Bayesian updating is a disciplined way to change your prediction market estimate as new evidence arrives, moving neither too little nor too much.
Prior and Evidence
Your prior is the estimate before the news. The key question is how much more likely the news is if the outcome is true than if it is false.
Avoid Overreacting
Dramatic headlines often carry weak evidence. If the news would be almost as likely either way, your estimate should barely move.
Avoid Underreacting
Strong, verified evidence deserves a real update, even when it contradicts your earlier view.
Worked Example
Your prior is 30%. A report appears that you judge twice as likely if YES is true. Prior odds are 30:70; multiplying by 2 gives 60:70, about 46%. The market moves from 30¢ to 60¢, which looks larger than the evidence alone justifies, so you remain cautious.
Key Takeaways
- Start from an explicit prior.
- Ask how diagnostic the evidence really is.
- Update in proportion to evidence strength.
- Market moves are not automatically correct.