BTTS predictions this weekend
Football predictions for btts, ranked from the highest model probability down.
51% is the highest BTTS Yes probability this weekend: São Paulo FC vs Santos FC. ProbaPredict's model rates 1 of 2 matches as more likely than not to see both teams score, from each side's chance of scoring at least once.
These are this weekend's football predictions for btts, ranked by the model's probability. A 70% pick still fails three times in ten, so read the percentage, not just the position. Our public grading record grows with every matchday.
Pick = the single most likely outcome among home/away win, Over/Under 1.5/2.5/3.5 and BTTS. Shading = probability.Darker teal means a higher model probability. Bold marks the most likely of home, draw and away. ✓ and ✕ show graded results after full time.
Times in UK time| Fair odds (Yes) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sat, 3 Oct 2026 | ||||||||||||
00:00 · Brasileirão Série A São Paulo SantosSão PaulovSantos | 51% | 49% | 1.97 | 42% | 28% | 30% | 46% | 51% | U3.5 75% | 72% | 25% | |
22:30 · Brasileirão Série A Mineiro BragantinoMineirovBragantino | 49% | 51% | 2.03 | 45% | 28% | 27% | 43% | 49% | U3.5 78% | 69% | 22% | |
How do our BTTS predictions work?
Both teams to score (BTTS) asks a single question: will the home side and the away side each score at least once in normal time, including stoppage time? The final result does not matter: 1-1, 3-2 and 2-1 all count as Yes, while 0-0, 1-0 and 4-0 count as No. Extra time and penalties are ignored.
Our model rates every team's attack and defence from recent results, adds a measured home advantage, and builds a grid of every scoreline from 0-0 to 10-10. BTTS Yes is one minus the chance the home team scores nothing, minus the chance the away team scores nothing, plus the chance of 0-0 (which was subtracted twice).
Fair odds are 100 divided by the percentage: 60% equals 1.67 and 45% equals 2.22, with no bookmaker margin. A 65% BTTS Yes still fails about one time in three. The table ranks matches from the highest BTTS Yes probability down; switch to BTTS No to rank the other way.
New to this? Read: What is BTTS? · Which league has the most goals?
These football predictions rank matches by probability while keeping the uncertainty visible.
More football predictions
BTTS FAQ
What does BTTS mean?
BTTS stands for both teams to score. It is Yes when the home and away teams each score at least once in normal time, and No otherwise. Extra time and penalties do not count.
Which matches are most likely to have both teams score this weekend?
By ProbaPredict's model, the highest BTTS Yes probabilities this weekend are São Paulo FC vs Santos FC (51%), CA Mineiro vs RB Bragantino (49%). Even the top pick can fail, so read the percentage.
How is the BTTS probability calculated?
From the model's scoreline grid: one minus the chance the home team scores nothing, minus the chance the away team scores nothing, plus the chance of 0-0.
Our BTTS probabilities currently don't beat the average-rate benchmark (always predicting the typical BTTS rate) in out-of-sample testing, so BTTS is excluded from Pick and ACCAs until they do.
See our btts accuracy record, including the back-test.
Frequently asked questions
- What does Over 2.5 goals mean?
- Over 2.5 goals means the match finishes with three or more goals in total, counting both teams. A 2-1 or 3-0 counts; a 1-1 or 2-0 does not.
- What does a 60% probability mean?
- If the model gives an outcome 60%, it expects that outcome in roughly 6 of every 10 similar matches. It will still not happen in about 4 of them, so a single result says little about the model.
- How do I convert a probability into odds?
- Divide 100 by the percentage. 50% becomes fair decimal odds of 2.00, 40% becomes 2.50 and 25% becomes 4.00. Fair odds carry no bookmaker margin.
- How accurate are you?
- Every pick is graded in public after full time, including the ones that lose. Hit rates, calibration and Brier scores are on our accuracy page. See accuracy
- Do you guarantee winners?
- No. These are model estimates of likelihood, not guarantees. Even well-calibrated probabilities are wrong a predictable share of the time.