Liverpool
vs
Chelsea
League Cup prediction
Our model says: Liverpool 2–1 Chelsea
The model makes it 53% Liverpool, 23% draw, 24% Chelsea — fair odds of 1.89, 4.35 and 4.17. Fair odds are simply 100 ÷ the probability; they are the model's own number, not a bookmaker's price.
Most likely scoreline: 2-1 (9%).
Expected goals: Liverpool 2.22, Chelsea 1.33.
Both teams to score: 66%. Over 2.5 goals: 69%, under 2.5: 31%.
Model confidence: High.
Evenly matched: once home advantage is counted, neither side has a clear edge. In past matches rated this way the favourite won 41%, 28% were drawn and the underdog won 31%.
Published
Match centre: line-ups, live score and match stats →
Key facts
- Liverpool have won 2 of their last 5 games, most recently a 0-3 defeat at home to Manchester City.
- Chelsea have won 2 of their last 5 games, most recently a 5-1 win at home to Bournemouth.
- Last 6 meetings: Liverpool 3 wins to 2, 1 draw; the most recent, 9 May 2026, ended Liverpool 1–1 Chelsea.
Chance to go through
League Cup Round of 16 — a single match: the winner goes through.
- Liverpool57%
- Level after ninety minutes21%
- Chelsea22%
The model's chance of each side going through in ninety minutes. Level means extra time or penalties, which it does not forecast. A new figure: not yet checked against settled ties, so read it as an estimate, not a record.
More Carabao Cup predictions
Live win probability
From kick-offFrom kick-off, the split above updates here every few seconds as the model re-reads the score, the clock and any red cards.
Model insights: Liverpool vs Chelsea
Has the model earned this number?
It puts Liverpool at 53% here. Across 3,660 settled matches (as of 14 October 2026), 872 forecasts have landed in the 50–65% band; those outcomes actually happened 55.5% of the time, against an average forecast of 55.5%. In this range the two line up, so the figure above can be read literally.
Every settled prediction on this site is scored against an uninformed baseline, band by band, and published whether or not it flatters the model — see the full record.
Form going into Liverpool vs Chelsea
Liverpool
- L0-3Manchester City11 October 2026
- W1-0away to Bournemouth20 September 2026
- D0-0Fulham12 September 2026
- W2-1Atletico Madrid9 September 2026
- W2-0away to Ipswich4 September 2026
Chelsea
- W5-1Bournemouth10 October 2026
- L0-3away to Brentford18 September 2026
- D2-2Hull City12 September 2026
- L1-2away to Arsenal6 September 2026
- W4-3Brighton30 August 2026
What these results say
- Chelsea conceded in every one of these 5 matches.
- Every one of these 5 matches for Chelsea had three goals or more.
- Chelsea scored in 4 of these 5 matches.
Up to 5 matches before kick-off, most recent first. The model rates each side on its goals over this season and last, adjusted for the opposition, rather than on these results alone, so read them as context — how the ratings are built.
Head-to-head: Chelsea vs Liverpool
Chelsea and Liverpool have met 6 times in the records we hold, Liverpool has won 3 of them to Chelsea's 2, with 1 drawn.
The record
They drew 1–1 at Liverpool on 9 May 2026.
| Club | W | D | L | GF | GA |
|---|---|---|---|---|---|
| Chelsea | 2 | 1 | 3 | 8 | 10 |
| Liverpool | 3 | 1 | 2 | 10 | 8 |
Between them these 6 meetings produced 18 goals, an average of 3.00 a game.
Every meeting
Every recorded meeting, newest first.
- 9 May 2026
Premier LeagueLiverpool 1–1 Chelsea - 4 Oct 2025
Premier LeagueChelsea 2–1 Liverpool - 4 May 2025
Premier LeagueChelsea 3–1 Liverpool - 20 Oct 2024
Premier LeagueLiverpool 2–1 Chelsea - 25 Feb 2024
League CupChelsea 0–1 Liverpool - 31 Jan 2024
Premier LeagueLiverpool 4–1 Chelsea
How the model rates them now
Attack and defence are multipliers on an average side: above 1.00 in attack means the club scores more than the average team in its competition, and below 1.00 in defence means it concedes fewer. The methodology sets out how they are derived, including where the model is known to be wrong.
Leading scorers: Liverpool and Chelsea
Liverpool
No leading scorers listed yet.
Chelsea
GoalsAssists
- João Pedro53
- M. Rogers41
- C. Palmer23
Top 3 by league goals this season. Player availability is not part of the model — injuries and suspensions are a planned improvement.
Two sides the ratings cannot separate
Liverpool carry a model rating of 7.04 against 6.95 for Chelsea. The scale is anchored so that an average side in one of Europe's big five leagues sits at 6.85, and it is adjusted for the strength of each team's league, so the two numbers are comparable even across borders.
A gap that small is inside the noise of the rating itself — treat the two as level.
The model expects goals
Under the hood the model expects 2.22 goals for Liverpool and 1.33 for Chelsea. Those two numbers, not the scoreline, are what the model actually estimates; everything else is derived from them.
Liverpool rate 1.23 in attack and 0.84 in defence, where 1 is league average and a lower defensive number is better; Chelsea rate 1.31 and 1.11.
How this was produced
Publicly tracked record so far: 1,771 correct results out of 3,709 settled predictions (48%), with 376 exact scorelines (10%). That log is updated automatically after every match and is never edited.
Full method: how the model works.
From the first whistle this page keeps going: the win probabilities at the top are recalculated as the match plays out, reading the score, the clock and any sendings-off. The same model, asked how many goals are left rather than how many there will be.
Team comparison
Bars show each side's share of the two-team total, based on the model's attack/defence ratings and expected goals for this fixture.
Why the model favours this side
The model gives Liverpool the highest probability at 53%, based on each side's rolling attack and defence ratings estimated from recent match data. The model projects 2.22 expected goals for Liverpool against 1.33 for Chelsea. Liverpool's attack/defence rating is 1.23/0.84 versus the league average (1.00 = average), compared with 1.31/1.11 for Chelsea.
Goal trends
Over 1.5 goals
Over 2.5 goals
Over 3.5 goals
Both teams score
Outer rings: how often each side's most recent league matches before kick-off (up to 10 games; the count is under each ring) went over the line. Middle ring: the model's own probability for this match, from the same expected goals as the prediction above. Past matches are context; the middle ring is the estimate. Not betting advice.
Season stats
Liverpool at home · Chelsea away
League matches only, Premier League, this season. Rates are per game, since the two sides have not always played the same number; the tinted figure is the better one, where better is not a matter of taste. Source: API-Football season statistics, refreshed after each league match.
Goals scored and conceded, by minute
All Premier League matches this season: Liverpool 6, Chelsea 6.
Liverpool score most often in the 46–60 minute window (3 goals); Chelsea have let in 3 there.
Chelsea score most often in the 0–15 minute window (7 goals); Liverpool have let in 1 there.
Each cell counts the league goals that went in during that 15-minute window. Each side's goals sit above the goals its opponent has let in over the same minutes; the darker the cell, the more goals in that window. Stoppage time counts in the last window of its half. Source: API-Football season statistics, refreshed after each league match; own goals are not always counted on the same side as in the table, so a row can be a goal off it. Context, not a prediction.
The next 4 matches, by difficulty
Liverpool
- vs Arsenal1 Nov · Premier Leaguewin 27%
- at Fenerbahçe4 Nov · UEFA Champions Leaguewin 29%
Chelsea
- vs Manchester United31 Oct · Premier Leaguewin 34%
- at Sunderland8 Nov · Premier Leaguewin 42%
- vs Leeds21 Nov · Premier Leaguewin 40%
- at Nottingham Forest27 Nov · Premier Leaguewin 37%
Each side's next matches on our board after this one. The colour is how strong the opponent is, from easier to harder as on the scale above: its model rating ranked among the teams of that competition, in fifths, so bright red is the top fifth whoever is playing it. The win figure is this side's own chance in that match, which is where venue and its own level come in. A match already on our board links to its own prediction; the rest follow as their week comes up.