Quantifying Non-Shot Chances¶

This notebook uses the Logistic Regression Expected Goals model trained using StatsBomb event data on Metrica's sample tracking data (match 2) to calculate the expected goals and the non-shot expected goals of the match.

Starts with loading and preprocessing the Metrica tracking and event data by transforming the coordinates to be cinsistent with StatsBombs.

Traditional match stats are calculated from the event data.

Possessions are defined and identified using Metrica's possession definitions with event types 'BALL LOST' and 'BALL OUT'.

Expected goals and non-shot expected goals are calculated for both Home and Away teams across all frames of the match.

Expected goals are summed for each shot taken whilst non-shot expected goals include possessions which no shots were attempted. In these possessions, the maximum non-shot expected goal value is taken across the whole possession and attributed to that possession.

Outcomes are at the bottom. This allows possessions where shots weren't taken to still be accounted for in post match statistics, further analysis such as difference between expected goals and non-shot expected goals can be extrapolated upon and discussed. But having the ability to start the conversation is great.

In [1]:

import pandas as pd
import numpy as np


import Metrica_IO as mio
import Metrica_Functions_TLMAnalytics as mfun

import warnings
warnings.filterwarnings('ignore')

pd.options.display.max_rows = 100
pd.options.display.max_columns = 100

Load Metrica Tracking Data¶

Loading data using MetricaIO, need to convert to StatsBomb consistent coordinates since xG model was trained using that coordinate system. The coefficients of the Logistic Regressions are dependent on those coordinate distances and angles.

In [2]:

# Load in Metrica events and tracking data using Metrica_IO from Friends of Tracking
events = mio.read_event_data('sample-data-master/data',2)
tracking_home = mio.tracking_data('sample-data-master/data'
                                  , 2
                                  , 'Home')
tracking_away = mio.tracking_data('sample-data-master/data/'
                                  , 2
                                  , 'Away')
tracking_home.head(3)

Reading team: home
Reading team: away

Out[2]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y
Frame
1	1	0.04	0.94275	0.50413	0.64812	0.28605	0.67752	0.42803	0.69882	0.55606	0.64348	0.76253	0.23339	0.25869	0.60269	0.47397	0.58508	0.53017	0.53163	0.75084	0.50637	0.30137	0.50194	0.61123	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	1	0.08	0.94275	0.50413	0.64812	0.28605	0.67752	0.42803	0.69882	0.55606	0.64348	0.76253	0.23339	0.25869	0.60269	0.47397	0.58508	0.53017	0.53163	0.75084	0.50637	0.30137	0.50194	0.61123	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	1	0.12	0.94275	0.50413	0.64812	0.28605	0.67752	0.42803	0.69882	0.55606	0.64348	0.76253	0.23339	0.25869	0.60269	0.47397	0.58508	0.53017	0.53163	0.75084	0.50637	0.30137	0.50194	0.61123	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

In [3]:

# Convert Metrica coordinates to StatsBomb coordinates
events = mfun.to_statsbomb_coordinates(events)
tracking_home = mfun.to_statsbomb_coordinates(tracking_home)
tracking_away = mfun.to_statsbomb_coordinates(tracking_away)
tracking_home.tail(3)

Out[3]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y
Frame
141154	2	5646.16	7.542	39.0056	46.1988	67.7152	27.0696	46.916	32.5704	54.2328	NaN	NaN	48.7104	73.5984	46.7832	34.9368	43.4592	35.0256	NaN	NaN	33.9864	51.1464	NaN	NaN	49.0044	68.4408	61.2348	4.2312	33.6372	50.6192	NaN	NaN
141155	2	5646.20	7.542	39.0056	46.1988	67.7152	27.0696	46.916	32.5704	54.2328	NaN	NaN	48.7104	73.5984	46.8012	35.0336	43.4592	35.0256	NaN	NaN	33.9192	51.1240	NaN	NaN	49.0044	68.4408	61.2336	4.1664	33.6708	50.6352	NaN	NaN
141156	2	5646.24	7.542	39.0056	46.1988	67.7152	27.0696	46.916	32.5704	54.2328	NaN	NaN	48.7104	73.5984	46.8084	35.0000	43.4592	35.0256	NaN	NaN	33.8412	51.1016	NaN	NaN	49.0044	68.4408	61.2348	4.1232	33.6780	50.8072	NaN	NaN

In [4]:

# reverse direction of play in the second half so that home team is always attacking from right->left
tracking_home,tracking_away,events = mfun.to_single_playing_direction(tracking_home,tracking_away,events)
tracking_home.tail(3)

Out[4]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y
Frame
141154	2	5646.16	112.458	40.9944	73.8012	12.2848	92.9304	33.084	87.4296	25.7672	NaN	NaN	71.2896	6.4016	73.2168	45.0632	76.5408	44.9744	NaN	NaN	86.0136	28.8536	NaN	NaN	70.9956	11.5592	58.7652	75.7688	86.3628	29.3808	NaN	NaN
141155	2	5646.20	112.458	40.9944	73.8012	12.2848	92.9304	33.084	87.4296	25.7672	NaN	NaN	71.2896	6.4016	73.1988	44.9664	76.5408	44.9744	NaN	NaN	86.0808	28.8760	NaN	NaN	70.9956	11.5592	58.7664	75.8336	86.3292	29.3648	NaN	NaN
141156	2	5646.24	112.458	40.9944	73.8012	12.2848	92.9304	33.084	87.4296	25.7672	NaN	NaN	71.2896	6.4016	73.1916	45.0000	76.5408	44.9744	NaN	NaN	86.1588	28.8984	NaN	NaN	70.9956	11.5592	58.7652	75.8768	86.3220	29.1928	NaN	NaN

Match Stats¶

Traditional stats such as goals, shots and passes are calculated here.

In [5]:

shots = events[events['Type'] == 'SHOT']
print('There are {} shots in total.'.format(shots.shape[0]))
home_shots = shots[shots['Team'] == 'Home']
print('There are {} shots by the Home team.'.format(home_shots.shape[0]))
home_goals = home_shots[home_shots['Subtype'].str.contains('GOAL')]
print('There are {} goals by the Home team.'.format(home_goals.shape[0]))
away_shots = shots[shots['Team'] == 'Away']
print('There are {} shots by the Away team.'.format(away_shots.shape[0]))
away_goals = away_shots[away_shots['Subtype'].str.contains('GOAL')]
print('There are {} goals by the Away team.'.format(away_goals.shape[0]))
passes = events[events['Type'] == 'PASS']
print('There are {} passes in total.'.format(passes.shape[0]))
home_passes = passes[passes['Team'] == 'Home']
print('There are {} passes by the Home team.'.format(home_passes.shape[0]))
away_passes = passes[passes['Team'] == 'Away']
print('There are {} passes by the Away team.'.format(away_passes.shape[0]))

There are 24 shots in total.
There are 13 shots by the Home team.
There are 3 goals by the Home team.
There are 11 shots by the Away team.
There are 2 goals by the Away team.
There are 964 passes in total.
There are 543 passes by the Home team.
There are 421 passes by the Away team.

Define Possessions¶

Possessions change everytime a BALL OUT or BALL LOST occurs. Separate for Home / Away.

In [6]:

events['Type'].value_counts()

Out[6]:

PASS              964
CHALLENGE         311
RECOVERY          248
BALL LOST         233
SET PIECE          80
BALL OUT           49
SHOT               24
FAULT RECEIVED     20
CARD                6
Name: Type, dtype: int64

Every time the event type is 'BALL LOST' or 'BALL OUT', the possession ends. The count increases to identify a new possession for the next event.

In [7]:

home_events = events[events['Team'] == 'Home']
home_events.loc[:, 'Type_shift'] = home_events.loc[:, 'Type'].shift(1)
possession_count = 1
for index, row in home_events.iterrows():
    if (home_events.loc[index, 'Type_shift'] == 'BALL LOST') | (home_events.loc[index, 'Type_shift'] == 'BALL OUT'):
        possession_count += 1
        home_events.loc[index, 'Possession'] = possession_count
    else:
        home_events.loc[index, 'Possession'] = possession_count
    possession_count = possession_count
home_events.drop(['Type_shift'], axis=1, inplace=True)
home_events.tail(10)

Out[7]:

	Team	Type	Subtype	Period	Start Frame	Start Time [s]	End Frame	End Time [s]	From	To	Start X	Start Y	End X	End Y	Possession
1919	Home	CHALLENGE	GROUND-FAULT-WON	2	137429	5497.16	137429	5497.16	Player1	NaN	82.8	4.0	NaN	NaN	146.0
1920	Home	FAULT RECEIVED	NaN	2	137429	5497.16	137429	5497.16	Player1	NaN	82.8	4.0	NaN	NaN	146.0
1921	Home	SET PIECE	FREE KICK	2	138250	5530.00	138250	5530.00	Player3	NaN	NaN	NaN	NaN	NaN	146.0
1922	Home	BALL LOST	INTERCEPTION	2	138250	5530.00	138334	5533.36	Player3	NaN	80.4	4.8	34.8	42.4	146.0
1924	Home	CHALLENGE	AERIAL-FAULT-WON	2	138334	5533.36	138334	5533.36	Player7	NaN	36.0	40.0	NaN	NaN	147.0
1926	Home	SET PIECE	FREE KICK	2	139891	5595.64	139891	5595.64	Player9	NaN	NaN	NaN	NaN	NaN	147.0
1927	Home	SHOT	ON TARGET-SAVED	2	139891	5595.64	139925	5597.00	Player9	NaN	36.0	41.6	2.4	41.6	147.0
1931	Home	CHALLENGE	AERIAL-LOST	2	140180	5607.20	140180	5607.20	Player1	NaN	82.8	24.0	NaN	NaN	147.0
1933	Home	RECOVERY	NaN	2	140211	5608.44	140211	5608.44	Player12	NaN	79.2	25.6	NaN	NaN	147.0
1934	Home	BALL OUT	CLEARANCE	2	140265	5610.60	140298	5611.92	Player12	NaN	80.4	12.8	70.8	-3.2	147.0

Useful to record the frames which each possession starts and ends to merge onto tracking data later on.

In [8]:

home_possession_frames = home_events.groupby(['Possession']).agg({'Start Frame': 'min', 'End Frame':'max'}).reset_index()
print('There are {} possessions for the Home team.'.format(home_possession_frames.shape[0]))
home_possession_frames.head(5)

There are 147 possessions for the Home team.

Out[8]:

	Possession	Start Frame	End Frame
0	1.0	504	534
1	2.0	1084	1121
2	3.0	1431	1611
3	4.0	1672	1724
4	5.0	1874	1918

Repeat for Away team.

In [9]:

away_events = events[events['Team'] == 'Away']
away_events.loc[:, 'Type_shift'] = away_events.loc[:, 'Type'].shift(1)
possession_count = 1
for index, row in away_events.iterrows():
    if (away_events.loc[index, 'Type_shift'] == 'BALL LOST') | (away_events.loc[index, 'Type_shift'] == 'BALL OUT'):
        possession_count += 1
        away_events.loc[index, 'Possession'] = possession_count
    else:
        away_events.loc[index, 'Possession'] = possession_count
    possession_count = possession_count
away_events.drop(['Type_shift'], axis=1, inplace=True)
away_possession_frames = away_events.groupby(['Possession']).agg({'Start Frame': 'min', 'End Frame':'max'}).reset_index()
print('There are {} possessions for the Away team.'.format(away_possession_frames.shape[0]))
away_events.tail(10)

There are 136 possessions for the Away team.

Out[9]:

	Team	Type	Subtype	Period	Start Frame	Start Time [s]	End Frame	End Time [s]	From	To	Start X	Start Y	End X	End Y	Possession
1902	Away	CHALLENGE	AERIAL-WON	2	136060	5442.40	136060	5442.40	Player 26	NaN	114.0	36.8	NaN	NaN	135.0
1903	Away	SHOT	HEAD-OFF TARGET-OUT	2	136060	5442.40	136078	5443.12	Player 26	NaN	114.0	36.8	123.6	32.8	135.0
1909	Away	CHALLENGE	GROUND-LOST	2	137026	5481.04	137026	5481.04	Player21	NaN	55.2	61.6	NaN	NaN	135.0
1918	Away	CHALLENGE	GROUND-FAULT-LOST	2	137416	5496.64	137416	5496.64	Player20	NaN	81.6	5.6	NaN	NaN	135.0
1923	Away	CHALLENGE	AERIAL-FAULT-LOST	2	138327	5533.08	138327	5533.08	Player17	NaN	34.8	38.4	NaN	NaN	135.0
1925	Away	CARD	YELLOW	2	138445	5537.80	138445	5537.80	Player17	NaN	NaN	NaN	NaN	NaN	135.0
1928	Away	RECOVERY	SAVED	2	139926	5597.04	139926	5597.04	Player25	NaN	1.2	41.6	NaN	NaN	135.0
1929	Away	PASS	NaN	2	140010	5600.40	140058	5602.32	Player25	Player21	15.6	40.0	43.2	64.0	135.0
1930	Away	BALL LOST	INTERCEPTION	2	140122	5604.88	140183	5607.32	Player21	NaN	55.2	61.6	85.2	19.2	135.0
1932	Away	CHALLENGE	AERIAL-LOST	2	140183	5607.32	140183	5607.32	Player24	NaN	81.6	24.0	NaN	NaN	136.0

Calculate xG for all Frames¶

For each frame in the tracking data, create features for calculating the xG for each frame according to each specified attacking team.

Eg. for Home team, the distance calculated will be the distance to the opponents (playing_direction = left to right). And opposite for the Away team.

Home¶

In [10]:

# Create distance, angle and defender features
home_tracking_frames = mfun.create_features(tracking_home = tracking_home,
                                       tracking_away = tracking_away,
                                       start_frame = 0,
                                       number_frames = tracking_home.shape[0],
                                       field_dimen = (120, 80),
                                       attacking_team = 'Home')
# Calculate xG using logistic regression model
home_tracking_frames = mfun.create_xG(home_tracking_frames)
# Calculate non-shot xG based on if the attacking team is near the ball
home_tracking_frames = mfun.create_non_shot_xG(home_tracking_frames, attacking_team = 'Home')
home_tracking_frames.head()

Out[10]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y	Away_25_x	Away_25_y	Away_15_x	Away_15_y	Away_16_x	Away_16_y	Away_17_x	Away_17_y	Away_18_x	Away_18_y	Away_19_x	Away_19_y	Away_20_x	Away_20_y	Away_21_x	Away_21_y	Away_22_x	Away_22_y	Away_23_x	Away_23_y	Away_24_x	Away_24_y	Away_26_x	Away_26_y	distance	angle	distance_nearest_defender	xG	distance_nearest_attacker	non_shot_xG
0	1	0.04	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1	1	0.08	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	1	0.12	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	1	0.16	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.9048	59.3040	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4032	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
4	1	0.20	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.8040	59.3032	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.3396	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

Away¶

In [11]:

# Create distance, angle and defender features
away_tracking_frames = mfun.create_features(tracking_home = tracking_home,
                                           tracking_away = tracking_away,
                                           start_frame = 0,
                                           number_frames = tracking_home.shape[0],
                                           field_dimen = (120, 80),
                                           attacking_team = 'Away',
                                           playing_direction = 'left-right')
# Calculate xG using logistic regression model
away_tracking_frames = mfun.create_xG(away_tracking_frames)
# Calculate non-shot xG based on if the attacking team is near the ball
away_tracking_frames = mfun.create_non_shot_xG(away_tracking_frames, attacking_team = 'Away')
away_tracking_frames.head()

Out[11]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y	Away_25_x	Away_25_y	Away_15_x	Away_15_y	Away_16_x	Away_16_y	Away_17_x	Away_17_y	Away_18_x	Away_18_y	Away_19_x	Away_19_y	Away_20_x	Away_20_y	Away_21_x	Away_21_y	Away_22_x	Away_22_y	Away_23_x	Away_23_y	Away_24_x	Away_24_y	Away_26_x	Away_26_y	distance	angle	distance_nearest_defender	xG	distance_nearest_attacker	non_shot_xG
0	1	0.04	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1	1	0.08	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	1	0.12	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	1	0.16	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.9048	59.3040	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4032	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
4	1	0.20	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.8040	59.3032	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.3396	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

Merge on Possession Indentifier by Frames¶

Respective Home / Away Possession columns identify the frames which that possession cover. So if Home Possession == 1 for a row, then it's the first possession for the Home team. Similar for Away team.

This is to ensure we only include the xG values for frames where the team has possession of the ball. We can also use the possessions to find the highest xG_available to the team during that possession.

Home¶

In [12]:

# Map the Possession identifiers to the respective tracking frames
for index1, row1 in home_tracking_frames.iterrows():
    for index2, row2 in home_possession_frames.iterrows():
        if (index1 > home_possession_frames.loc[index2, 'Start Frame']) & (index1 < home_possession_frames.loc[index2, 'End Frame']):
            home_tracking_frames.loc[index1, 'Home Possession'] = home_possession_frames.loc[index2, 'Possession']
            next
home_tracking_frames.head()

Out[12]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y	Away_25_x	Away_25_y	Away_15_x	Away_15_y	Away_16_x	Away_16_y	Away_17_x	Away_17_y	Away_18_x	Away_18_y	Away_19_x	Away_19_y	Away_20_x	Away_20_y	Away_21_x	Away_21_y	Away_22_x	Away_22_y	Away_23_x	Away_23_y	Away_24_x	Away_24_y	Away_26_x	Away_26_y	distance	angle	distance_nearest_defender	xG	distance_nearest_attacker	non_shot_xG	Home Possession
0	1	0.04	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1	1	0.08	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	1	0.12	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	1	0.16	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.9048	59.3040	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4032	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
4	1	0.20	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.8040	59.3032	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.3396	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

Away¶

In [13]:

# Map the Possession identifiers to the respective tracking frames
for index1, row1 in away_tracking_frames.iterrows():
    for index2, row2 in away_possession_frames.iterrows():
        if (index1 > away_possession_frames.loc[index2, 'Start Frame']) & (index1 < away_possession_frames.loc[index2, 'End Frame']):
            away_tracking_frames.loc[index1, 'Away Possession'] = away_possession_frames.loc[index2, 'Possession']
            next
away_tracking_frames.head()

Out[13]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y	Away_25_x	Away_25_y	Away_15_x	Away_15_y	Away_16_x	Away_16_y	Away_17_x	Away_17_y	Away_18_x	Away_18_y	Away_19_x	Away_19_y	Away_20_x	Away_20_y	Away_21_x	Away_21_y	Away_22_x	Away_22_y	Away_23_x	Away_23_y	Away_24_x	Away_24_y	Away_26_x	Away_26_y	distance	angle	distance_nearest_defender	xG	distance_nearest_attacker	non_shot_xG	Away Possession
0	1	0.04	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1	1	0.08	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	1	0.12	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	28.0068	59.3048	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4692	14.5232	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	1	0.16	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.9048	59.3040	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.4032	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
4	1	0.20	113.13	39.6696	77.7744	57.116	81.3024	45.7576	83.8584	35.5152	77.2176	18.9976	27.8040	59.3032	72.3228	42.0824	70.2096	37.5864	63.7956	19.9328	60.7644	55.8904	60.2328	31.1016	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1.4616	38.5896	39.3396	14.5224	36.8808	35.8544	38.682	48.8368	43.806	55.1608	59.4768	6.9088	47.7744	42.2256	51.9756	54.9168	59.8668	81.8104	60.2124	39.884	59.4204	51.0128	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

Include a Shot Indicator¶

The shot indicator will allow the tracking data to know when an actual shot took place. For possessions where shots took place, the xG of the shot can be used to override and other non_shot_xG since the shot takes priority.

In [14]:

# Create shot indicator and merge to tracking data
home_shots['Shot Indicator'] = 1
home_tracking_frames = pd.merge(home_tracking_frames, home_shots[['Start Time [s]', 'Shot Indicator']], how='left', left_on='Time [s]', right_on='Start Time [s]')
home_tracking_frames.drop(['Start Time [s]'], axis=1, inplace=True)
home_tracking_frames[home_tracking_frames['Shot Indicator'] == 1]

Out[14]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y	Away_25_x	Away_25_y	Away_15_x	Away_15_y	Away_16_x	Away_16_y	Away_17_x	Away_17_y	Away_18_x	Away_18_y	Away_19_x	Away_19_y	Away_20_x	Away_20_y	Away_21_x	Away_21_y	Away_22_x	Away_22_y	Away_23_x	Away_23_y	Away_24_x	Away_24_y	Away_26_x	Away_26_y	distance	angle	distance_nearest_defender	number_blocking_defenders	xG	distance_nearest_attacker	non_shot_xG	Home Possession	Shot Indicator
4418	1	176.76	89.2344	43.1432	12.7356	72.4264	41.0100	58.8976	48.4584	40.9808	29.9100	31.9096	15.5784	61.7808	28.9692	45.8976	10.8684	40.0832	15.0732	37.1216	5.1240	61.5696	6.8004	41.7656	NaN	NaN	NaN	NaN	NaN	NaN	15.2712	61.7568	1.2396	43.2752	6.7500	35.9824	5.9232	41.4272	5.3892	49.6744	6.8652	60.6816	11.5932	38.6544	9.8928	53.6088	11.0784	44.4784	18.1284	60.6552	37.4964	57.2328	32.2008	35.1632	NaN	NaN	26.581345	0.159782	3.062207	2.0	0.019134	0.308136	0.019134	7.0	1.0
12201	1	488.08	89.4348	42.1136	14.8872	60.0056	38.8380	58.7016	52.3488	39.1992	28.7604	28.3064	8.0796	48.0784	37.3908	44.5304	8.0688	48.1480	9.9564	37.8424	16.4064	17.3600	6.3096	35.6648	NaN	NaN	NaN	NaN	NaN	NaN	6.5580	36.5160	2.3316	40.3192	8.3508	33.7896	6.5988	35.7232	6.4284	46.4688	10.9176	57.9616	13.4808	34.9488	12.9936	48.7032	15.0960	40.7152	19.1748	51.6248	38.5152	58.7088	28.0848	29.6264	NaN	NaN	7.426010	0.854438	0.793849	1.0	0.330727	0.886704	0.330727	16.0	1.0
16483	1	659.36	101.1312	41.6400	66.6408	61.2456	65.7336	47.3912	74.8644	34.2560	75.7188	26.3896	27.5664	56.9176	58.0536	45.8032	62.2416	39.9184	55.2300	27.8584	17.5884	39.0392	12.9468	53.2672	NaN	NaN	NaN	NaN	NaN	NaN	13.0092	53.2904	4.1676	42.8440	14.7300	37.9832	11.2320	43.6728	11.4984	51.4880	36.6696	58.9584	63.9660	31.2696	20.4060	46.3936	35.5920	38.3232	73.2060	64.7120	77.6340	44.1040	75.4080	55.1432	NaN	NaN	18.597688	0.278761	2.351842	2.0	0.048740	0.066573	0.048740	21.0	1.0
27344	1	1093.80	87.1824	41.7352	18.1668	64.4352	39.3312	50.3256	10.2048	32.7296	57.2208	41.7944	44.1408	34.9896	17.1288	39.9992	8.9988	30.4144	14.9388	17.1008	26.0916	67.1160	8.3376	38.5528	NaN	NaN	NaN	NaN	NaN	NaN	10.3872	33.1424	3.8220	38.0144	24.4308	60.7608	9.9252	34.1216	8.0184	39.2304	13.2540	51.6048	14.8188	40.5552	18.6240	58.1008	12.8640	43.2800	28.3416	41.1200	21.6264	44.7488	11.0736	29.7216	NaN	NaN	12.446710	0.492677	1.082717	1.0	0.142581	0.451302	0.142581	33.0	1.0
29753	1	1190.16	92.1264	38.9536	19.3500	54.7016	49.6296	51.9928	55.6032	36.8200	24.5328	23.6600	15.0516	38.2616	42.5604	36.0304	13.9284	28.5000	12.5928	14.6424	28.0356	31.7720	10.1076	41.2696	NaN	NaN	NaN	NaN	NaN	NaN	24.6168	23.6496	3.0540	37.4104	12.6384	19.9912	13.5480	29.2376	12.0768	38.3864	11.8140	41.3768	18.0384	22.6400	20.9016	36.9336	18.8028	27.2640	22.4664	54.2024	47.0148	54.0720	48.4500	38.4328	NaN	NaN	29.552029	0.206422	6.655422	1.0	0.031972	0.084641	0.031972	35.0	1.0
56078	1	2243.16	86.8812	44.1416	13.6692	68.6040	42.4776	52.8176	50.6280	39.8784	34.0152	33.7248	12.5904	47.0608	35.8164	47.9080	12.4176	31.7128	20.8920	56.4576	28.3920	65.9024	11.3292	37.6912	NaN	NaN	NaN	NaN	NaN	NaN	12.6684	47.7416	2.1936	41.4696	9.7428	33.0792	9.6372	38.3760	8.3124	45.5016	11.7240	63.5288	15.8124	32.6072	18.2160	44.2304	14.0940	53.2208	20.8188	60.2800	40.1676	34.5504	30.3804	53.3064	NaN	NaN	14.846573	0.420855	4.898197	1.0	0.155117	0.685254	0.155117	65.0	1.0
67066	1	2682.68	91.7532	39.0696	23.9244	60.1184	53.9220	48.9056	53.6376	26.2224	32.9388	25.3240	23.7204	24.2776	42.2424	44.4088	15.6588	18.5792	20.3196	43.0264	14.0760	37.3648	9.5532	36.5432	NaN	NaN	NaN	NaN	NaN	NaN	15.6792	18.4888	4.0104	36.1312	13.9896	20.2760	9.3360	33.5616	10.3080	38.5128	9.6936	45.1272	14.6256	23.3424	18.3372	29.7848	17.7864	37.2848	20.6772	50.4344	50.5416	26.2856	23.3520	32.2592	NaN	NaN	26.618960	0.163513	2.459437	1.0	0.024509	0.092673	0.024509	79.0	1.0
73982	2	2959.32	91.3824	39.2240	29.4516	51.5744	55.9548	48.0952	62.1876	31.5896	55.9188	26.2144	4.7496	30.3312	33.2496	39.5912	32.1756	27.8720	7.2216	16.0280	6.3564	41.2136	17.8644	37.2440	NaN	NaN	NaN	NaN	NaN	NaN	6.3708	40.8568	2.7444	41.3576	4.2372	22.4352	4.8828	31.6624	3.7728	37.2168	6.7716	43.0744	14.7588	28.0568	11.0328	31.0512	10.0524	37.1088	23.1012	47.5288	54.4308	51.0536	56.1504	27.5168	NaN	NaN	6.428157	1.030677	2.253528	1.0	0.467064	0.357090	0.467064	89.0	1.0
98879	2	3955.20	86.3016	38.5376	51.9936	34.6272	42.8700	39.0688	9.6252	44.9400	NaN	NaN	32.3616	19.2896	14.3040	35.4184	33.3216	25.5128	15.0228	20.9504	5.3808	15.2384	4.4640	34.9560	23.9040	44.0848	NaN	NaN	NaN	NaN	4.6152	33.9928	1.8072	36.4120	6.3540	30.7448	8.3532	40.7712	6.1692	38.4368	7.9884	45.2904	13.1844	33.2592	16.2024	47.7312	9.8808	39.5720	NaN	NaN	21.3288	29.5368	36.1884	14.3280	6.1344	17.9936	7.575389	0.654387	3.684146	1.0	0.335298	0.974995	0.335298	112.0	1.0
117217	2	4688.72	93.6876	40.1840	15.8316	75.2896	49.1016	59.0504	53.8308	42.0976	NaN	NaN	13.6968	41.5352	32.8848	51.2432	40.6488	30.0432	NaN	NaN	12.6096	48.7112	10.2204	37.9640	16.8984	61.4096	24.9816	68.9048	NaN	NaN	12.8364	50.4360	2.1888	40.7768	10.1400	38.4248	10.1352	38.4816	9.4632	47.5456	12.4680	69.9824	12.9000	64.4872	19.1904	48.0384	17.6412	56.5608	NaN	NaN	48.9984	58.9608	35.8188	46.9456	20.9604	37.5144	16.543375	0.346205	4.442172	1.0	0.113531	1.739647	NaN	131.0	1.0
121026	2	4841.08	91.3128	40.2960	35.7120	58.2232	58.8360	45.7520	20.4252	25.7336	NaN	NaN	31.7268	14.2688	40.8936	32.0384	59.1588	31.2504	NaN	NaN	24.0180	69.8560	17.9856	32.1200	25.6464	46.3376	44.3664	43.0208	NaN	NaN	25.8120	46.2784	4.7952	41.0040	18.0672	28.5584	17.8620	33.7064	21.0048	45.7616	19.2600	56.6736	26.3580	45.7536	36.2004	41.4320	26.1072	40.7344	NaN	NaN	47.7360	54.1264	55.6632	33.1264	32.5908	22.4032	26.564594	0.266267	0.757318	1.0	0.023571	0.175864	0.023571	136.0	1.0
124335	2	4973.44	94.3608	38.6968	42.3276	53.3024	64.3848	44.0480	68.6820	32.1432	NaN	NaN	32.8656	16.2472	53.7192	34.7920	64.3380	24.4680	NaN	NaN	8.1456	40.3680	-3.9684	23.1144	13.0908	47.9296	4.3320	35.3000	NaN	NaN	8.1456	40.3680	0.2244	37.4368	9.4896	25.3984	-1.2660	26.6296	5.4288	34.2952	6.8448	41.2848	10.4388	43.9848	28.1112	37.3976	6.0432	38.4512	NaN	NaN	54.0588	48.8016	65.9460	22.1352	37.6608	19.5336	8.153908	0.842822	1.591415	1.0	0.331111	0.000000	NaN	137.0	1.0
139890	2	5595.64	89.5548	39.0832	43.4292	30.7312	53.9796	49.4448	14.8056	26.2432	NaN	NaN	17.4324	27.3952	45.3240	39.9208	15.8436	54.1128	NaN	NaN	35.7576	41.5592	NaN	NaN	29.4336	33.0984	31.7184	48.5024	45.1224	62.1672	35.7576	41.5592	0.9948	39.6032	15.9228	29.7464	15.3060	27.5984	18.4644	39.8696	14.7072	54.8600	16.7748	50.9856	23.8248	47.1288	19.5660	43.4496	NaN	NaN	23.9496	41.5544	36.3408	31.1816	46.1400	53.8056	35.791578	0.203488	10.393974	2.0	0.018879	0.000000	NaN	147.0	1.0

In [15]:

# Create shot indicator and merge to tracking data
away_shots['Shot Indicator'] = 1
away_tracking_frames = pd.merge(away_tracking_frames, away_shots[['Start Time [s]', 'Shot Indicator']], how='left', left_on='Time [s]', right_on='Start Time [s]')
away_tracking_frames.drop(['Start Time [s]'], axis=1, inplace=True)
away_tracking_frames[away_tracking_frames['Shot Indicator'] == 1]

Out[15]:

	Period	Time [s]	Home_11_x	Home_11_y	Home_1_x	Home_1_y	Home_2_x	Home_2_y	Home_3_x	Home_3_y	Home_4_x	Home_4_y	Home_5_x	Home_5_y	Home_6_x	Home_6_y	Home_7_x	Home_7_y	Home_8_x	Home_8_y	Home_9_x	Home_9_y	Home_10_x	Home_10_y	Home_12_x	Home_12_y	Home_13_x	Home_13_y	Home_14_x	Home_14_y	ball_x	ball_y	Away_25_x	Away_25_y	Away_15_x	Away_15_y	Away_16_x	Away_16_y	Away_17_x	Away_17_y	Away_18_x	Away_18_y	Away_19_x	Away_19_y	Away_20_x	Away_20_y	Away_21_x	Away_21_y	Away_22_x	Away_22_y	Away_23_x	Away_23_y	Away_24_x	Away_24_y	Away_26_x	Away_26_y	distance	angle	distance_nearest_defender	number_blocking_defenders	xG	distance_nearest_attacker	non_shot_xG	Away Possession	Shot Indicator
18514	1	740.60	115.7292	41.3944	101.7096	51.4880	102.6804	45.6096	104.2860	34.5560	101.4156	18.2360	94.1424	44.6888	94.0284	40.5592	91.1124	29.8064	87.3444	7.3488	80.8764	51.5000	69.0252	31.6600	NaN	NaN	NaN	NaN	NaN	NaN	95.3916	52.0408	26.5776	40.1832	73.7928	13.0480	63.0900	30.1120	65.1696	37.2776	77.9616	57.6632	95.6904	20.1792	97.3188	41.2296	76.5504	35.1056	95.5584	52.0760	106.2144	38.1336	103.0992	49.0856	NaN	NaN	27.396245	0.239509	6.342138	0.0	0.055348	0.170474	0.055348	26.0	1.0
53048	1	2121.96	116.9964	39.2760	109.1556	42.4840	109.7688	31.2960	111.4464	30.9216	106.7136	21.7816	99.8724	30.6280	97.8324	41.4744	99.0972	22.0168	88.8564	58.6744	70.8384	29.1272	69.0648	40.7312	NaN	NaN	NaN	NaN	NaN	NaN	110.4768	43.5784	24.3168	42.4752	80.6364	13.6480	67.1796	27.2304	63.7668	44.9872	64.6680	61.3024	101.6724	21.8384	85.1256	34.1976	71.6448	45.7160	94.4760	52.7136	107.8152	30.7568	110.5464	43.5840	NaN	NaN	10.173312	0.658078	1.715599	0.0	0.293708	0.069825	0.293708	62.0	1.0
63361	1	2534.48	119.6184	37.8928	108.7056	49.2408	110.2896	58.3304	111.5364	41.9384	113.8632	34.4856	107.9868	40.0192	109.4280	45.0960	111.4728	40.4592	101.0352	58.7056	71.1408	24.7664	59.8560	40.1104	NaN	NaN	NaN	NaN	NaN	NaN	111.8136	40.6848	12.5148	43.7640	68.9568	22.0584	57.3408	30.7000	58.9320	41.1952	62.4420	55.6864	109.1964	31.5000	91.6428	37.8504	81.4476	47.8312	111.2712	63.6040	111.7860	40.5560	102.4452	43.8400	NaN	NaN	8.214992	0.836034	0.408705	1.0	0.295756	0.131724	0.295756	71.0	1.0
64771	1	2590.88	118.6536	39.5288	108.0276	50.3304	108.7104	41.7848	105.5148	38.0216	108.5532	37.5944	97.7232	43.4792	94.8216	37.7832	98.7408	33.5928	91.9584	51.7032	84.5772	52.1408	64.2132	35.7504	NaN	NaN	NaN	NaN	NaN	NaN	92.3496	35.6200	20.9700	41.0080	82.0140	16.4920	62.3688	29.9392	60.7644	42.1768	110.3040	39.4992	110.2524	34.3472	92.3712	35.6464	77.7204	35.6136	89.3280	49.4040	77.7000	23.0112	102.3516	65.0192	NaN	NaN	27.995161	0.256770	3.284847	0.0	0.037192	0.034110	0.037192	73.0	1.0
69886	2	2795.48	118.2540	41.0528	107.0988	61.2120	110.2344	52.8184	107.8584	39.8696	108.8448	35.5288	98.3520	64.0488	104.3760	53.7040	104.2800	46.5400	99.2052	45.7536	75.6612	56.5632	70.0488	39.9408	NaN	NaN	NaN	NaN	NaN	NaN	94.4976	52.8360	25.4388	40.2296	83.4084	24.3944	67.3500	39.6136	71.5368	56.0128	103.2228	68.9864	100.5588	26.1696	85.2900	42.3160	94.4988	52.7880	110.2020	66.6744	109.1784	48.0176	108.2184	34.7400	NaN	NaN	28.550574	0.228615	8.504228	1.0	0.045788	0.048015	0.045788	77.0	1.0
86190	2	3447.64	118.9272	41.1416	110.2620	44.1624	109.1400	34.3856	109.7580	36.4784	112.7628	38.5080	106.6056	46.0120	107.9280	41.0328	111.1560	47.7968	114.7332	64.3152	100.0380	39.2408	108.9084	56.8624	NaN	NaN	NaN	NaN	NaN	NaN	100.6596	45.8616	40.8108	40.6200	93.0864	28.1600	108.6024	31.7000	110.1012	34.5176	83.8056	52.5504	96.7020	43.5840	101.4264	66.4248	108.0024	51.4328	115.1136	71.4016	104.9304	43.2512	112.8504	40.5792	NaN	NaN	20.209142	0.343913	5.947902	0.0	0.125456	4.566187	NaN	90.0	1.0
90164	2	3606.60	117.8256	36.0600	97.8828	47.4856	114.2388	32.6272	106.7592	22.9080	109.8708	27.5520	110.4948	43.1968	109.3404	33.9176	102.6252	29.9360	73.8012	21.6016	103.5780	38.3576	55.3764	26.6400	NaN	NaN	NaN	NaN	NaN	NaN	112.8816	26.5600	17.4936	39.6232	71.8992	17.9728	59.6748	32.5728	56.8284	48.8296	76.0368	58.1536	112.9464	26.5808	99.5568	49.4936	76.7976	42.1960	NaN	NaN	107.9160	49.6400	99.9048	24.9312	113.5740	32.4112	15.208722	0.234547	3.170013	0.0	0.127539	0.068056	0.127539	91.0	1.0
111757	2	4470.32	119.2752	39.9336	104.9100	62.4872	117.1308	53.5912	117.5400	42.9600	NaN	NaN	108.6228	35.5400	111.5436	42.3448	113.7768	36.1216	NaN	NaN	95.1612	52.6920	72.2556	34.9448	111.0648	51.5024	114.0744	45.2552	NaN	NaN	109.6128	31.2592	29.0244	39.4032	86.0148	22.4856	71.0316	31.9936	66.1524	44.6920	121.0188	54.3520	99.4464	62.9528	104.7312	45.9600	87.6168	43.6288	NaN	NaN	117.8844	47.9832	109.5936	31.1280	117.1932	37.8976	13.575548	0.418328	4.393785	2.0	0.124502	0.132597	0.124502	120.0	1.0
115008	2	4600.36	119.0316	41.1160	101.5236	50.4768	101.8308	47.2080	102.2148	30.2440	NaN	NaN	101.2728	31.6992	97.7556	45.3744	99.3024	32.6424	NaN	NaN	95.7048	36.5312	101.2224	25.9176	101.4912	48.9832	101.1036	27.7488	NaN	NaN	106.0740	41.7048	29.5752	38.1336	91.9236	30.6560	100.7460	49.2504	65.8416	42.5920	77.0952	49.1272	96.2040	41.6240	106.0740	41.7048	97.5420	38.7176	NaN	NaN	99.5928	47.6608	100.8948	32.5512	101.8176	28.8968	14.029961	0.506883	6.949098	0.0	0.291622	0.000000	NaN	120.0	1.0
132569	2	5302.80	118.6140	40.4328	109.8048	30.8808	109.9668	29.4800	108.9072	32.1056	NaN	NaN	107.5620	42.0824	108.5196	32.7824	109.2684	31.5272	NaN	NaN	102.8796	43.0448	NaN	NaN	108.4716	33.8584	108.3960	38.7736	108.480	47.4840	98.2176	30.2880	30.9600	42.3808	94.9140	33.4280	108.0216	36.3912	109.6212	36.9440	98.2176	30.2880	108.6828	36.0584	108.6444	33.9464	97.9332	42.9760	NaN	NaN	106.2960	48.7168	108.2868	42.8752	109.5972	29.6752	23.849442	0.279384	10.599681	0.0	0.134575	0.000000	NaN	130.0	1.0
136059	2	5442.40	118.7184	39.7176	113.8944	34.5312	113.9988	45.9784	114.7884	38.9888	NaN	NaN	108.2520	38.0960	108.5676	45.8416	107.3700	58.0152	NaN	NaN	86.2080	53.1856	NaN	NaN	94.4352	36.9800	105.0480	48.4744	112.716	56.2368	113.9304	36.7648	43.2720	39.9992	102.7572	23.0456	73.7988	36.8320	80.5620	53.3936	99.9312	42.7688	114.7512	44.6920	93.1032	40.6800	90.7572	54.8608	NaN	NaN	108.5700	66.6752	113.7768	30.3336	113.9412	36.7088	6.877977	0.918300	2.233890	0.0	0.490506	0.057032	0.490506	135.0	1.0

Sum xG and Non Shot xG¶

Sum the shot expected goals and the non-shot expected goals using xG and non_shot_xG respectively for the possessions that the team has.

Home¶

In [16]:

# Sum the xG for the shots
home_shot_xg = home_tracking_frames[home_tracking_frames['Shot Indicator'] == 1][['xG', 'Home Possession']]
print('Total xG for all shots: {}.'.format(home_shot_xg['xG'].sum()))
home_shot_xg

Total xG for all shots: 2.0422338503137234.

Out[16]:

	xG	Home Possession
4418	0.019134	7.0
12201	0.330727	16.0
16483	0.048740	21.0
27344	0.142581	33.0
29753	0.031972	35.0
56078	0.155117	65.0
67066	0.024509	79.0
73982	0.467064	89.0
98879	0.335298	112.0
117217	0.113531	131.0
121026	0.023571	136.0
124335	0.331111	137.0
139890	0.018879	147.0

In [18]:

# Sum the non-shot xG for all possessions
home_non_shot_xG = home_tracking_frames.groupby('Home Possession').max()['non_shot_xG']
print('Total non-shot xG for all possessions: {}.'.format(home_non_shot_xG.sum()))
home_non_shot_xG

Total non-shot xG for all possessions: 6.471366602506773.

Out[18]:

Home Possession
1.0      0.000173
2.0      0.000373
3.0      0.000130
4.0      0.000085
5.0      0.000184
           ...   
143.0         NaN
144.0    0.000018
145.0    0.000009
146.0    0.001678
147.0    0.014475
Name: non_shot_xG, Length: 146, dtype: float64

In [20]:

# If a possession had a shot, then attribute the xG to that possession
home_xG_combined = home_non_shot_xG.copy()
row_count=0
for possession in list(home_shot_xg['Home Possession']):
    for index in home_xG_combined.index:
        if possession == index:
            home_xG_combined[index] = home_shot_xg['xG'].iloc[row_count]
    row_count+=1
print('Combined xG and non-shot xG for all possessions: {}.'.format(home_xG_combined.sum()))
home_xG_combined

Combined xG and non-shot xG for all possessions: 5.3129587422610225.

Out[20]:

Home Possession
1.0      0.000173
2.0      0.000373
3.0      0.000130
4.0      0.000085
5.0      0.000184
           ...   
143.0         NaN
144.0    0.000018
145.0    0.000009
146.0    0.001678
147.0    0.018879
Name: non_shot_xG, Length: 146, dtype: float64

Away¶

In [22]:

away_shot_xg = away_tracking_frames[away_tracking_frames['Shot Indicator'] == 1][['xG', 'Away Possession']]
print('Total xG for all shots: {}.'.format(away_shot_xg['xG'].sum()))
away_shot_xg

Total xG for all shots: 2.0219916974382524.

Out[22]:

	xG	Away Possession
18514	0.055348	26.0
53048	0.293708	62.0
63361	0.295756	71.0
64771	0.037192	73.0
69886	0.045788	77.0
86190	0.125456	90.0
90164	0.127539	91.0
111757	0.124502	120.0
115008	0.291622	120.0
132569	0.134575	130.0
136059	0.490506	135.0

In [23]:

away_non_shot_xG = away_tracking_frames.groupby('Away Possession').max()['non_shot_xG']
print('Total non-shot xG for all possessions: {}.'.format(away_non_shot_xG.sum()))
away_non_shot_xG[:10]

Total non-shot xG for all possessions: 3.8990711322285203.

Out[23]:

Away Possession
1.0     0.002531
2.0     0.000991
3.0     0.002739
4.0     0.003080
5.0     0.007574
6.0     0.000847
7.0     0.000031
8.0     0.000782
9.0     0.000069
10.0    0.000391
Name: non_shot_xG, dtype: float64

In [24]:

away_xG_combined = away_non_shot_xG.copy()
row_count=0
for possession in list(away_shot_xg['Away Possession']):
    for index in away_xG_combined.index:
        if possession == index:
            away_xG_combined[index] = away_shot_xg['xG'].iloc[row_count]
    row_count+=1
print('Combined xG and non-shot xG for all possessions: {}.'.format(away_xG_combined.sum()))
away_xG_combined

Combined xG and non-shot xG for all possessions: 3.4645735037493606.

Out[24]:

Away Possession
1.0      0.002531
2.0      0.000991
3.0      0.002739
4.0      0.003080
5.0      0.007574
           ...   
131.0    0.000900
132.0    0.016164
133.0    0.015819
134.0    0.006342
135.0    0.490506
Name: non_shot_xG, Length: 134, dtype: float64

In [ ]: