Liverpool vs Manchester City: prediction and stats
Premier LeagueRegular Season - 6Kick-off on October 10, 2026 at 02:00 PM (UTC)Anfield, Liverpool
The Footlab model makes Manchester City the favourite at 43%, against 25% for a draw and 31% for Liverpool. The most likely score is 1-1.
This model is measured at 53.9% 1X2 accuracy on a temporal hold-out, against 33.3% for random guessing. See the measured accuracy
Probabilities recomputed on Aug 23, 2026. The full match card is recomputed continuously until kick-off.
Who will win Liverpool vs Manchester City?
Manchester City starts as the favourite with a 163-point Elo edge (2,369 vs 2,206).
What is the recent form of both teams?
Over their last 15 matches in all competitions, Liverpool have 5 wins, 4 draws and 6 defeats; Manchester City have 10, 3 and 2.
The competition is shown for every match: in pre-season, several of these games are friendlies.
| Date | Opponent | Score | Competition |
|---|---|---|---|
| Aug 23, 2026 | Newcastle(away) | 2-2 | Premier League |
| Aug 16, 2026 | Como(home) | 0-0 | Friendlies Clubs |
| Aug 9, 2026 | Monaco(home) | 2-3 | Friendlies Clubs |
| Aug 2, 2026 | Leeds(home) | 2-4 | Premier League - Summer Series |
| Jul 29, 2026 | Wrexham(home) | 1-0 | Premier League - Summer Series |
| Date | Opponent | Score | Competition |
|---|---|---|---|
| Aug 23, 2026 | Bournemouth(home) | 2-1 | Premier League |
| Aug 16, 2026 | Arsenal(away) | 0-3 | Community Shield |
| Aug 9, 2026 | Atletico Madrid(home) | 3-1 | Friendlies Clubs |
| Aug 5, 2026 | K-League Stars(away) | 3-1 | Friendlies Clubs |
| Aug 1, 2026 | Inter(home) | 1-1 | Friendlies Clubs |
What does the head-to-head say?
Across the last 6 meetings: 2 Liverpool win(s), 1 draw(s), 3 Manchester City win(s).
| Date | Score | Competition |
|---|---|---|
| Apr 4, 2026 | Manchester City 4-0 Liverpool | FA Cup |
| Feb 8, 2026 | Liverpool 1-2 Manchester City | Premier League |
| Nov 9, 2025 | Manchester City 3-0 Liverpool | Premier League |
| Feb 23, 2025 | Manchester City 0-2 Liverpool | Premier League |
| Dec 1, 2024 | Liverpool 2-0 Manchester City | Premier League |
| Mar 10, 2024 | Liverpool 1-1 Manchester City | Premier League |
Last archived meeting: Manchester City vs Liverpool on Apr 4, 2026
How many goals should we expect?
Liverpool score 1.6 and concede 1.7 goals per match over their last 15; Manchester City score 1.9 and concede 1.0. The model gives a 58% probability of over 2.5 goals and 61% for both teams to score.
| Club | Goals scored / match | Goals conceded / match | Clean sheets | Both teams scored | Over 2.5 goals |
|---|---|---|---|---|---|
| Liverpool | 1.6 | 1.7 | 3 | 10/15 | 8/15 |
| Manchester City | 1.9 | 1.0 | 5 | 9/15 | 11/15 |
Over the last 15 matches in all competitions.
Where do both teams sit in the table?
| # | Club | P | W | D | L | GF | GA | Pts |
|---|---|---|---|---|---|---|---|---|
| 6 | Manchester City | 1 | 1 | 0 | 0 | 2 | 1 | 3 |
Verdict, probable XI and verified absences
The full match card recomputes this prediction continuously until kick-off: multi-market verdict, player-by-player probable XI and web-verified absences.
Open the full match card in the app
No odds, no affiliation: the method is published and the accuracy is measured.
How does Footlab compute this prediction?
The model combines both teams' Elo ratings (anchored by competition pool), their recent form weighted by opponent strength, home advantage and a Poisson matrix for scorelines. Its accuracy is measured continuously on the predictions actually published. Methodology and measured accuracy
Frequently asked questions
- Who is the favourite for Liverpool vs Manchester City?
- The Footlab model makes Manchester City the favourite at 43%, against 25% for a draw and 31% for Liverpool.
- Where and when is Liverpool vs Manchester City played?
- Liverpool vs Manchester City is played on October 10, 2026 at 02:00 PM (UTC) at Anfield, in the Premier League.
- Will there be over 2.5 goals?
- The Footlab model gives a 58% probability of at least three goals in this match.
- How reliable is this prediction?
- The Footlab model is measured at 53.9% correct 1X2 outcomes on a temporal hold-out, against 33.3% for random guessing. The breakdown by competition and Elo gap is published on the accuracy page.