Sunday, 20/09/2026 at 15:00
Full timeManchester City 5:3 Sunderland
Manchester City were the model's likeliest winner before kick-off at 55%. They won 5–3. The biggest difference was Sunderland's attack: 3.57 xG against 0.94 projected (+2.63). Manchester City produced 2.82 xG against 1.71 projected (+1.11).
Goals
- 9' Enzo Fernández Manchester City · 0.04 xG chance 1–0
- 12' Brian Brobbey Sunderland · 0.61 xG chance 1–1
- 29' Rayan Cherki Manchester City · 0.08 xG chance 2–1
- 33' Brian Brobbey Sunderland · 0.89 xG chance 2–2
- 43' Antoine Semenyo Manchester City · 0.10 xG chance 3–2
- 57' Antoine Semenyo Manchester City · 0.05 xG chance 4–2
- 59' Brian Brobbey Sunderland · 1.00 xG chance 4–3
- 81' Erling Haaland Manchester City · 0.70 xG chance 5–3
The model against the result
- Manchester City
- 55% What happened
- Draw
- 28%
- Sunderland
- 18%
Data coverage Both teams have the full 20-match model window available.
Projected vs produced
Sunderland produced 2.63 xG more than projected
Projected before kick-off Produced on the night
The shots behind it
- Shots
- 16
- 17
- On target
- 6
- 7
- Inside the box
- 12
- 14
- xG per shot
- 0.17
- 0.25
- Open play xG
- 2.27
- 3.28
- Set-piece xG
- 0.44
- 0.99
- Penalties
- —
- —
Shot-level xG may differ from the published match total
The provider publishes the match total separately from its shot-by-shot values. Those values are rounded to two decimals and, leaving out penalties and own goals, add up to its non-penalty xG: 2.72 for Manchester City and 4.34 for Sunderland.
Shot map
Both teams shown attacking the same goal for comparison.
All shots
33 shots · 6.98 xG
Hover, focus or tap a shot to read it.
- Manchester City
- Sunderland
- Filled = goal
- Dot area = xG (the smallest chances drawn at a minimum size)
Recent form
1.75 xG created 1.28 xGA
- D 1.48 created, 2.25 conceded against Bournemouth
- L 1.34 created, 1.77 conceded against Aston Villa
- W 2.25 created, 0.65 conceded against Bournemouth
- W 2.21 created, 0.66 conceded against Crystal Palace
- W 2.12 created, 1.37 conceded against Coventry City
- W 1.10 created, 1.00 conceded against Manchester United
1.42 xG created 1.31 xGA
xG created xG conceded (xGA) W / D / L: the match result
Head to head
Pre-match model & market record
9 opportunities flagged3 won6 lost
- Lost
- Lost
- Lost
- Lost
- Won
- Lost
- Won
- Won
- Lost
View pre-match record Hide pre-match record
| Selection | Price | Edge | Result |
|---|---|---|---|
| Sunderland to win | 9.50 | +66.4% | Lost |
| Draw | 5.75 | +59.2% | Lost |
| Manchester City not to score | 10.00 | +55.0% | Lost |
| Manchester City at most 1 goal | 3.00 | +41.0% | Lost |
| Sunderland 2+ goals | 5.50 | +31.9% | Won |
| Under 2.5 goals | 2.38 | +16.5% | Lost |
| Sunderland to score | 1.80 | +15.4% | Won |
| Both teams to score | 1.95 | +8.8% | Won |
| Under 1.5 goals | 5.00 | +7.3% | Lost |
Pre-match model vs market
| Outcome | Model | Market | Difference | Fair odds |
|---|---|---|---|---|
| Manchester City | 54.8% | 73.6% | −18.8 pp | 1.82 |
| Draw | 27.7% | 16.5% | +11.2 pp | 3.61 |
| Sunderland | 17.5% | 10.0% | +7.6 pp | 5.71 |
Other markets
| Outcome | Model | Market | Difference | Fair odds |
|---|---|---|---|---|
| Manchester City to score | 84.5% | 90.4% | −5.9 pp | 1.18 |
| Manchester City not to score | 15.5% | 9.6% | +5.9 pp | 6.45 |
| Sunderland to score | 64.1% | 51.5% | +12.6 pp | 1.56 |
| Sunderland not to score | 35.9% | 48.5% | −12.6 pp | 2.78 |
| Both teams to score | 55.8% | 48.0% | +7.8 pp | 1.79 |
| Both teams not to score | 44.2% | 52.0% | −7.8 pp | 2.26 |
| Manchester City 2+ goals | 53.0% | 68.7% | −15.7 pp | 1.89 |
| Manchester City at most 1 goal | 47.0% | 31.3% | +15.7 pp | 2.13 |
| Sunderland 2+ goals | 24.0% | 17.2% | +6.8 pp | 4.17 |
| Sunderland at most 1 goal | 76.0% | 82.8% | −6.8 pp | 1.32 |
| Over 1.5 goals | 78.5% | 81.1% | −2.5 pp | 1.27 |
| Under 1.5 goals | 21.5% | 18.9% | +2.5 pp | 4.66 |
| Over 2.5 goals | 50.9% | 60.2% | −9.3 pp | 1.96 |
| Under 2.5 goals | 49.1% | 39.8% | +9.3 pp | 2.04 |
Prices and probabilities are the ones stored before kick-off; nothing here is recalculated. Edge is the model's probability times the price, minus one, so a flagged outcome that lost is the ordinary case rather than a contradiction.
Model details
Show Hide
- Fit window
- the last 20 matches, recency weighted (a match 107 days old counts half)
- Manchester City sample
- 20 matches
- Sunderland sample
- 20 matches
- Computed
- 1 day before kick-off
- Prices taken
- 1 day before kick-off
Limitations
The model reads expected goals and nothing else. It does not know the lineups, who is injured or suspended, what either side has to play for, the weather, or how congested the schedule around this match is.