Yesterday's BTTS.
24 predictions across 16 leagues. BTTS is the model's noisiest market — we publish the gap so you can see how often it lands. What do these numbers mean?
Football is an inherently variable sport. Both Teams To Score requires two independent scoring processes to line up, which makes it the noisiest market we display. Other markets we publish — outright home/away wins, double-chance (1X, X2), and over 2.5 — are materially more reliable. We label our predictions with honest probabilities, not certainties.
Hit rate over the last 90days, computed on finished matches that passed each market's on-page confidence threshold. All markets exclude leagues where the model is known to overpredict (women's, youth, tier-gap cups, and a handful of low-scoring lower divisions).
championship
1 prediction
1v2
Wolves
ligue 2
1 prediction
1v2
PAU
bundesliga
1 prediction
1v0
1. FC Köln2. bundesliga
2 predictions
1v1
1. FC Nürnberg
0v0
Hertha BSC3. liga
1 prediction
0v1
Stuttgart II
eredivisie
1 prediction
4v0
Excelsioreerste divisie
4 predictions
2v1
FC OSS
0v1
Vitesse
0v2
NAC Breda
4v4
Jong AZ
ekstraklasa
2 predictions
1v2
Raków Częstochowa
0v0
Wisla Krakow
Úrvalsdeild
1 prediction
3v1
FH hafnarfjordur
a-league
1 prediction
0v0
Western Sydney Wanderers
primera división
1 prediction
2v1
ADT
pro league
1 prediction
0v2
Al-Wasl FC
stars league
2 predictions
0v3
Al Wakrah
0v6
Al-Arabi SC
pro league
1 prediction
fnl
3 predictions
1v0
Hanácká
3v0
Dukla Praha
1v3
Vysočina Jihlava
premier division
1 prediction
2v1
WaterfordWhat does both teams to score mean?
Both Teams To Score (BTTS) is a single yes/no question about a football match: will bothsides find the net? “Yes” needs at least one goal from the home team and one from the away team. A 1-1, 2-1 or 3-2 all settle BTTS as yes; a 0-0, 1-0 or 3-0 settle it as no. The final score and the match winner are irrelevant — only whether each team scored at least once.
How these BTTS predictions are made
Each fixture's BTTS probability comes from a calibrated statistical model, not a tipster's opinion. Two Poisson regressors estimate how many goals the home and away side are each expected to score, using recent venue-specific form, Elo ratings, rest days, head-to-head history and recency-weighted shot and expected-goals data. We turn those two goal expectations into a full scoreline distribution, apply a Dixon-Coles correction for the low-scoring scorelines where goals are correlated, and read off the probability that both teams score.
A final isotonic calibration step pulls the raw number toward its honest long-run frequency, so a fixture shown at 70% lands BTTS close to 70% of the time rather than being quietly overconfident. On top of that, each league gets its own correction, refreshed every day from recent results: some divisions produce both-teams-to-score far more often than a general-purpose model expects, and some far less. The full pipeline is documented on the methodology page, and the live calibration gap per market is on the track record.
How to read this page
Predictions refresh every day. Each row is one fixture with its UK kick-off time, the two teams, and — once the match finishes — the actual score and a tick or cross showing whether BTTS landed. We only surface fixtures above the model's BTTS confidence threshold, grouped by competition.
A fixture also has to have a bookmaker price for BTTS, and our probability can't sit above the one that price implies. When the model rates BTTS more likely than the bookmaker does, the bookmaker has usually been right, so we leave those out. Prices move through the day, so a fixture near that line can appear or drop off before kick-off, and fixtures further ahead fill in as bookmakers price them.
Women's, youth, and a handful of tier-gap cup and structurally low-scoring competitions are excluded, because the model is known to overpredict BTTS there. Exact percentages on each fixture are reserved for signed-in users; the public view shows which fixtures qualified and how the market has performed lately. Looking for more than BTTS? The predictions page carries match-result, double-chance and over/under picks too.
Frequently asked questions
- What does BTTS mean in football?
- BTTS stands for Both Teams To Score. It settles “yes” when the home team and the away team each score at least one goal, and “no” otherwise. The final score and the match winner don't matter — only whether both sides found the net. A 1-1 or 2-1 is BTTS yes; a 0-0, 1-0 or 3-0 is BTTS no.
- How accurate are these BTTS predictions?
- Every BTTS prediction is a calibrated probability, fitted so that fixtures shown at (say) 70% land BTTS close to 70% of the time. The summary band at the top of this page reports the model's real hit rate over the last 90 days against what it predicted — the closer those two numbers, the better calibrated the model. You can see the full per-market breakdown on the track record page. BTTS is the noisiest market we publish, so we're deliberate about showing that gap rather than hiding it.
- Is BTTS harder to predict than the match result?
- Yes. BTTS needs two independent scoring processes to line up in the same match, which makes it noisier than markets like home/away win, double chance, or over/under goals. Those markets are materially more reliable, which is why we show their hit rates alongside BTTS for context.
- What's the difference between BTTS and over 2.5 goals?
- BTTS asks who scores; over 2.5 asks how many. They can disagree. A 1-1 is BTTS yes but under 2.5 goals (only two were scored). A 3-0 is over 2.5 goals but BTTS no (one team was shut out). They only coincide when there are at least three goals and both teams are among the scorers, like 2-1 or 2-2.
- How often are the predictions updated?
- Daily. New fixtures appear as each day's schedule fills in, and once a match finishes we show the final score and a tick or cross marking whether BTTS landed. Use the date pills near the top to move between recent days and tomorrow.
- Why is a match with a high BTTS chance missing?
- Two reasons beyond the confidence threshold. We only list a fixture once a bookmaker prices BTTS on it, and we leave it out when our probability is higher than the bookmaker's price implies — in our testing those were the picks that let us down, landing about half the time against the 60% we'd claimed. Matches further ahead often have no price yet, so they appear as bookmakers open their markets.
- Do you cover women's and youth matches for BTTS?
- No — women's and youth competitions are excluded from public BTTS predictions. The model's joint-Poisson assumption breaks down on lopsided fixtures, where a much stronger side regularly shuts out a weaker one, and it overpredicts BTTS in those competitions by a wide margin. We'd rather hide a market than show one we know is miscalibrated.