Narrative coherence is the silent enemy of accurate thinking. When a sequence of events forms a satisfying story — this caused that, which caused the other — the human mind treats the story as evidence, even when it's just the mind doing what it does best: making things feel inevitable after the fact. The philosopher and historian R.G. Collingwood, in his 1946 work *The Idea of History*, argued that understanding any event requires reconstructing the 'inside' — the intentions and reasoning of the agents involved — not just the 'outside' sequence of what happened. Most post-mortems, sprint retrospectives, and product reviews do the opposite: they trace the outside story backward until it sounds logical, then stop. What Collingwood's method demands, and what almost no one actually does, is to hold the narrative loosely enough to ask whether the agents at the time could even have seen the decision the way we're now framing it. The practical move: when your mental model of why something failed feels satisfying and complete, treat that feeling as a warning. Completeness is what a story feels like, not what reality looks like.

What is the last conclusion you reached that felt airtight — and what would you have to believe about the people involved for that conclusion to be wrong?

Drawing from Philosophy of History / Collingwoodian Hermeneutics — R.G. Collingwood (The Idea of History, 1946)