1. The three settlement outcomes
If the selected team wins, the market is settled as a win. If the selected team loses, it is settled as a loss. If the relevant match period ends level, the stake is returned and the market is void.
- Team wins → winning settlement
- Teams draw → stake returned
- Team loses → losing settlement
2. Handicap 0 uses the same core idea
Applying a handicap of zero does not change either team's score. Its role is to define the refund condition when the adjusted result is level.
The label can vary between interfaces, so the exact market title and rules should always be checked before interpreting the price.
3. Regulation time and overtime matter
Football markets normally use regulation time plus added time unless the rules say otherwise. Extra time and penalty shoot-outs are often excluded.
In hockey and other sports, a full-game winner market may include overtime while a regulation handicap does not. The same team name can therefore refer to different settlement conditions.
4. Why the price differs from a straight win
The refund on a draw reduces one losing scenario, so DNB usually has a shorter price than a straight match win for the same team. That difference is the price of protection, not free additional probability.
To compare the markets, examine the model's win, draw and loss probabilities rather than looking only at the favourite label.
5. A worked example
Suppose a model estimates 46% for home win, 30% for draw and 24% for away win. A home DNB view conditions the decision on the 70% of cases that are not draws: 46 / (46 + 24), or about 65.71% before considering market margin.
This conditional number helps explain the market but does not prove that the offered price is favourable or that the estimate is calibrated.
6. Common interpretation errors
DNB does not mean the team cannot lose, and a draw does not count as a win. It also should not be confused with double chance 1X, where a draw is a winning outcome rather than a refund.
Always retain the original event, period and settlement rule when a prediction is later verified.