Settlement from the score
| Regulation-time score | BTTS Yes | BTTS No | Reason |
|---|---|---|---|
| 0-0 | Loss | Win | Neither team scores |
| 1-0 | Loss | Win | Away scores zero |
| 0-2 | Loss | Win | Home scores zero |
| 1-1 | Win | Loss | Both teams score |
| 3-1 | Win | Loss | Both teams score |
The table assumes a regulation-time market. Betfair's current football rules provide one operator example in which default match settlement uses 90 minutes plus stoppage time unless stated otherwise.
BTTS is not Over 2.5
| Score | BTTS Yes | Over 2.5 |
|---|---|---|
| 1-1 | Win | Loss |
| 3-0 | Loss | Win |
| 2-1 | Win | Win |
| 1-0 | Loss | Loss |
BTTS depends on the distribution of goals between teams. Match totals depend only on the combined count. A total-goals forecast cannot be substituted without retaining team-level scoring states.
Illustrative independent-Poisson calculation
Suppose a deliberately simplified model uses home mean 1.6 and away mean 1.1. Under independent Poisson assumptions, illustrative zero-goal probabilities are:
- P(home scores 0) = 0.2019
- P(away scores 0) = 0.3329
- P(both score 0) = 0.0672 under the stated independent Poisson assumptions
The probability that both teams score is found by inclusion-exclusion:
P(BTTS Yes) = 1 - 0.2019 - 0.3329 + 0.0672 = 0.5324 under the cited Poisson construction
The OpenStax Poisson chapter defines the distribution and assumptions. The multiplication of the two zero-goal probabilities requires independence, a condition explained by OpenStax's independence chapter.
Dependence and model validation
Football scores can share match state and low-score dependence. Dixon and Coles proposed an adjustment for low-scoring association football results (Dixon and Coles, 1997). That model does not make 53.24% a real forecast for any fixture; the inputs above are illustrative and need chronological validation and calibration.
Price example
At an illustrative decimal price of 1.95 and model probability 0.5324:
EV = (0.5324 * 0.95) - (0.4676 * 1) = 0.03818
At probability 0.50, the same price has negative expected value:
EV = (0.50 * 0.95) - (0.50 * 1) = -0.025
The calculation is sensitive to the team scoring probabilities, dependence treatment, accepted price, and settlement period.
Checks before analysis
- Confirm regulation time, half, or another period.
- Confirm whether extra time and shoot-outs are excluded.
- Preserve any operator rule for abandoned matches and awarded results.
- Estimate the joint scoring event, not two unrelated recent percentages.
- Check calibration on later matches and compare with a relevant baseline.
- Record the available and accepted price.
Validate the probability grid
Build BTTS from a complete home-goal by away-goal grid and verify it two ways. First, sum every cell in which both goal counts are at least one. Second, calculate one minus the home-zero row, minus the away-zero column, plus the shared 0-0 cell. The two results must agree apart from rounding.
Keep three boundary tests beside the model. If the away team has probability one of scoring zero, BTTS Yes must be zero. If both teams have probability one of scoring at least once, BTTS Yes must be one. Every BTTS Yes probability must also be no greater than either team's individual probability of scoring. A failed identity is a data, truncation, or aggregation error, not a price signal.
Record the score-grid tail, calibration period, and provider definition with the result. This makes the probability reproducible and reveals whether an apparently attractive BTTS price came from football evidence or a spreadsheet mapping mistake.
Next step
Use Btts Meaning for the next part of this topic.
Continue learning
- Next guide: Correct Score Betting
- Related guide: Double Chance Betting
- Definition: BTTS Meaning
Assumptions and limitations
The model and prices are illustrative. DraftKings' soccer rules and Betfair are current operator examples, not universal settlement. Own goals, awarded matches, VAR changes, and abandoned fixtures follow the accepted product rules.

