xgEdge

Sunday, 18/10/2026 at 18:00

Full time
Match report · Serie A

Milan 3:2 Atalanta

Milan Home 2.58 xG · 11 shots
Atalanta Away 1.74 xG · 12 shots
32 The scoreline reflected the chances

Milan were the model's likeliest winner before kick-off at 44%. They won 3–2. The biggest difference was Milan's attack: 2.58 xG against 1.44 projected (+1.14). Atalanta produced 1.74 xG against 1.07 projected (+0.67).

Goals

  1. 24' Gonçalo Ramos (pen.) Milan · 0.82 xG chance 1–0
  2. 43' Adrien Rabiot Milan · 0.15 xG chance 2–0
  3. 57' Mario Gila Milan · 0.38 xG chance 3–0
  4. 66' Gianluca Gaetano Atalanta · 0.44 xG chance 3–1
  5. 81' Nikola Krstović Atalanta · 0.16 xG chance 3–2

The model against the result

priced before kick-off
Milan
44% What happened
Draw
31%
Atalanta
25%

Data coverage Both teams have the full 20-match model window available.

Projected vs produced

Expected goals
Milan The model projected 1.44 Produced 2.58 Difference +1.14
Atalanta The model projected 1.07 Produced 1.74 Difference +0.67

Milan produced 1.14 xG more than projected

Projected before kick-off Produced on the night

The shots behind it

Milan Atalanta
Shots
11
12
On target
6
4
Inside the box
7
6
xG per shot
0.24
0.14
Open play xG
0.65
1.69
Set-piece xG
0.38
0.00
Penalties
2
0

Milan's shots add up to 2.68 xG and Atalanta's to 1.69; the published figures are 2.58 and 1.74. Leaving out penalties and own goals, Milan's shots come to 1.03 against a published non-penalty xG of 1.01: more than rounding can explain, so the two figures were not read at the same point of the provider's revisions. Leaving out penalties and own goals, Atalanta's shots come to 1.69 against a published non-penalty xG of 1.74: more than rounding can explain, so the two figures were not read at the same point of the provider's revisions. The provider computes a match's total xG and its non-penalty xG separately, so the difference between the two is not what a penalty was worth. Milan's 2 penalties are worth 1.64 xG in the shot data.

Shot map

Both teams shown attacking the same goal for comparison.

Milan · Gonçalo Ramos · 24' · Goal · 12 m · 0.82 xG Milan · Gonçalo Ramos · 57' · No goal · 12 m · 0.82 xG Atalanta · Gianluca Gaetano · 66' · Goal · 6 m · 0.44 xG Atalanta · Davide Zappacosta · 83' · No goal · 9 m · 0.42 xG Milan · Mario Gila · 57' · Goal · 3 m · 0.38 xG Atalanta · Nikola Krstović · 19' · No goal · 11 m · 0.22 xG Milan · Gonçalo Ramos · 90' · No goal · 15 m · 0.19 xG Atalanta · Nikola Krstović · 51' · No goal · 11 m · 0.18 xG Atalanta · Nikola Krstović · 81' · Goal · 12 m · 0.16 xG Milan · Adrien Rabiot · 43' · Goal · 14 m · 0.15 xG Milan · Yunus Musah · 22' · No goal · 18 m · 0.10 xG Milan · Samuel Chukwueze · 48' · No goal · 15 m · 0.07 xG Atalanta · Nikola Krstović · 63' · No goal · 17 m · 0.07 xG Milan · Adrien Rabiot · 10' · No goal · 14 m · 0.06 xG Atalanta · Lazar Samardžić · 74' · No goal · 18 m · 0.05 xG Milan · Adrien Rabiot · 54' · No goal · 19 m · 0.04 xG Atalanta · Éderson · 74' · No goal · 25 m · 0.04 xG Atalanta · Charles De Ketelaere · 16' · No goal · 19 m · 0.04 xG Atalanta · Davide Zappacosta · 90' · No goal · 15 m · 0.03 xG Milan · Ardon Jashari · 88' · No goal · 25 m · 0.03 xG Atalanta · Lazar Samardžić · 42' · No goal · 19 m · 0.03 xG Milan · Diego Moreira · 45' · No goal · 26 m · 0.02 xG Atalanta · Lazar Samardžić · 66' · No goal · 27 m · 0.02 xG

All shots

23 shots · 4.37 xG

Hover or tap a shot, or step through them with the arrows.

  • Milan
  • Atalanta
  • Filled = goal
  • Dot area = xG (the smallest chances drawn at a minimum size)

Figures read 7 November, 17:08 · final

Recent form

Last 6 · xG per match
Milan

1.52 xG created 1.18 xGA

  • W 2.15 created, 1.29 conceded against Torino
  • W 2.21 created, 0.80 conceded against Venezia
  • D 0.11 created, 1.59 conceded against Juventus
  • D 1.25 created, 2.01 conceded against Lazio
  • W 2.19 created, 0.21 conceded against Lecce
  • D 1.18 created, 1.19 conceded against Sassuolo
Atalanta

1.04 xG created 1.97 xGA

  • W 0.88 created, 2.70 conceded against Sassuolo
  • W 0.15 created, 1.05 conceded against Bologna
  • L 0.70 created, 2.97 conceded against Roma
  • L 1.74 created, 1.32 conceded against Cagliari
  • L 0.60 created, 1.31 conceded against Juventus
  • W 2.19 created, 2.48 conceded against Venezia

xG created xG conceded (xGA) W / D / L: the match result

Head to head

Milan: 0 won · 1 drawn · 1 lost

Pre-match model & market record

6 signals flagged0 won6 lost

  • Lost
  • Lost
  • Lost
  • Lost
  • Lost
  • Lost
View pre-match record
The outcomes the model flagged on Milan against Atalanta before kick-off, in the order they ranked, with the price, the model-implied edge and what became of each.
Selection Price Model-implied edge Result
Milan not to score 6.00 +26.3% Lost
Draw 4.00 +23.5% Lost
Milan at most 1 goal 2.10 +18.8% Lost
Under 2.5 goals 2.20 +16.7% Lost
Atalanta to win 4.50 +13.4% Lost
Under 1.5 goals 4.50 +9.0% Lost

Pre-match model vs market

Outcome Model Market Disagreement Model price
Milan 43.9% 55.1% −11.2 pp 2.28
Draw 30.9% 23.8% +7.1 pp 3.24
Atalanta 25.2% 21.1% +4.1 pp 3.97

Other markets

Outcome Model Market Disagreement Model price
Milan to score 78.9% 84.2% −5.3 pp 1.27
Milan not to score 21.1% 15.8% +5.3 pp 4.75
Atalanta to score 68.7% 64.5% +4.1 pp 1.46
Atalanta not to score 31.3% 35.5% −4.1 pp 3.19
Both teams to score 56.0% 57.7% −1.6 pp 1.78
Both teams not to score 44.0% 42.3% +1.6 pp 2.27
Milan 2+ goals 43.4% 55.7% −12.3 pp 2.30
Milan at most 1 goal 56.6% 44.3% +12.3 pp 1.77
Atalanta 2+ goals 29.1% 29.1% +0.0 pp 3.44
Atalanta at most 1 goal 70.9% 70.9% −0.0 pp 1.41
Over 1.5 goals 75.8% 78.9% −3.2 pp 1.32
Under 1.5 goals 24.2% 21.1% +3.2 pp 4.13
Over 2.5 goals 47.0% 56.9% −9.9 pp 2.13
Under 2.5 goals 53.0% 43.1% +9.9 pp 1.89

Prices and probabilities are the ones stored before kick-off; nothing here is recalculated. The model-implied edge is the model's probability times the price, minus one: what a stake of one would return on average if the model is right, not observed return, so a flagged outcome that lost is the ordinary case rather than a contradiction.

Model details

Show
Fit window
the last 20 matches, recency weighted (a match 107 days old counts half)
Milan sample
20 matches
Atalanta sample
20 matches
Computed
2 hours before kick-off
Prices taken
5 hours 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.