness — the understanding that any realized path is one draw from a distribution, and the distribution is what deserves evaluation.
For the full framework, read Fooled by Randomness: How Luck Masquerades as Skill.