Field guide
Value betting
Expected value, model disagreement, and the difference between an edge estimate and a profitable fact.
8 min read · Updated August 2026
01
What value means
A value estimate exists when a model probability is higher than the break-even probability implied by the available price. It is a disagreement between estimates, not proof that the market is wrong.
02
Expected value
Expected value weights possible profit and loss by estimated probabilities. A positive number is only as reliable as the probability estimate, the current price, and the assumptions used. Small input errors can erase a small calculated edge.
03
Demand input quality
Check whether the line is current, the market is liquid, the teams are correctly mapped, and key information is present. Treat sparse or conflicting data as a reason for lower confidence.
04
Judge over a real sample
Short-term outcomes are noisy. Evaluate calibration, closing-line context, sport, market type, and sample size alongside win rate. Do not erase losses or promote a brief streak into a durable claim.