Re: nuclear weapon detonation, the implied probability of 9% (I take it source is: https://polymarket.com/event/will-a-nuclear-weapon-detonate-by https://polyma

Re: nuclear weapon detonation, the implied probability of 9% (I take it source is: https://polymarket.com/event/will-a-nuclear-weapon-detonate-by https://polymarket.com/event/will-a-nuclear-weapon-detonate-...) might be factoring in the premium for the insurance that participants that predict a detonation are interested in acquiring. Let's say that the real odds of detonation are 1% and that everyone participating in the market knows and agrees to this. You would expect the implied probability that the market produces to be 1%. But in practice, a nuclear detonation would be a highly disruptive event where the impact is hard to assess. This creates an asymmetry of interests. If you want to protect yourself financially from such an event, you would pay a premium for it (which in a prediction market implies placing a higher bet on "detonation will happen" than a perfect gambler). If you want to protect yourself from the event _not_ happening, you would also do the same, but no one really does that other than speculators. Similarly, you can sell tornado insurance to a lot of people, but very few people are interested in insuring that a tornado _will_ happen (maybe concrete bunker architect studios?). So the underlying prediction market would skew towards overestimating the likelihood of tornados.