Contingency vs Convergence Diagnostic
Tell whether an outcome was a fluke or fated before you draw the wrong lesson from it
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 85%
Borrowed from evolutionary biologists debating why species rise or fall, contingency explains outcomes driven by one-off, path-dependent accidents — the asteroid that killed the dinosaurs hit the one patch of gypsum-rich rock on Earth that produced a lethal toxic gas; a second earlier or later and most life continues as before. Convergence explains outcomes that happen more than once through independent paths because the underlying problem only has a few good solutions — the octopus eye and the human eye evolved the same structure 600 million years apart on completely separate evolutionary branches. Klaas argues both forces are constantly acting on human lives and societies, and that most people misdiagnose one for the other after the fact because we only ever observe the single timeline that happened and instinctively narrate it as if it had to happen that way.
Origin
Klaas says he borrowed the contingency/convergence pair directly from evolutionary biologists who use it to explain why species rise or fall, and applies it to how humans should interpret their own life trajectories and societies.
Core principles
- 01Contingency is the 'stuff happens' version of reality: path-dependent, one-off, could have gone any number of ways.
- 02Convergence is the 'it was going to happen anyway' version: independent paths arrive at the same solution because the underlying problem has one good answer.
- 03Both forces operate on every outcome simultaneously — the diagnostic question is which one dominates.
- 04Getting this wrong in either direction produces a bad lesson: over-crediting convergence breeds false certainty, over-crediting contingency breeds paralysis.
How to run it
- 1
Identify the outcome you're trying to explain
State precisely what event or result you're diagnosing before assigning a cause.
- 2
Search for a single point of failure
Check whether one small, specific detail — timing, a person, a location — plausibly would have changed the whole outcome if nudged.
Pro tip If you can name the one grain that would have flipped it, you're likely looking at contingency.
- 3
Search for independent repetition
Check whether the same outcome, structure, or solution has emerged more than once, via unrelated paths.
Pro tip Independent convergence is strong evidence the outcome was closer to inevitable.
- 4
Weigh the two signals against each other
Most real outcomes carry both contingent and convergent elements; decide which one is doing more explanatory work here.
Watch out Don't force a single-cause verdict when the honest answer is 'a mix.'
- 5
Resist the after-the-fact narrative pull
Consciously separate 'this is the story I've stitched together because it's the only thing that happened' from 'this was actually likely to happen.'
Watch out Your brain will always produce a coherent-sounding story regardless of which is true — that fluency is not evidence.
- 6
State your causal claim explicitly before acting on it
When using the outcome to plan or predict, say out loud whether you're treating the driver as contingent (unlikely to repeat) or convergent (likely to recur), since that changes what you should do next.
In the wild
The asteroid that ended the dinosaurs struck the one Yucatán site rich in gypsum, producing a toxic gas cloud that caused a second, deadlier die-off beyond the initial heat blast; a delay of even one second in the asteroid's path may have spared most species. Meanwhile, human and octopus eyes evolved nearly identical structures independently, 600 million years after their lineages split, because vision only has a few workable designs.
→ One event (extinction) is a textbook contingency case; the other (the eye) is a textbook convergence case, and Klaas uses the pair to teach listeners how to tell them apart.
Klaas uses the classic time-travel thought experiment — would you kill Baby Hitler — to argue the real question is causal, not moral: killing him might avert WWII, or, as in a novel by Stephen Fry that Klaas cites, could produce a more disciplined Nazi leader who wins the war by developing the atomic bomb first.
→ Because decades of intervening contingency stand between the act and the outcome, certainty about the consequence is unjustified even when the moral intuition feels obvious.
Common mistakes
Treating a contingent outcome as convergent
Assuming an election winner, a company's rise, or a historical event 'had to happen' erases how close a small detail came to flipping it, leading to false confidence in forecasts.
Treating a convergent outcome as pure luck
Dismissing a repeated pattern, like independent species evolving the same trait, as coincidence ignores real structural forces that make the outcome likely to recur.
Is it for you?
Best for
Analyzing why a specific outcome (an election, a company's rise, a historical event) happened, before extracting a strategic lesson from it.
Not ideal for
Situations with a large, verifiable sample size where you can just measure the actual base rate directly rather than reasoning from a single case.
From the transcript
“contingency in a nutshell is the [__] happens version of reality”
“sometimes things work and when they work they might happen more than once”
“contingent moments divert our trajectories forever and convergent moments sort of push us towards an outcome”
From the episode
Chaos Theory: The Hidden Force That Secretly Controls Your Life - Brian Klaas - #806
Brian Klaas