Free EFL Cup predictions: the model's favourite on every fixture with the best price beside it, and the model's calls graded on the public record, misses included.
8 EFL Cup matches from Tuesday 27 October
market favourite Bradford 1.83 55% implied
Over 2.5 1.95, market 51%
market favourite Arsenal 1.07 93% implied
Over 2.5 1.33, market 75%
not priced yet
market favourite Sunderland 2.60 38% implied
Over 2.5 1.85, market 54%
market favourite Fulham 2.30 43% implied
Over 2.5 1.85, market 54%
market favourite Bournemouth 2.30 43% implied
Over 2.5 1.80, market 56%
market favourite Liverpool 2.25 44% implied
Over 2.5 1.67, market 60%
market favourite Everton 2.35 43% implied
Over 2.5 2.00, market 50%
The EFL Cup, Carabao Cup, League Cup, whatever sponsor it carries this season, is the single noisiest competition we model. Wholesale rotation from Premier League sides in the early rounds, semifinals played over two legs, finals that sometimes feature lineups closer to a Champions League starting XI than the third-round version. We split our model into two regimes: rounds 1-3 (rotation expected, lean on home form and squad depth) and quarter-finals onwards (starting XIs likely full strength, revert to standard probability logic). Below: every upcoming EFL Cup fixture with model probabilities for match result, both-teams-to-score and over/under 2.5 goals. The fixture page for each tie includes a head-to-head record where available, full bookmaker comparison and an honest read of which side is likely to play their first-choice XI. Goals tend to come in this competition. The over 2.5 hit rate has historically outperformed our overall site average in the rounds played at Premier League grounds, but we don't publish that as a rule because the sample is too small to be statistically reliable on a per-season basis.
Last 5 results in the league per team. Sorted by form points (W=3, D=1, L=0).
How often each market has landed in the EFL Cup, with an honest read on where the model has a proven record.
Every Carabao Cup call the model makes is logged before kick-off and graded win or loss on the public record, misses included. The record link on this page opens it, with the full settled history on the track record page. The competition is one of our smallest by volume.
We split our model into two regimes: rounds 1-3 where Premier League sides routinely rotate, and quarter-finals onwards where starting XIs are typically full strength. The probability logic differs significantly between the two.
Historically yes, particularly in the early rounds played at Premier League grounds. The over 2.5 hit rate in those ties has run higher than our overall site average, but the sample size is too small to build a rule on.
The final is one of the most-data-rich fixtures of the season: starting XIs are confirmed, motivation is symmetrical, the data feed is dense. Our model treats it on equal footing with Champions League knockouts in terms of probability confidence.
Yes, each leg is treated as its own fixture for prediction purposes, but the reasoning text on the second-leg preview accounts for the first-leg aggregate scoreline and the in-form side's tactical adjustment.
The match-result market is genuinely volatile in this competition because rotation patterns are hard to forecast. The BTTS and over/under markets carry less line-movement noise and have historically delivered the steadier hit rate.
Our model reads the prices of the UK bookmakers we track, the form and the head to head to produce EFL Cup predictions. Every fixture shows the best price we can find beside the model's number, on the match result, both teams to score, the goals lines and more.
Every pick published before kick-off, graded in public. The losing ones stay on the page.
Bradford
Peterborough
Fleetwood Town
Arsenal
Manchester City
Brighton
Sunderland
Brentford
Fulham
Crystal Palace
Bournemouth
Aston Villa
Liverpool
Chelsea
Everton
Newcastle
