Fulham v Chelsea
Fulham vs Chelsea
Premier League- Model pick
- Chelsea win 36%
- Likely score
- 1-1 (13%)
- Kickoff
- Mon 24 Aug, 20:00 UK · Fulham home venue
Why
- Fulham rated 62 Elo points higher, home advantage included
- Fulham in better recent form (0.6 more points a game)
“Elo alone” is the plain rating model before form, attack, defence and the goals model are added; the gap shows what those add.
What the model weighed
- Fulham rated 62 Elo points higher, home advantage included
- Fulham in better recent form (0.6 more points a game)
- Fulham stronger on head-to-head ratings
- Fulham are at home
- Fulham concede less (0.4 goals a game fewer)
- Chelsea have the better goal difference (0.3 a game)
Goals
Total goals
Both teams to score
Most likely scores
- 1-113.3%
- 2-19.5%
- 1-08.9%
- 2-08.6%
- 0-08.1%
- 1-26.5%
Score by result
- Fulham win2-19.5%
- Draw1-113.3%
- Chelsea win · pick1-26.5%
From the goals model (expected goals with the Dixon-Coles low-score correction), so these can differ slightly from the calibrated win/draw/win forecast. A single score is rarely right (about one match in eight); the actual score is among the top three about a quarter of the time. Model estimates, not betting prices.
- Cairney MIDinjuredKnee injury - Expected back 21 Nov
No readable news on the club's site (it may load its news in the browser).
fulhamfc.com- Gittens MIDinjuredAnkle injury - Unknown return date
- Hato DEFinjuredUnspecified injury - Unknown return date
- James DEFinjuredKnee injury - Unknown return date
- João Pedro FWDinjuredThigh injury - Unknown return date
- Lacroix DEFinjuredHamstring injury - Unknown return date
- Acheampong DEFdoubtfulUnspecified injury - 75% chance of playing
- Martinez GKPdoubtfulHead injury - 75% chance of playing
No team news on the club's site in the last 7 days.
chelseafc.comToday's injury list, not the one for this match; news from each club's own website, scanned Mon 16 Nov, 18:00 UK. The forecast does not use this news.
Goalkeepers
Defenders
Midfielders
Forwards
Goalkeepers
Defenders
Midfielders
- Cole Palmer3 goals
- Jordan Henderson2 goals
- Mahdi Nicoll-Jazuli
- Moisés Caicedo
- Morgan Rogers5 goals
- Romeo Lavia1 goal
- Valentin Barco
Forwards
Season goals from the competition's top scorers; absences from the injury list.
Looking for lineups…
Weather at kickoff
No weather record for this match.
Rest
- Fulham
- no earlier match found
- Chelsea
- no earlier match found
Prices
| Result | Model | Model fair odds | Bookmaker |
|---|---|---|---|
| Fulham win | 34.6% | 2.89 | 3.66 |
| Draw | 29.4% | 3.40 | 3.76 |
| Chelsea win | 36.0% | 2.78 | 1.89 |
Bookmaker prices from Market average (football-data.co.uk), retrieved Tue 17 Nov, 08:00 UK. Shown for comparison only: picks use the model's chances alone.
Model
- Version
- football-outcome-ensemble-v1+dc-mle-xi0.0019+goalcal
- Confidence
- 66.7% (59.9% after data quality)
- Data quality
- 89.8%
- Simulations
- 3,000
- Basis
- Finished Premier League fixture, forecast as of kickoff (pre-match Elo 1492 v 1485). Final score: Fulham 2-3 Chelsea. Chelsea were favoured at 36% from 389 completed matches (2025-08-15 to 2026-08-23). Value signals compare this with pre-match bookmaker prices from football-data.co.uk.
- Sources
- Fixtures and results: openfootball and football-data.org · weather: Open-Meteo · injuries: league injury list · prices: football-data.co.uk
How The Model Does In Premier League
100 gradedAre its chances honest?
- Said 20–30%happened 27%143 calls
- Said 30–40%happened 33%78 calls
- Said 40–50%happened 42%50 calls
- Said 50–60%happened 53%19 calls
On average its chances are 1.6% away from what happened. Full record
How Fresh It Is
updated by the worker- Results and fixtures51m agoon schedule
- Injury list57m agoon schedule
- Weather57m agoon schedule
- Bookmaker prices5h agoon schedule
Sources
- openfootball (fixtures, results)
- football-data.org (results, squads)
- ESPN (live scores, lineups)
- Open-Meteo (weather)
- Premier League injury list
- football-data.co.uk (prices, display only)
Checks The Model Passes
6 of 7- reviewPatterns are not assumed to be causes
- passedTested on matches it was not trained on
- passedOverlapping signals kept apart
- passedUses only data from before each match
- passedRelegated and promoted clubs kept in
- passedEvery fixture forecast, not a chosen few
- passedNo hindsight in the features
Hover a check for how it is enforced.