ARS5–1
MCIARS wonEmirates Stadium
Football's answer engine
The AI analyst for football.
The product
An answer is a widget and a story. With SWOF, both at once.
Swof gives you neither a wall of text nor a bare stat. Every answer you get brings together live data you can see and reasoning you can follow.
Who wins Man City vs Arsenal?
Thought for 7 secondsThe market has City as favourites: 1.65, an implied probability of around 61%. The data backs that up, but two things narrow the gap.
What the market says
A home win is priced at 1.65, which works out at an implied probability of about 61%.
Odds1.65Man City winImplied probability61%what the market pricesThe numbers on both sides
City score 2.47 goals per game at home and took 10 points from their last five. Arsenal average 1.88 points per game away and have the best defence in the league: 0.80 conceded per game, 1st of 20.
Manchester C.MCIGoals per game at home2.47Last 510 pointsWWDWLArsenalARSPoints per game away1.88Conceded per game0.801st of 20Starting XI
Manchester City average 7.50 in a 4-3-3, Arsenal 7.22 in a 4-3-3; the highest rating on the pitch is Haaland (8.7).
Manchester City
Formation 4-3-3
- 31Ederson
- 6Aké
- 25Akanji
- 3Dias
- 2Walker
- 20Silva
- 16Rodri
- 17De Bruyne
- 11Doku
- 9Haaland
- 47Foden
- 11Martinelli
- 9Jesus
- 7Saka
- 29Havertz
- 41Rice
- 8Ødegaard
- 35Zinchenko
- 6Gabriel
- 2Saliba
- 4White
- 22Raya
Arsenal
Formation 4-3-3
In the last 10 meetings Arsenal won 4, City won 2 and 4 were drawn. The most recent, in February at the Emirates, finished 5-1 to Arsenal.
Arsenal are without Calafiori (injured) and Lewis-Skelly (suspended). Both are left backs, and City's most productive attacking side is the right.
Bottom line: City are slight favourites, but the gap is narrower than the market suggests. I'd expect a tight game.
This is a prediction, not betting advice.
Will it go over 2.5?
Thought for 5 secondsOver 2.5 is 1.72, so the market is saying around 58%. I don't agree in this game.
Goals in this fixture
Both sides score freely across the season: City 2.26 per game, Arsenal 2.03. But this fixture behaves differently: five of the last 10 meetings finished with two goals or fewer, against 43% league-wide.
Last 10 meetings
Arsenal won the last meeting 5–1. In the last 10 matches: Manchester City 4 wins, Arsenal 3, 3 draws.
- 08 Feb 2026ARS5–1MCIARS wonEmirates Stadium
- 22 Sept 2025MCI2–2ARSDrawEtihad Stadium
- 02 Feb 2025ARS5–1MCIARS wonEmirates Stadium
- 22 Sept 2024MCI2–2ARSDrawEtihad Stadium
- 31 Mar 2024MCI0–0ARSDrawEtihad Stadium
- 08 Oct 2023ARS1–0MCIARS wonEmirates Stadium
- 26 Apr 2023MCI4–1ARSMCI wonEtihad Stadium
- 15 Feb 2023ARS1–3MCIMCI wonEmirates Stadium
- 01 Jan 2022ARS1–2MCIMCI wonEmirates Stadium
- 28 Aug 2021MCI5–0ARSMCI wonEtihad Stadium
The reason is the defences. Arsenal concede 0.80 per game, City 0.89. Both sides cut their risk against each other.
Arsenal are also missing both left backs. Arteta tends to set up more conservatively in that situation.
Bottom line: under 2.5 at 2.10 looks the better value.
This is a prediction, not betting advice.
- 08 Feb 2026
Will Haaland score?
Thought for 4 secondsHis chance of scoring is 52%. That's the highest in the match, but below his own season average.
Chance of scoring
The highest figure in the match is Haaland's, and it still sits below his own season average.
Season numbers
27 goals in 33 games, 0.83 per 90. He leads the scoring charts by five.
Erling HaalandManchester City · ForwardGoals27Games33Per 900.83Where he puts his shots
69% of his shots on target go to the corners; he converts 56% there against 16% through the middle, where the keeper stands. His best zone is bottom right: 8 goals from 12 shots.
- 5 goals9 shots · xGOT 4.3
- 1 goal7 shots · xGOT 1.3
- 5 goals8 shots · xGOT 4.0
- 6 goals14 shots · xGOT 5.4
- 2 goals12 shots · xGOT 2.1
- 8 goals12 shots · xGOT 5.9
From the shooter's viewDeeper green means more goals
Finishing
27 goals, +4.0 above the expected value of his shots on target; placement is winning him extra goals. Against Arsenal the question is whether he gets the shot, not what he does with it.
Shots on target62this seasonGoals2744% conversionxGOT23.0expected goalsDifference+4.0above expectedWhat he did in the last 5
He found the net in 3 of the last 5: 0.85 goals per 90, just above his season rate of 0.83, so the form matches the season line rather than beating it. The two blanks came against Chelsea and Leeds, both sides that sat deep.
Date Opponent Venue Result Min Goals Assists 26 Apr 2026 
BrightonAway L1–2 90’ 1 0 18 Apr 2026 
WolvesHome W3–0 82’ 1 1 11 Apr 2026 
ChelseaAway D1–1 90’ 0 0 04 Apr 2026 
BournemouthHome W4–0 71’ 2 0 21 Mar 2026 
LeedsHome W2–0 90’ 0 1 The drop comes down to the opponent. Arsenal keep a clean sheet in 43% of their games and concede 0.80 per game, both the best in the league. That is exactly the profile that shuts Haaland down.
In his favour: the game is at the Etihad, and City score 2.47 per game at home.
Bottom line: he is the most likely scorer here, but the price doesn't fully account for Arsenal's defence.
This is a prediction, not betting advice.
Questions
Frequently asked
The short answers. Anything else, ask us at hello@swof.ai.
What does Swof actually do?
You ask a football question in plain language. Swof pulls professional match data, checks it against the source and answers with a card you can read at a glance plus a short narrative.
When does the beta open?
First invites go out in October. Confirm your email after joining to lock your place; invites go out in order.
Does it cost anything?
No. Swof is free during the beta.
Which leagues are covered?
Süper Lig first, then the major European leagues. Tell us what you'd use Swof for when you join and we'll shape the beta around it.
Where does the data come from?
Professional match data, not screenshots of someone else's tables. Every answer is tied to its source.
What happens to my email?
It's used only for the beta invite. Every email carries an unsubscribe link, and you can leave the list at any time.
Stop looking it up. Start asking.
We open in October. Get your place in line.