Dataset Viewer
Auto-converted to Parquet Duplicate
match_id
string
season
string
minute
int64
second
int64
attacking_team
string
opponent
string
home_away
string
score_for
int64
score_against
int64
field_tilt_5min
float64
crosses_5min
int64
box_occupancy_avg
float64
clearance_success_5min
float64
second_ball_wins_5min
int64
xg_next_sequence
float64
shots_next_sequence
int64
goal_next_sequence
int64
label_goal_next_sequence
int64
EPL_2025_021
2025-26
18
44
ARS
CRY
home
0
0
0.64
3
4.2
0.82
2
0.11
1
0
0
EPL_2025_021
2025-26
21
9
ARS
CRY
home
0
0
0.71
5
5.1
0.63
3
0.28
2
1
1
EPL_2025_022
2025-26
37
33
LIV
BRE
away
1
0
0.58
2
3.8
0.88
1
0.07
0
0
0
EPL_2025_022
2025-26
39
2
LIV
BRE
away
1
0
0.73
6
6.4
0.59
4
0.34
2
1
1
EPL_2025_023
2025-26
52
11
TOT
WHU
home
2
1
0.61
4
4.7
0.79
2
0.14
1
0
0
EPL_2025_023
2025-26
55
26
TOT
WHU
home
2
1
0.69
7
6.9
0.54
5
0.39
3
1
1
EPL_2025_024
2025-26
66
48
MCI
NEW
home
1
1
0.66
3
4.5
0.84
2
0.09
1
0
0
EPL_2025_024
2025-26
69
14
MCI
NEW
home
1
1
0.76
8
7.2
0.51
6
0.42
3
1
1
EPL_2025_025
2025-26
81
5
CHE
AVL
away
0
0
0.59
2
3.1
0.9
1
0.05
0
0
0
EPL_2025_025
2025-26
83
22
CHE
AVL
away
0
0
0.72
6
6
0.57
4
0.31
2
1
1

EPL In-Play Quad Sustained Pressure Breakthrough v0.1

What this dataset is

You test whether a model can detect when sustained attacking pressure converts into a breakthrough goal.

Each row represents a live attacking sequence snapshot.

Core quad coupling

Field tilt
Cross frequency
Box occupancy
Clearance success rate

The label asks

Will the attacking team score in the next sequence

Why this matters

Pressure alone does not cause goals.

Pressure combined with clearance instability and second-ball wins drives collapse.

This dataset tests pressure-to-breakthrough transition detection.

Columns

match_id
season
minute
second
attacking_team
opponent
home_away
score_for
score_against
field_tilt_5min
crosses_5min
box_occupancy_avg
clearance_success_5min
second_ball_wins_5min
xg_next_sequence
shots_next_sequence
goal_next_sequence
label_goal_next_sequence

Target label

label_goal_next_sequence

Evaluation

Use scorer.py.

Predictions accepted

0 or 1
goal or no_goal
probability float

Files

data/train.csv
scorer.py

License

MIT

Downloads last month
17