Calibration and Forecasting: How to Be Right More Often

Calibration is the match between your confidence and reality: if you're well-calibrated, the things you call 70% likely happen about 70% of the time. It's the measurable core of good judgment, and both elite forecasters and strong poker players train it relentlessly.

What good forecasters do

Research on superforecasting found the best predictors share habits anyone can copy:

• They assign specific probabilities, not vague words like "likely."
• They break big questions into smaller, answerable parts.
• They update incrementally as new evidence arrives — neither anchoring stubbornly nor overreacting.
• They keep score to see where they were over- or under-confident, and adjust.

These are the same habits that make a poker player accurate about ranges and equities.

Why calibration beats confidence

Loud confidence is rewarded socially but is often poorly calibrated. The skilled forecaster who says "60%" and is right 60% of the time has better judgment than the pundit who says "definitely" and is wrong a third of the time. Calibration is judgment you can actually trust and improve, because it's measurable.

Keeping score is the secret

You can't improve calibration without feedback. Write down predictions with probabilities, then check them against outcomes. Over time you'll see your biases — maybe you're overconfident on close calls, or too timid on strong evidence — and you can correct them. Poker players do this through honest hand review; forecasters do it with a prediction log.

The takeaway

Being right more often isn't about confidence — it's about calibration. Assign real probabilities, break problems down, update on evidence, and keep score. Judgment, like a poker read, gets sharper only when you measure it against reality.