In [61]:

import pandas as pd
from mplsoccer.pitch import Pitch

In [62]:

#read in csv
df = pd.read_csv('valladolidA.csv')

In [63]:

#filter df to get only the team we want
df = df[df['teamId']=='Barcelona']

In [64]:

#now we want to find our passes and recipients and then filter for only passes
df['passer'] = df['playerId']
df['recipient'] = df['playerId'].shift(-1)

#find passes and then only look for the successful passes
passes = df[df['type']=='Pass']
successful = passes[passes['outcome']=='Successful']

In [65]:

successful

Out[65]:

	id	eventId	minute	second	teamId	x	y	period	type	outcome	playerId	endX	endY	passer	recipient
2	2248226929	3	0	1.0	Barcelona	50.0	50.0	1	Pass	Successful	9.0	43.9	46.9	9.0	8.0
3	2248226941	4	0	2.0	Barcelona	43.8	46.9	1	Pass	Successful	8.0	36.4	56.0	8.0	21.0
4	2248226951	5	0	4.0	Barcelona	36.6	56.6	1	Pass	Successful	21.0	28.5	73.8	21.0	15.0
5	2248226973	6	0	5.0	Barcelona	28.3	74.1	1	Pass	Successful	15.0	11.6	51.8	15.0	1.0
6	2248226975	7	0	8.0	Barcelona	11.2	53.7	1	Pass	Successful	1.0	11.6	81.4	1.0	15.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1651	2248256867	1018	93	14.0	Barcelona	76.7	31.8	2	Pass	Successful	8.0	76.7	43.0	8.0	10.0
1652	2248256869	1019	93	17.0	Barcelona	79.6	24.2	2	Pass	Successful	10.0	77.1	53.6	10.0	14.0
1653	2248256983	1020	93	18.0	Barcelona	77.1	56.4	2	Pass	Successful	14.0	69.0	56.9	14.0	44721.0
1654	2248256989	1021	93	20.0	Barcelona	68.3	55.2	2	Pass	Successful	44721.0	87.0	17.3	44721.0	2.0
1667	2248257081	1027	93	50.0	Barcelona	75.6	4.6	2	Pass	Successful	2.0	77.1	26.8	2.0	10.0

635 rows × 15 columns

In [66]:

#find the first subsititution and filter the successful dataframe to be less than that minute
subs = df[df['type']=='SubstitutionOff']
subs = subs['minute']
firstSub = subs.min()

successful = successful[successful['minute'] < firstSub]

In [67]:

successful

Out[67]:

	id	eventId	minute	second	teamId	x	y	period	type	outcome	playerId	endX	endY	passer	recipient
2	2248226929	3	0	1.0	Barcelona	50.0	50.0	1	Pass	Successful	9.0	43.9	46.9	9.0	8.0
3	2248226941	4	0	2.0	Barcelona	43.8	46.9	1	Pass	Successful	8.0	36.4	56.0	8.0	21.0
4	2248226951	5	0	4.0	Barcelona	36.6	56.6	1	Pass	Successful	21.0	28.5	73.8	21.0	15.0
5	2248226973	6	0	5.0	Barcelona	28.3	74.1	1	Pass	Successful	15.0	11.6	51.8	15.0	1.0
6	2248226975	7	0	8.0	Barcelona	11.2	53.7	1	Pass	Successful	1.0	11.6	81.4	1.0	15.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1230	2248252877	782	69	30.0	Barcelona	83.1	42.8	2	Pass	Successful	9.0	89.9	21.5	9.0	2.0
1231	2248252881	783	69	32.0	Barcelona	90.5	23.0	2	Pass	Successful	2.0	78.6	36.9	2.0	8.0
1234	2248252897	785	69	36.0	Barcelona	84.6	27.7	2	Pass	Successful	8.0	90.3	14.3	8.0	2.0
1235	2248252905	786	69	37.0	Barcelona	90.7	14.3	2	Pass	Successful	2.0	88.4	15.3	2.0	8.0
1236	2248252913	787	69	38.0	Barcelona	88.4	15.3	2	Pass	Successful	8.0	80.1	33.0	8.0	16.0

505 rows × 15 columns

In [68]:

#this makes it so our passer and recipients are float values
pas = pd.to_numeric(successful['passer'],downcast='integer')
rec = pd.to_numeric(successful['recipient'],downcast='integer')
successful['passer'] = pas
successful['recipient'] = rec

<ipython-input-68-7dc55db856cb>:4: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  successful['passer'] = pas
<ipython-input-68-7dc55db856cb>:5: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  successful['recipient'] = rec

In [69]:

#now we need to find the average locations and counts of the passes
average_locations = successful.groupby('passer').agg({'x':['mean'],'y':['mean','count']})
average_locations.columns = ['x','y','count']

In [70]:

average_locations

Out[70]:

	x	y	count
passer
1	7.124000	50.324000	25
2	68.274359	11.915385	39
4	32.895833	43.518750	48
8	53.332927	41.608537	82
9	69.916667	51.383333	6
10	63.201613	42.351613	62
15	38.409091	74.378182	55
16	69.594286	69.471429	35
18	55.242553	83.793617	47
21	55.663636	52.263636	55
28	49.156863	16.282353	51

In [71]:

successful

Out[71]:

	id	eventId	minute	second	teamId	x	y	period	type	outcome	playerId	endX	endY	passer	recipient
2	2248226929	3	0	1.0	Barcelona	50.0	50.0	1	Pass	Successful	9.0	43.9	46.9	9	8
3	2248226941	4	0	2.0	Barcelona	43.8	46.9	1	Pass	Successful	8.0	36.4	56.0	8	21
4	2248226951	5	0	4.0	Barcelona	36.6	56.6	1	Pass	Successful	21.0	28.5	73.8	21	15
5	2248226973	6	0	5.0	Barcelona	28.3	74.1	1	Pass	Successful	15.0	11.6	51.8	15	1
6	2248226975	7	0	8.0	Barcelona	11.2	53.7	1	Pass	Successful	1.0	11.6	81.4	1	15
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1230	2248252877	782	69	30.0	Barcelona	83.1	42.8	2	Pass	Successful	9.0	89.9	21.5	9	2
1231	2248252881	783	69	32.0	Barcelona	90.5	23.0	2	Pass	Successful	2.0	78.6	36.9	2	8
1234	2248252897	785	69	36.0	Barcelona	84.6	27.7	2	Pass	Successful	8.0	90.3	14.3	8	2
1235	2248252905	786	69	37.0	Barcelona	90.7	14.3	2	Pass	Successful	2.0	88.4	15.3	2	8
1236	2248252913	787	69	38.0	Barcelona	88.4	15.3	2	Pass	Successful	8.0	80.1	33.0	8	16

505 rows × 15 columns

In [72]:

#now we need to find the number of passes between each player
pass_between = successful.groupby(['passer','recipient']).id.count().reset_index()
pass_between.rename({'id':'pass_count'},axis='columns',inplace=True)

#merge the average location dataframe. We need to merge on the passer first then the recipient
pass_between = pass_between.merge(average_locations, left_on='passer',right_index=True)
pass_between = pass_between.merge(average_locations, left_on='recipient',right_index=True,suffixes=['', '_end'])

In [73]:

pass_between

Out[73]:

	passer	recipient	pass_count	x	y	count	x_end	y_end	count_end
0	1	2	3	7.124000	50.324000	25	68.274359	11.915385	39
18	4	2	1	32.895833	43.518750	48	68.274359	11.915385	39
26	8	2	8	53.332927	41.608537	82	68.274359	11.915385	39
35	9	2	2	69.916667	51.383333	6	68.274359	11.915385	39
39	10	2	6	63.201613	42.351613	62	68.274359	11.915385	39
...	...	...	...	...	...	...	...	...	...
52	15	10	1	38.409091	74.378182	55	63.201613	42.351613	62
60	16	10	10	69.594286	69.471429	35	63.201613	42.351613	62
70	18	10	7	55.242553	83.793617	47	63.201613	42.351613	62
77	21	10	14	55.663636	52.263636	55	63.201613	42.351613	62
87	28	10	5	49.156863	16.282353	51	63.201613	42.351613	62

90 rows × 9 columns

In [74]:

#set minimum threshold of combinations.. I like 5 for high passing teams. 2 or 3 for low passing.
pass_between = pass_between[pass_between['pass_count']>5]

In [75]:

pass_between

Out[75]:

	passer	recipient	pass_count	x	y	count	x_end	y_end	count_end
26	8	2	8	53.332927	41.608537	82	68.274359	11.915385	39
39	10	2	6	63.201613	42.351613	62	68.274359	11.915385	39
83	28	2	10	49.156863	16.282353	51	68.274359	11.915385	39
27	8	4	8	53.332927	41.608537	82	32.895833	43.518750	48
50	15	4	9	38.409091	74.378182	55	32.895833	43.518750	48
84	28	4	13	49.156863	16.282353	51	32.895833	43.518750	48
11	2	8	11	68.274359	11.915385	39	53.332927	41.608537	82
19	4	8	6	32.895833	43.518750	48	53.332927	41.608537	82
41	10	8	17	63.201613	42.351613	62	53.332927	41.608537	82
51	15	8	6	38.409091	74.378182	55	53.332927	41.608537	82
68	18	8	6	55.242553	83.793617	47	53.332927	41.608537	82
76	21	8	13	55.663636	52.263636	55	53.332927	41.608537	82
85	28	8	13	49.156863	16.282353	51	53.332927	41.608537	82
21	4	15	20	32.895833	43.518750	48	38.409091	74.378182	55
30	8	15	6	53.332927	41.608537	82	38.409091	74.378182	55
71	18	15	7	55.242553	83.793617	47	38.409091	74.378182	55
78	21	15	7	55.663636	52.263636	55	38.409091	74.378182	55
31	8	16	8	53.332927	41.608537	82	69.594286	69.471429	35
53	15	16	6	38.409091	74.378182	55	69.594286	69.471429	35
72	18	16	10	55.242553	83.793617	47	69.594286	69.471429	35
32	8	18	6	53.332927	41.608537	82	55.242553	83.793617	47
54	15	18	15	38.409091	74.378182	55	55.242553	83.793617	47
62	16	18	11	69.594286	69.471429	35	55.242553	83.793617	47
23	4	21	6	32.895833	43.518750	48	55.663636	52.263636	55
33	8	21	10	53.332927	41.608537	82	55.663636	52.263636	55
46	10	21	12	63.201613	42.351613	62	55.663636	52.263636	55
55	15	21	10	38.409091	74.378182	55	55.663636	52.263636	55
73	18	21	9	55.242553	83.793617	47	55.663636	52.263636	55
16	2	28	7	68.274359	11.915385	39	49.156863	16.282353	51
24	4	28	7	32.895833	43.518750	48	49.156863	16.282353	51
34	8	28	11	53.332927	41.608537	82	49.156863	16.282353	51
47	10	28	7	63.201613	42.351613	62	49.156863	16.282353	51
81	21	28	7	55.663636	52.263636	55	49.156863	16.282353	51
48	15	1	6	38.409091	74.378182	55	7.124000	50.324000	25
13	2	10	11	68.274359	11.915385	39	63.201613	42.351613	62
29	8	10	23	53.332927	41.608537	82	63.201613	42.351613	62
60	16	10	10	69.594286	69.471429	35	63.201613	42.351613	62
70	18	10	7	55.242553	83.793617	47	63.201613	42.351613	62
77	21	10	14	55.663636	52.263636	55	63.201613	42.351613	62

In [76]:

#plot the pitch

pitch = pitch = Pitch(pitch_type='statsbomb', orientation='horizontal',
              pitch_color='#22312b', line_color='#c7d5cc', figsize=(13, 8),
              constrained_layout=True, tight_layout=False)
fig, ax = pitch.draw()

#plot the arrows

arrows = pitch.arrows(1.2*pass_between.x,.8*pass_between.y,1.2*pass_between.x_end,.8*pass_between.y_end,
                     width = 5, headwidth = 5, color = 'w', ax = ax, zorder = 1, alpha = .5)

#plot the nodes

nodes = pitch.scatter(1.2*average_locations.x,.8*average_locations.y,
                     s = 300, color = '#d3d3d3', edgecolors = 'black', linewidth = 2.5, alpha = 1, zorder = 1, ax=ax)

#make annotations

In [ ]: