Both Teams to Score (BTTS) Explained — How It Works & How to Predict It
BTTS is one of the simplest and most popular football betting markets — but predicting it accurately requires understanding attacking strength, defensive weaknesses, and the statistical models behind the numbers. Here's everything you need to know.
What Is BTTS?
Both Teams to Score (BTTS) is a betting market where you predict whether both teams will score at least one goal during the match. It doesn't matter who wins, what the final score is, or when the goals are scored — only whether each team finds the net at least once.
The market has two options:
- BTTS Yes — Both teams score at least one goal (e.g. 1-1, 2-1, 3-2)
- BTTS No — At least one team fails to score (e.g. 1-0, 0-0, 3-0)
Own goals count. Extra time and penalties do not — only regular time (90 minutes + stoppage time) matters.
How BTTS Odds Work
Because BTTS is a binary market (Yes or No), the odds are structured similarly to Over/Under markets. Typical ranges you'll see:
| Selection | Typical Odds Range | Implied Probability |
|---|---|---|
| BTTS Yes (attacking teams) | 1.50 – 1.70 | 59% – 67% |
| BTTS Yes (average match) | 1.70 – 1.90 | 53% – 59% |
| BTTS Yes (defensive teams) | 2.00 – 2.40 | 42% – 50% |
| BTTS No (defensive match) | 1.50 – 1.70 | 59% – 67% |
| BTTS No (open match) | 2.00 – 2.30 | 43% – 50% |
The bookmaker margin is typically 5-8% on BTTS markets, meaning the combined implied probabilities of Yes and No will add up to roughly 105-108%.
Key Factors That Influence BTTS
Attacking Strength
Teams with high expected goals (xG) per match are more likely to score regardless of opponent. Look at goals scored per game over the last 10-15 matches, not just the season average — recent form matters more than early-season data.
Defensive Weakness
A team that concedes frequently makes BTTS Yes more likely for their opponents. The key metric is goals conceded per game, especially away from home where defensive records tend to be worse.
League Patterns
Some leagues are structurally more open than others. The Bundesliga consistently produces higher BTTS rates than Serie A, for example. This isn't random — it reflects tactical culture, pressing intensity, and league-wide playing styles.
Home vs Away Splits
Teams often have dramatically different scoring profiles at home versus away. A team might score 2.1 goals per game at home but only 0.8 away. Always check the venue-specific stats, not just overall averages.
ExPrysm's feature engineering calculates separate home and away attacking/defensive ratings using Pi-ratings and rolling form windows. These venue-specific features feed directly into the prediction models.
How ExPrysm Predicts BTTS
ExPrysm doesn't predict BTTS directly as a classification problem. Instead, it models the expected goals for each team independently using Poisson regression, then derives BTTS probability mathematically.
Step 1: Estimate Expected Goals
The goals regression models (home_goals.cbm and away_goals.cbm) predict λ_home and λ_away — the expected number of goals for each team. These CatBoost models use 150+ features including Pi-ratings, ELO ratings, form indices, and head-to-head records.
Step 2: Calculate Individual Scoring Probabilities
Using the Poisson distribution, we calculate the probability that each team scores at least one goal:
P(team scores ≥ 1) = 1 − P(team scores 0) = 1 − e^(−λ)
If λ_home = 1.6, then P(home scores ≥ 1) = 1 − e^(−1.6) = 1 − 0.202 = 79.8%
If λ_away = 1.1, then P(away scores ≥ 1) = 1 − e^(−1.1) = 1 − 0.333 = 66.7%
Step 3: Combine for BTTS
Assuming independence (adjusted by Dixon-Coles correction for low-scoring outcomes):
P(BTTS Yes) = P(home ≥ 1) × P(away ≥ 1)
P(BTTS Yes) = 0.798 × 0.667 = 53.2%
The Dixon-Coles correction adjusts for the empirical observation that 0-0, 1-0, 0-1, and 1-1 scorelines occur at slightly different rates than a pure Poisson model predicts. This correction is particularly important for BTTS since it directly affects the probability of zero-goal outcomes.
BTTS Statistics by League
Historical BTTS Yes rates vary significantly across Europe's top leagues. These averages are based on multiple recent seasons:
| League | BTTS Yes % | Avg Goals/Match | Character |
|---|---|---|---|
| 🇩🇪 Bundesliga | ~55% | 3.1 | High-pressing, open play |
| 🇳🇱 Eredivisie | ~56% | 3.2 | Attacking philosophy |
| 🏴 Premier League | ~52% | 2.8 | Physical, end-to-end |
| 🇫🇷 Ligue 1 | ~48% | 2.6 | Mixed, PSG dominance |
| 🇪🇸 La Liga | ~45% | 2.5 | Possession-based, tactical |
| 🇮🇹 Serie A | ~46% | 2.6 | Defensively organized |
| 🇵🇹 Primeira Liga | ~50% | 2.7 | Competitive mid-table |
| 🇹🇷 Süper Lig | ~49% | 2.7 | Unpredictable, volatile |
ExPrysm covers 100+ leagues worldwide. League-specific patterns are captured through the model's training data and Pi-rating system, which adapts to each league's scoring tendencies.
Common BTTS Mistakes
1. Ignoring Defensive Form
Many bettors focus only on attacking stats when evaluating BTTS. But a team's defensive record is equally important — if one side has kept 5 clean sheets in their last 8 matches, BTTS Yes becomes significantly less likely regardless of the opponent's attack.
2. Chasing High Odds on BTTS Yes
When BTTS Yes is priced at 2.20+, there's usually a good reason — one or both teams have poor scoring records. High odds don't mean value. Value exists only when the true probability exceeds the implied probability from the odds.
3. Small Sample Sizes
Looking at only the last 3-4 matches can be misleading. A team might have faced three weak defences in a row, inflating their BTTS rate temporarily. ExPrysm uses rolling windows of 10-15 matches combined with season-long Pi-ratings to avoid this trap.
4. Ignoring Match Context
Cup finals, relegation battles, and dead rubbers produce very different BTTS profiles. A team fighting relegation may park the bus and play for 0-0, while a mid-table team with nothing to play for might open up. Context matters.
5. Treating BTTS as Independent from Other Markets
BTTS Yes is correlated with Over 2.5 Goals but they're not the same. A 2-0 result is Over 2.5 but BTTS No. A 1-1 result is BTTS Yes but Under 2.5. Understanding these overlaps helps build smarter combination bets.
Conclusion
BTTS is a straightforward market on the surface, but predicting it well requires understanding the interplay between attacking quality and defensive solidity on both sides. The Poisson-based approach — modelling each team's goal expectation independently and combining them — provides a principled, mathematically grounded way to estimate BTTS probability.
ExPrysm publishes BTTS predictions for every match it covers, derived from the same goals regression pipeline that powers Over/Under and Asian handicap markets. Check today's predictions on the Dashboard.
See today's BTTS predictions across 100+ leagues: Create a free account