Premier League Gameweek 8: Model Estimates for Chelsea–Spurs, Villa–City and Liverpool–Brighton
Gameweek 8 lands with three fixtures that each ask a different question of the numbers: can the league's most prolific attack travel to Anfield and keep its numbers intact? Has the only perfect start in the division survived a week of bad injury news? And is one league win enough to rewrite what seven games of data say about a side that has scored 6 goals? Here are our Dixon-Coles estimates for Aston Villa vs Manchester City, Chelsea vs Tottenham and Liverpool vs Brighton — model estimates, not betting tips — plus the form and team-news context behind them.
Aston Villa vs Manchester City — Sat 24 Oct, 12:30 BST, Villa Park
Manchester City arrive with the only perfect record in the division: seven wins from seven, 19 scored and 7 conceded, three points clear of Arsenal at the top. Recent league results include a 3-2 win at Ipswich, a 3-0 win at Liverpool and a 5-3 win over Sunderland. But the week has not been kind to Enzo Maresca: Antoine Semenyo is out for up to eight weeks with a knee injury, and Erling Haaland — the league's top scorer with 8 goals — is a major doubt with the foot and ankle problem he picked up against Ipswich. A 4-2 Champions League defeat at PSG on Tuesday adds fatigue to the list.
Aston Villa are 12th on 8 points (2W 2D 3L), but the direction of travel is positive: a 1-0 win at Newcastle last weekend, with record signing Johan Manzambi scoring the winner, and a 2-2 draw with Brentford the week before have them unbeaten in five in all competitions, including a 6-0 Champions League win over Fenerbahçe. Unai Emery's headache is at full-back: Matteo Ruggeri (hamstring) and Aaron Wan-Bissaka (hamstring) are out, Amadou Onana (ACL) is a long-term absentee, Ian Maatsen is nearing a return from an ankle injury but remains a doubt, and Manzambi is being monitored after a knock. Villa also host Viking FK in the Champions League on Wednesday night, so rotation is a live question.
| Model output | Villa win | Draw | Man City win |
|---|---|---|---|
| 1X2 | 7.5% | 15.3% | 77.1% |
| Over/Under 2.5 | 69.0% / 31.0% | — | — |
| BTTS | 52.8% | ||
City are the strongest favourites of the three previews despite the Haaland doubt: their attack strength (1.97, second only to Brighton's 2.39) against Villa's concession rate (1.57 goals per game) gives an expected-goals gap the model can't bridge (λH 0.81 vs λA 2.75). Note that the estimate assumes Haaland plays; a Haaland-less City is a materially weaker version of the side that produced those numbers.
Chelsea vs Tottenham Hotspur — Sat 24 Oct, 17:30 BST, Stamford Bridge
Chelsea are 4th on 13 points (4W 1D 2L), 17 scored and 13 conceded, and arrive in good nick: three wins in their last four league games — 2-0 at Everton (ending Everton's unbeaten league start), 5-1 over Bournemouth and a 4-3 thriller against Brighton. Xabi Alonso's injury list, however, is growing: Emiliano Martínez is out after fracturing his jaw and undergoing surgery, Reece James has had knee surgery, Jamie Gittens and Emmanuel Emegha are sidelined, and Pep Chavarria (who withdrew before the Everton match) and Malo Gusto (illness) are doubts. Jorrel Hato deputised impressively at left wing-back at Goodison.
Tottenham finally have a win to talk about. Monday's 3-1 victory over Coventry — James Maddison scoring on his first start in 17 months, Omar Marmoush with his first goal for the club, Ben Davies sealing it in stoppage time — was their first league win of the season after three draws (1-1 at Manchester United, 0-0 with Everton, 0-0 at Nottingham Forest) and three defeats. Roberto De Zerbi has confirmed that Tosin Adarabioyo and Andy Robertson, who took knocks against Coventry, will be available, and Dejan Kulusevski is back in the squad after missing all of last season. Spurs are 16th on 6 points (6 scored, 10 conceded) and have not won at Stamford Bridge since January 2017.
| Model output | Chelsea win | Draw | Spurs win |
|---|---|---|---|
| 1X2 | 73.5% | 16.2% | 10.3% |
| Over/Under 2.5 | 73.7% / 26.3% | — | — |
| BTTS | 61.1% | ||
The biggest home lean of the three previews: Chelsea's attack strength (1.77) against the league's 16th-placed defence gives λH 2.81 vs λA 1.03. Tottenham's win over Coventry barely dents the seven-game numbers — the model still sees a side that has scored 6 in 7 — while over 2.5 at 73.7% reads Chelsea's open games (their matches average 4.3 goals) more than Spurs' tight ones.
Liverpool vs Brighton & Hove Albion — Sun 25 Oct, 14:00 GMT, Anfield
This is the game of the weekend on the numbers. Brighton are 3rd on 16 points (5W 1D 1L) and are the league's top scorers with 23 goals in seven games. Their last two league games were a 5-0 demolition of Crystal Palace and a 2-0 win at Sunderland, and their only defeat was a 4-3 thriller at Chelsea back in August. Fabian Hürzeler may have Mats Wieffer nearing a return from the knee injury that has kept him out since that Chelsea game, though his availability for Sunday remains uncertain.
Liverpool are 5th on 12 points (3W 3D 1L), 9 scored and 8 conceded. Andoni Iraola's side have two wins, two draws and one defeat in their last five league games: a 2-1 win at Brentford with Alexander Isak scoring the winner, a 1-0 win over Bournemouth, a 2-2 draw at Newcastle and a 3-0 defeat to Manchester City. The absentee list is long — Dominik Szoboszlai (thigh), Jeremie Frimpong (thigh), Giovanni Leoni (ACL), Hugo Ekitike (Achilles) and Conor Bradley (knee) are all out, with Víctor Muñoz (knee) doubtful after limping off at Brentford. Ryan Gravenberch returns from a one-match suspension, and Liverpool face Villarreal in the Champions League on Tuesday night before Brighton arrive.
| Model output | Liverpool win | Draw | Brighton win |
|---|---|---|---|
| 1X2 | 8.5% | 18.0% | 73.5% |
| Over/Under 2.5 | 61.3% / 38.7% | — | — |
| BTTS | 48.8% | ||
The model's most provocative output of the week: it makes Brighton clear favourites at Anfield. Their attack strength (2.39, the best in the division) against Liverpool's mid-table defence gives λA 2.42 vs λH 0.74 — the largest expected-goals gap in either direction of the three previews. Two caveats: Brighton's 23 goals include a 5-0 and a 4-0 that flatter the average, and Liverpool's absentee list is already fully reflected nowhere in the numbers — the model scores what has happened, not who is missing.
How we computed this
Each team's attack and defence strength was estimated from their seven league games so far: goals scored and conceded per game, divided by the league-wide per-game average of 1.375 (the predictor's EPL preset uses 1.53 home / 1.22 away goals). Those strengths feed expected goals (λH, λA), and the Dixon-Coles correction (ρ = −0.13) adjusts the four low scorelines for dependence.
| Fixture | Home attack | Home defence | Away attack | Away defence | λH | λA |
|---|---|---|---|---|---|---|
| Villa–Man City | 0.73 | 1.14 | 1.97 | 0.73 | 0.81 | 2.75 |
| Chelsea–Spurs | 1.77 | 1.35 | 0.62 | 1.04 | 2.81 | 1.03 |
| Liverpool–Brighton | 0.94 | 0.83 | 2.39 | 0.52 | 0.74 | 2.42 |
For informational and educational purposes only. Nothing here is betting advice.