Over/Under 2.5 goals, explained properly
What the market measures, how it settles, and why predicting goals is different from deciding whether the current price is fair.
The line turns total goals into a binary event
Over 2.5 goals wins when the settled match total is three or more; Under 2.5 wins when it is two or fewer. The half-goal line removes the possibility of a push under the ordinary two-sided definition.
The relevant match period depends on the operator's published settlement rules. A data model and a market quote must refer to the same regulation-time definition; extra time, abandoned matches and awarded results cannot be assumed from a generic label.
A likely Over can still be a poor price
Suppose a model estimates Over 2.5 at 60%. Its mathematical break-even decimal price is about 1.67 before any uncertainty buffer. At 1.80, the model sees positive theoretical value; at 1.55, it does not—even though Over remains the model's more likely side.
This is why FORM/PRICE displays win probability and price quality separately. Probability answers how often; the price determines whether that frequency is being offered cheaply or expensively.
What a serious totals model must preserve
Goal rates vary across teams, leagues and time. A usable estimate needs a fixed information cutoff and a model version so later line-ups, results or market moves cannot leak into the original forecast.
- Exact competition and fixture identity.
- Regulation-time market definition.
- Feature and odds timestamps.
- Full probability, not only the chosen side.
- Complete settlement and correction policy.
Sources
Direct links are preserved so the editorial reasoning can be checked independently.