Performance literacy
How to read sports betting results beyond win percentage.
A betting record can show a high win rate and still lose money. Read price, units risked, ROI, drawdown, sample size, and publication integrity together.
Win percentage ignores the cost of each bet
American odds determine how much a win earns and how often a selection must win to break even. A bettor laying -200 can win far more than half the time and still lose if the true win rate does not clear the high implied threshold. An underdog bettor can finish below 50 percent and still profit when wins pay enough. A transparent record therefore displays the published price for every selection. Overall win percentage can remain a useful description, but it should never replace profit measured from the actual prices.
Units make stakes comparable
A unit is a standardized measure of stake chosen by the bettor. Reporting in units helps compare records without revealing or assuming anyone's bankroll size. If one unit represents one percent of a personal bankroll, a 0.5-unit play risks half that amount and a 1.5-unit play risks one and a half times it. The convention must be explicit. Some records risk one unit to win a variable amount; others size the stake so a win returns one unit. Mixing those conventions changes reported profit. Winnor's public methodology states how risk and profit are calculated so the curve can be reconstructed.
ROI needs the correct denominator
Betting ROI is commonly calculated as net profit divided by total units risked. If a record earns 5 units after risking 100 units, ROI is 5 percent. Dividing by the number of bets or by starting bankroll answers a different question and should not be labeled the same way. Pending selections do not belong in settled profit or risk. Counting pending stakes in the denominator before outcomes are known can depress ROI; counting them as wins can inflate it. A clean report separates pending counts and reveals the as-of date.
Drawdown shows the path, not just the destination
Two strategies can finish with the same profit while creating very different experiences. Maximum drawdown measures the largest decline from a prior cumulative-profit peak to a later trough. It helps describe the severity of losing periods and the bankroll pressure a bettor would have faced. Drawdown is backward-looking and tends to increase as a record grows. It is not a promise about the worst future decline. Still, hiding the equity curve or reporting only the final profit removes important context from performance claims.
Check grading and duplicate rules
Pushes should return the stake and contribute zero profit. Voids and canceled events need a stated treatment. Duplicate picks can overstate exposure and sample size if the same publication is imported twice. A trustworthy methodology explains how unique events are identified and how corrections are recorded. The most defensible approach derives results from immutable publications plus append-only settlement events. That lets an auditor identify what was originally posted and distinguish it from the later grade.
Read the record in layers
Start with evidence quality, then examine economics, then consider uncertainty. A small positive sample with strong receipts is still a small sample. A large legacy record can provide context while remaining weaker proof of prospective timing. The labels should make that distinction obvious. Evidence: publication timing, price source, receipt integrity. Economics: prices, units risked, net units, and ROI. Path: equity curve, streaks, and maximum drawdown. Scope: sports, markets, date range, and settled sample. Uncertainty: pending picks, unavailable closes, and prior-standard history.