How to Think in Probabilities: A Trader’s Guide
Imagine a Polymarket contract trading at 15%, but you think the event is more like 30%. If you’re right, you make $0.85 per share. If you’re wrong, you lose $0.15. Should you buy?
Most people answer with their gut. That’s System 1 thinking — fast, intuitive, and often wrong. Probability thinking trains System 2: deliberate, analytical, and aware of bias.
Why Our Brains Struggle with Probabilities
Humans evolved to make fast decisions, not calculate precise odds. Your ancestors needed to decide quickly: Is that a predator? Eat this or not? In these situations, speed beat accuracy.
Today, that shortcut can mislead you. When you glance at a market and think “this feels right,” that’s System 1. The discipline of probability thinking is about slowing down, checking your assumptions, and updating your beliefs based on evidence.
Update Beliefs Like a Pro
Bayesian thinking is at the heart of rational probability assessment. Start with the base rate — how often does this kind of event actually happen? Then add new evidence gradually, asking how informative it really is.
Reframe percentages into real-world counts. Instead of thinking “15% chance,” imagine 15 out of 100 similar events. This simple mental shift improves intuition and reduces mistakes.
Biases That Cost You Money
- Availability: Recent or vivid events feel more likely than they really are. Markets spike after dramatic news, not because probabilities changed, but because the event is top of mind.
- Anchoring: The current market price strongly influences your estimate. Your adjustment may never reach the true probability.
- Overconfidence: Most people overestimate how likely they are correct. If you think you have a 90% edge, it’s probably closer to 60%.
- Confirmation Bias: Once you take a position, you notice information that supports it and ignore anything against it.
- Loss Aversion: You hold losers too long and sell winners too early, distorting rational decisions.
- Narrative Fallacy: Every event seems obvious in hindsight, which leads to overconfidence in predicting the next one.
Cheap longshots are often overpriced because humans overweight small probabilities — just like buying lottery tickets.
Lessons from Probability Puzzles
- Monty Hall: Every piece of evidence matters. The host’s deliberate choice concentrates probability. Apply this thinking when new, surprising data appears in a market.
- Birthday Problem: Risks compound across positions. Twenty “high confidence” bets aren’t independent — one failure can hurt your portfolio.
- Prosecutor’s Fallacy: Market prices are signals, not absolute truths. Extreme odds are systematically biased; a 1-cent contract doesn’t mean a 1% chance.
How Superforecasters Think
Superforecasters aren’t smarter than everyone else. They are disciplined.
- Start with the outside view — historical trends first, not gut instincts.
- Break big questions into smaller, answerable parts.
- Update predictions frequently, but in small increments.
- Actively look for evidence that challenges your view.
- Think in probabilities, not “yes or no.”
- Track every prediction and compare it to reality. Over time, your estimates become more accurate.
Practical Approach Before Every Trade
- Find the base rate. Start with history, not gut.
- Translate percentages into counts for better intuition.
- Update gradually with each new piece of evidence.
- Check for biases: anchoring, availability, confirmation, narrative, and loss aversion.
- Challenge your own view. Ask what would have to happen for the opposite outcome.
- Size positions according to your edge. Small edge? Small stake.
- Record every trade and learn from outcomes. Calibration improves intuition over time.
Bottom Line
Probability thinking is a skill, not a talent. Every prediction market trade is practice. Start with the base rate, update with evidence, check your biases, size bets wisely, and track results. Repeat.
Over time, your intuition becomes sharper, your edge grows, and your decisions reflect real probabilities instead of instinctive guesses.