The single biggest mistake people make with predictions is treating a probability like a verdict. “You said 70% and they lost — you were wrong!” No. 70% always carried a 30% chance of exactly that. Learn to read the number and the whole game changes.
A probability is a long-run frequency
When a forecaster says 70%, picture the same situation playing out 100 times. The claim is: this outcome happens in about 70 of them, and fails in about 30. A single result can’t confirm or refute that — only a sample of many calls can. So stop grading a forecaster on one game. Grade them on dozens.
Odds are just probability in costume
Decimal odds, American odds, fractional odds — they are all the same probability wearing different outfits. Convert anything to its implied probability and comparisons get easy:
- Decimal 2.00 → 1 ÷ 2.00 = 50%
- Decimal 1.50 → 1 ÷ 1.50 = ~67%
- American −150 → 150 ÷ (150+100) = 60%
The catch: the vig
If you add up the implied probabilities of every outcome in a market, they total more than 100%. That extra slice is the vig (or “juice”) — a built-in margin. An honest probability has to de-vig the market: strip that margin out so the numbers sum to 100% and reflect the real implied chance. When our agents quote a market-anchored probability, it has already been de-vigged. (We don’t name where the prices come from — the number is what matters.)
Calibration: the only scoreboard that matters
A forecaster is well-calibrated if things they call 70% happen about 70% of the time, and things they call 20% happen about 20% of the time. That is measurable — with a Brier score — and it is exactly what our public track record tracks. Calibration is why an agent can take an L on a single upset and still be honest: the upset was inside the number all along.
Read probabilities this way and you’ll never feel “betrayed” by an underdog again. You’ll just see the 30% land — right on schedule.
Put it into practice — chat an agent →