Match predictions

The model expects roughly 1.4 goals from Wolverhampton Wanderers and 0.6 from Blackburn Rovers — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Wolverhampton Wanderers the favourite at home at 56%, with the draw at 28%.

Both teams score in 37% of the model's outcomes, and the match clears 2.5 goals in 34%. The single most likely scoreline is 1–0, at just 17% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.3 goals from Bolton Wanderers and 1.1 from Preston North End — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Bolton Wanderers the favourite at home at 39%, with the draw at 29%.

Both teams score in 51% of the model's outcomes, and the match clears 2.5 goals in 44%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.0 goals from Charlton Athletic and 1.2 from Derby County — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Derby County the favourite away at 38%, with the draw at 31%.

Both teams score in 45% of the model's outcomes, and the match clears 2.5 goals in 37%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.1 goals from Bristol City and 1.1 from Millwall — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Millwall the favourite away at 35%, with the draw at 30%.

Both teams score in 47% of the model's outcomes, and the match clears 2.5 goals in 39%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.5 goals from Portsmouth and 1.3 from Queens Park Rangers — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Portsmouth the favourite at home at 42%, with the draw at 26%.

Both teams score in 58% of the model's outcomes, and the match clears 2.5 goals in 54%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.8 goals from Middlesbrough and 0.8 from Lincoln City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Middlesbrough the favourite at home at 59%, with the draw at 24%.

Both teams score in 48% of the model's outcomes, and the match clears 2.5 goals in 49%. The single most likely scoreline is 1–0, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.1 goals from Stoke City and 1.1 from Swansea City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Stoke City the favourite at home at 36%, with the draw at 30%.

Both teams score in 47% of the model's outcomes, and the match clears 2.5 goals in 39%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.5 goals from Norwich City and 1.0 from West Bromwich Albion — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Norwich City the favourite at home at 47%, with the draw at 28%.

Both teams score in 50% of the model's outcomes, and the match clears 2.5 goals in 45%. The single most likely scoreline is 1–1, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.5 goals from Sheffield United and 1.0 from Birmingham City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Sheffield United the favourite at home at 48%, with the draw at 27%.

Both teams score in 51% of the model's outcomes, and the match clears 2.5 goals in 47%. The single most likely scoreline is 1–1, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.1 goals from Watford and 1.8 from Southampton — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Southampton the favourite away at 52%, with the draw at 25%.

Both teams score in 57% of the model's outcomes, and the match clears 2.5 goals in 56%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.2 goals from Burnley and 0.5 from West Ham United — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Burnley the favourite at home at 52%, with the draw at 33%.

Both teams score in 27% of the model's outcomes, and the match clears 2.5 goals in 23%. The single most likely scoreline is 1–0, at just 22% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.3 goals from Cardiff City and 1.7 from Wrexham — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Wrexham the favourite away at 45%, with the draw at 25%.

Both teams score in 61% of the model's outcomes, and the match clears 2.5 goals in 58%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.4 goals from Millwall and 1.0 from Norwich City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Millwall the favourite at home at 44%, with the draw at 28%.

Both teams score in 50% of the model's outcomes, and the match clears 2.5 goals in 44%. The single most likely scoreline is 1–1, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.4 goals from Lincoln City and 1.2 from Portsmouth — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Lincoln City the favourite at home at 41%, with the draw at 28%.

Both teams score in 52% of the model's outcomes, and the match clears 2.5 goals in 47%. The single most likely scoreline is 1–1, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.4 goals from Birmingham City and 1.1 from Bristol City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Birmingham City the favourite at home at 43%, with the draw at 29%.

Both teams score in 49% of the model's outcomes, and the match clears 2.5 goals in 43%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.3 goals from Swansea City and 1.3 from Sheffield United — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Sheffield United the favourite away at 36%, with the draw at 28%.

Both teams score in 53% of the model's outcomes, and the match clears 2.5 goals in 47%. The single most likely scoreline is 1–1, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 2.0 goals from Southampton and 0.9 from Stoke City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Southampton the favourite at home at 64%, with the draw at 22%.

Both teams score in 51% of the model's outcomes, and the match clears 2.5 goals in 55%. The single most likely scoreline is 2–0, at just 11% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 0.9 goals from Preston North End and 1.4 from Wolverhampton Wanderers — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Wolverhampton Wanderers the favourite away at 47%, with the draw at 29%.

Both teams score in 46% of the model's outcomes, and the match clears 2.5 goals in 41%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.6 goals from Queens Park Rangers and 1.2 from Bolton Wanderers — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Queens Park Rangers the favourite at home at 46%, with the draw at 26%.

Both teams score in 56% of the model's outcomes, and the match clears 2.5 goals in 53%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.8 goals from Derby County and 0.9 from Cardiff City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Derby County the favourite at home at 57%, with the draw at 24%.

Both teams score in 51% of the model's outcomes, and the match clears 2.5 goals in 52%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 0.9 goals from Blackburn Rovers and 1.1 from Middlesbrough — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Middlesbrough the favourite away at 40%, with the draw at 32%.

Both teams score in 42% of the model's outcomes, and the match clears 2.5 goals in 34%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.9 goals from Wrexham and 1.1 from Watford — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Wrexham the favourite at home at 54%, with the draw at 24%.

Both teams score in 58% of the model's outcomes, and the match clears 2.5 goals in 57%. The single most likely scoreline is 1–1, at just 11% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 2.0 goals from West Ham United and 0.6 from Charlton Athletic — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes West Ham United the favourite at home at 71%, with the draw at 20%.

Both teams score in 39% of the model's outcomes, and the match clears 2.5 goals in 49%. The single most likely scoreline is 2–0, at just 15% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.0 goals from Deportivo Alavés and 0.8 from Getafe — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Deportivo Alavés the favourite at home at 39%, with the draw at 33%.

Both teams score in 33% of the model's outcomes, and the match clears 2.5 goals in 25%. The single most likely scoreline is 0–0, at just 18% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.2 goals from Sevilla and 1.1 from Rayo Vallecano — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Sevilla the favourite at home at 38%, with the draw at 28%.

Both teams score in 44% of the model's outcomes, and the match clears 2.5 goals in 38%. The single most likely scoreline is 1–1, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.2 goals from Racing Santander and 1.8 from Villarreal — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Villarreal the favourite away at 52%, with the draw at 23%.

Both teams score in 58% of the model's outcomes, and the match clears 2.5 goals in 58%. The single most likely scoreline is 1–1, at just 10% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.5 goals from Espanyol and 1.1 from Levante — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Espanyol the favourite at home at 45%, with the draw at 25%.

Both teams score in 52% of the model's outcomes, and the match clears 2.5 goals in 48%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.7 goals from Celta de Vigo and 1.0 from Osasuna — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Celta de Vigo the favourite at home at 53%, with the draw at 24%.

Both teams score in 52% of the model's outcomes, and the match clears 2.5 goals in 51%. The single most likely scoreline is 1–0, at just 11% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.4 goals from Deportivo La Coruña and 1.3 from Elche — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Deportivo La Coruña the favourite at home at 40%, with the draw at 25%.

Both teams score in 54% of the model's outcomes, and the match clears 2.5 goals in 50%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 2.2 goals from Atlético de Madrid and 0.7 from Málaga — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Atlético de Madrid the favourite at home at 70%, with the draw at 18%.

Both teams score in 46% of the model's outcomes, and the match clears 2.5 goals in 56%. The single most likely scoreline is 2–0, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.3 goals from Rayo Vallecano and 0.8 from Deportivo Alavés — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Rayo Vallecano the favourite at home at 48%, with the draw at 28%.

Both teams score in 41% of the model's outcomes, and the match clears 2.5 goals in 36%. The single most likely scoreline is 1–0, at just 16% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 2.0 goals from Real Betis and 1.1 from Real Sociedad — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Real Betis the favourite at home at 58%, with the draw at 21%.

Both teams score in 59% of the model's outcomes, and the match clears 2.5 goals in 61%. The single most likely scoreline is 2–1, at just 10% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.6 goals from Athletic Club and 1.0 from Sevilla — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Athletic Club the favourite at home at 53%, with the draw at 24%.

Both teams score in 49% of the model's outcomes, and the match clears 2.5 goals in 47%. The single most likely scoreline is 1–0, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.4 goals from Valencia and 1.3 from Celta de Vigo — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Valencia the favourite at home at 40%, with the draw at 25%.

Both teams score in 55% of the model's outcomes, and the match clears 2.5 goals in 52%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 0.9 goals from Espanyol and 1.8 from Real Madrid — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Real Madrid the favourite away at 57%, with the draw at 23%.

Both teams score in 50% of the model's outcomes, and the match clears 2.5 goals in 51%. The single most likely scoreline is 0–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 2.2 goals from Arsenal and 0.6 from Coventry City — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Arsenal the favourite at home at 74%, with the draw at 19%.

Both teams score in 41% of the model's outcomes, and the match clears 2.5 goals in 54%. The single most likely scoreline is 2–0, at just 15% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 1.2 goals from Hull City and 1.7 from Manchester United — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Manchester United the favourite away at 48%, with the draw at 26%.

Both teams score in 58% of the model's outcomes, and the match clears 2.5 goals in 55%. The single most likely scoreline is 1–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

Treat this one as provisional: at least one side changed division this summer. A promoted side enters at a deliberately below-average prior — history says that is where promoted teams land — while a relegated side keeps its old-league strength, translated down by a backtested league gap. Either way, real matches replace the guess within weeks.

The model expects roughly 0.9 goals from Ipswich Town and 1.7 from Sunderland — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Sunderland the favourite away at 54%, with the draw at 27%.

Both teams score in 49% of the model's outcomes, and the match clears 2.5 goals in 47%. The single most likely scoreline is 1–1, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.5 goals from Nottingham Forest and 1.1 from Leeds United — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Nottingham Forest the favourite at home at 45%, with the draw at 28%.

Both teams score in 54% of the model's outcomes, and the match clears 2.5 goals in 49%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.3 goals from Everton and 1.0 from Crystal Palace — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Everton the favourite at home at 42%, with the draw at 31%.

Both teams score in 46% of the model's outcomes, and the match clears 2.5 goals in 39%. The single most likely scoreline is 1–1, at just 15% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.9 goals from Brentford and 1.3 from Tottenham Hotspur — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Brentford the favourite at home at 52%, with the draw at 24%.

Both teams score in 62% of the model's outcomes, and the match clears 2.5 goals in 61%. The single most likely scoreline is 1–1, at just 11% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 0.9 goals from Udinese and 1.6 from Como — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Como the favourite away at 52%, with the draw at 26%.

Both teams score in 49% of the model's outcomes, and the match clears 2.5 goals in 47%. The single most likely scoreline is 0–1, at just 12% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 3.5 goals from Inter and 0.5 from Monza — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Inter the favourite at home at 90%, with the draw at 7%.

Both teams score in 41% of the model's outcomes, and the match clears 2.5 goals in 77%. The single most likely scoreline is 3–0, at just 13% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 1.1 goals from Parma and 1.1 from Cagliari — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Parma the favourite at home at 35%, with the draw at 31%.

Both teams score in 43% of the model's outcomes, and the match clears 2.5 goals in 35%. The single most likely scoreline is 1–1, at just 14% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.

The model expects roughly 0.8 goals from Genoa and 1.5 from Napoli — expected scoring rates from each side's attack and defence strengths, the league's home advantage, and form weighted toward the last year. That makes Napoli the favourite away at 54%, with the draw at 27%.

Both teams score in 43% of the model's outcomes, and the match clears 2.5 goals in 40%. The single most likely scoreline is 0–1, at just 15% — no scoreline in football is ever likely, which is why the expected goals and the full grid carry more information than any one score.