A 95% market feels like a done deal. It looks like consensus. It trades like gravity. And then, without warning, it's sitting at 50/50 — and everyone pretends they saw it coming.
They didn't. Here's what actually happened.
The Anatomy of a Crowded Trade
Markets don't sprint to 95% because the underlying probability is 95%. They get there because of narrative momentum — a story so clean, so repeatable, that new money piles in not because of fresh analysis, but because the chart is pointing up. Recency bias does the heavy lifting. If nothing bad has happened recently, the crowd prices nothing bad into the future. The probability isn't a forecast anymore; it's a comfort level.
Overconfidence compounds this. Traders who bought at 70% and watched it climb to 90% feel vindicated. That feeling is not signal — it's noise. The position looks right because it's been right, not because the fundamentals changed. Meanwhile, the actual tail risk — the scenarios that would break the narrative — gets systematically underweighted. Nobody wants to be the person pricing in the 5% when the crowd is at 95%.
What Breaks the Spell
The collapse rarely comes from a single bombshell. It comes from information that the narrative couldn't absorb — a data point that doesn't fit, a delay that shouldn't have happened, a statement that contradicts the script. Suddenly the 95% traders are asking questions they should have asked at 70%. Liquidity dries up. The market reprices violently, not because the world changed dramatically, but because the crowd's confidence was always borrowed.
The 50/50 that follows isn't the truth either. It's panic pricing — the opposite bias wearing different clothes.
The real edge in prediction markets isn't finding the 95% and riding it. It's asking, at 95%, what the crowd has stopped bothering to imagine.