Arsenal vs Chelsea — Premier League 2026
Match Prediction
Statistical Prediction
Based on xG from last 35+ matches
Arsenal48%
Chelsea27%✓ Correct in 49 of 100 matches — 5 more than always backing the home side
The model picks a winner (home or away) for each finished match using only pre-kickoff data, then checks it against the real result: 48.6% correct across 15,072 matches, vs 43.8% for always backing the home side.
Home picks 50.1% correct · Away picks 45.3% correct
Football matches only. Past performance is not indicative of future results.
Expected goals in this match
What is xG?
Every shot is scored 0–1 by how likely it is to be a goal from that spot — a chance in the middle of the box ≈ 0.4, a long shot from a tight angle ≈ 0.03, a penalty ≈ 0.76. Add them up over a match and that is the xG.
The scoreline can lie: you can win 1-0 on a fluke while creating half the chances. xG shows who was really closer to scoring — and over a season it predicts results better than goals alone.
Arsenal creates 42% more chances
Season form · 35 home / 35 away
creates per match
allows per match
finishing
Total goals
- Over52
- Under48
Close call
Both teams score
- Yes55
- No45
Both teams likely to score
Score probability
Probability of each scoreline, %
Arsenal ↓
Most likely 1–1 (12%) · grid covers 97% of scorelines
More markets
Total goals
Double chance
Winning margin
Team goals
Draw no bet
Combos
Model probabilities, not odds. For information only.
How is this calculated?
Every number below comes from this match — nothing is made up.
Recent form of each side
Arsenal at home — creates 1.77, concedes 0.81 · 35 matches
Chelsea away — creates 1.49, concedes 1.49 · 35 matches
⚑ Home and away form are counted separately — the home-field edge is already baked into these numbers.
Expected goals for this match
Arsenal attack 1.77 + Chelsea defence 1.49 → ÷2 → 1.63
Chelsea attack 1.49 + Arsenal defence 0.81 → ÷2 → 1.15
Each side's attack is blended with the other's defence, then averaged.
Chance of every scoreline
From those two numbers a Poisson distribution gives the probability of each exact score — the grid shown above.
The answer to "why 48%?"
Arsenal at 48% is simply the biggest of the three — that is the pick.
What the model can't see
It reads only the stats of recent matches. It knows nothing about:
- injuries & suspensions
- the matchday lineup
- motivation — cup, relegation, a title race
- weather & pitch condition
- in-game coaching decisions
So 48% does not mean "Arsenal will win" — it means "on the available stats, this is the most likely result."
Statistical model based on expected goals (xG). For informational purposes only.
Player of the match
Match Events
Arsenal
Aston Villa
Bournemouth
Brentford
BrightonChelsea
Coventry
Crystal Palace
Everton
Fulham
Hull City
Ipswich
Leeds
Liverpool
Manchester City
Manchester United
Newcastle
Nottingham Forest
Sunderland
TottenhamIf points are equal: 1. Goal difference 2. Goals scored 3. Head-to-head
P = Played · W = Won · D = Drawn · L = Lost · GF = Goals for · GA = Goals against · GD = Goal difference · Pts = Points
Possession
Shots
Pass accuracy
Statistics
Arsenal host Chelsea on Wednesday, 3 January 2018 at 19:45. The match is part of the Premier League 2026/2027 season.
Premier League: Arsenal 2–2 Chelsea
Arsenal and Chelsea drew 2-2 in Premier League on January 3, 2018.
Goals: J. Wilshere (63'), E. Hazard (67' pen), Marcos Alonso (84'), Bellerín (90+2').
Arsenal controlled possession (58%) and registered 14 shots to 19.
The match was played at Emirates Stadium in London.