Data Match Analysing¶

Introduction¶

its' a data gathering most of clubs and players around countries in different leagues around world in host country from season 2008 to 2016

Questions¶

1- How many leagues in data?
2- Which league has a high scoring matches?
3- what's the most scoring club in most scoring league?
4- what's the players coordinates for the most scoring league?
5- Is there a relation between players coordinates and chance to win?
6- Is there a relation between numbers of players and scoring more gools?

In [1]:
#import packages
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from subprocess import call
call(['python', '-m', 'nbconvert', 'Investigate_a_Dataset.ipynb'])
Out[1]:
1
In [2]:
#read CSV file
df = pd.read_csv('match.csv/Match.csv')

#set display options
pd.set_option('display.max_columns', 85)
df.shape
Out[2]:
(25979, 85)
In [3]:
df.head(10)
Out[3]:
id country_id league_id season stage date match_api_id home_team_api_id away_team_api_id home_team_goal away_team_goal home_player_X1 home_player_X2 home_player_X3 home_player_X4 home_player_X5 home_player_X6 home_player_X7 home_player_X8 home_player_X9 home_player_X10 home_player_X11 away_player_X1 away_player_X2 away_player_X3 away_player_X4 away_player_X5 away_player_X6 away_player_X7 away_player_X8 away_player_X9 away_player_X10 away_player_X11 home_player_Y1 home_player_Y2 home_player_Y3 home_player_Y4 home_player_Y5 home_player_Y6 home_player_Y7 home_player_Y8 home_player_Y9 home_player_Y10 home_player_Y11 away_player_Y1 away_player_Y2 away_player_Y3 away_player_Y4 away_player_Y5 away_player_Y6 away_player_Y7 away_player_Y8 away_player_Y9 away_player_Y10 away_player_Y11 home_player_1 home_player_2 home_player_3 home_player_4 home_player_5 home_player_6 home_player_7 home_player_8 home_player_9 home_player_10 home_player_11 away_player_1 away_player_2 away_player_3 away_player_4 away_player_5 away_player_6 away_player_7 away_player_8 away_player_9 away_player_10 away_player_11 goal shoton shotoff foulcommit card cross corner possession
0 1 Belgium Belgium Jupiler League 2008/2009 1 17/08/2008 0:00 492473 KRC Genk Beerschot AC 1 1 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
1 2 Belgium Belgium Jupiler League 2008/2009 1 16/08/2008 0:00 492474 SV Zulte-Waregem Sporting Lokeren 0 0 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
2 3 Belgium Belgium Jupiler League 2008/2009 1 16/08/2008 0:00 492475 KSV Cercle Brugge RSC Anderlecht 0 3 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
3 4 Belgium Belgium Jupiler League 2008/2009 1 17/08/2008 0:00 492476 KAA Gent RAEC Mons 5 0 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
4 5 Belgium Belgium Jupiler League 2008/2009 1 16/08/2008 0:00 492477 FCV Dender EH Standard de Liège 1 3 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
5 6 Belgium Belgium Jupiler League 2008/2009 1 24/09/2008 0:00 492478 KV Mechelen Club Brugge KV 1 1 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
6 7 Belgium Belgium Jupiler League 2008/2009 1 16/08/2008 0:00 492479 KSV Roeselare KV Kortrijk 2 2 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
7 8 Belgium Belgium Jupiler League 2008/2009 1 16/08/2008 0:00 492480 Tubize Royal Excel Mouscron 1 2 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
8 9 Belgium Belgium Jupiler League 2008/2009 1 16/08/2008 0:00 492481 KVC Westerlo Sporting Charleroi 1 0 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
9 10 Belgium Belgium Jupiler League 2008/2009 10 01/11/2008 0:00 492564 Club Brugge KV KV Kortrijk 4 1 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN

Cleaning Data¶

In [4]:
#know how many duplicated and nan data
df.duplicated().sum(), df.isnull().sum()
Out[4]:
(0,
 id                0
 country_id        0
 league_id         0
 season            0
 stage             0
               ...  
 foulcommit    11762
 card          11762
 cross         11762
 corner        11762
 possession    11762
 Length: 85, dtype: int64)

Set suitable null values¶

first fill null values with median to resist outliers for quantitative data
second fill null values with (no data) value for categorical data

In [5]:
#make data frame for only null floats values to facilitate gain median
float_columns = df.select_dtypes(include=['float64'])
#replace null values with median
df.fillna(float_columns.median(), inplace= True)
In [6]:
df.fillna({'corner' : "no corners", 'card' : 'no cards', 'cross' : 'no cross','goal' : df['home_team_goal'] + df['away_team_goal'], 'foulcommit' : 'no foul', 'possession' : 'no data', 'shotoff' : 'no data', 'shoton' : 'no data'}, inplace= True)
In [7]:
#see if cleaning work
df.sample()
Out[7]:
id country_id league_id season stage date match_api_id home_team_api_id away_team_api_id home_team_goal away_team_goal home_player_X1 home_player_X2 home_player_X3 home_player_X4 home_player_X5 home_player_X6 home_player_X7 home_player_X8 home_player_X9 home_player_X10 home_player_X11 away_player_X1 away_player_X2 away_player_X3 away_player_X4 away_player_X5 away_player_X6 away_player_X7 away_player_X8 away_player_X9 away_player_X10 away_player_X11 home_player_Y1 home_player_Y2 home_player_Y3 home_player_Y4 home_player_Y5 home_player_Y6 home_player_Y7 home_player_Y8 home_player_Y9 home_player_Y10 home_player_Y11 away_player_Y1 away_player_Y2 away_player_Y3 away_player_Y4 away_player_Y5 away_player_Y6 away_player_Y7 away_player_Y8 away_player_Y9 away_player_Y10 away_player_Y11 home_player_1 home_player_2 home_player_3 home_player_4 home_player_5 home_player_6 home_player_7 home_player_8 home_player_9 home_player_10 home_player_11 away_player_1 away_player_2 away_player_3 away_player_4 away_player_5 away_player_6 away_player_7 away_player_8 away_player_9 away_player_10 away_player_11 goal shoton shotoff foulcommit card cross corner possession
20966 20967 Scotland Scotland Premier League 2013/2014 3 17/08/2013 0:00 1474143 Hibernian Dundee United 1 1 1.0 2.0 4.0 6.0 8.0 2.0 4.0 6.0 8.0 4.0 6.0 1.0 2.0 4.0 6.0 8.0 2.0 4.0 6.0 8.0 4.0 6.0 1.0 3.0 3.0 3.0 3.0 7.0 7.0 7.0 7.0 10.0 10.0 1.0 3.0 3.0 3.0 3.0 7.0 7.0 7.0 7.0 10.0 10.0 22962.0 300527.0 23507.0 112123.0 32630.0 35972.0 41432.0 32618.0 32559.0 23104.0 189075.0 13471.0 162548.0 25131.0 166423.0 357880.0 36610.0 237563.0 243160.0 34447.0 172186.0 32465.0 2 no data no data no foul no cards no cross no corners no data

let's make EDA(Exploratory Data Analysis)¶

Q1¶

How many leagues and some analysis?

In [8]:
df['league_id'].unique()
Out[8]:
array(['Belgium Jupiler League', 'England Premier League',
       'France Ligue 1', 'Germany 1. Bundesliga', 'Italy Serie A',
       'Netherlands Eredivisie', 'Poland Ekstraklasa',
       'Portugal Liga ZON Sagres', 'Scotland Premier League',
       'Spain LIGA BBVA', 'Switzerland Super League'], dtype=object)

Q2¶

Which league has a high scoring matches?

In [37]:
#function to put plot data
def plotting_data(xlabel, ylabel, title):
    [plot for plot in [mm.set_xlabel(xlabel), mm.set_ylabel(ylabel), plt.title(title), plt.legend()]]
In [93]:
#ploting leagues to find which league has a high intensity of scoring
fig, mm = plt.subplots(figsize =(8,6))
hex_colors = ["#F1A9A0", "#D2527F", "#2ECC71", "#E26A6A", "#9B59B6", "#F89406", "#3A539B", "#1F3A93", "#F62459", "#26A65B", "#663399"]
plt.barh(df['league_id'].unique(), df.groupby('league_id')['home_team_goal'].sum() + df.groupby('league_id')['away_team_goal'].sum(),
         alpha= 0.5, color= hex_colors, edgecolor= 'white')
plotting_data('Goals summation', 'Leagues', 'Summation goals for each league')

No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.

Q3¶

what's the most scoring club in spain league?

In [39]:
#plotting spain league clubs to show the most scoring one
fig, mm = plt.subplots(figsize =(10,6))
plt.barh(df[df['league_id'] == 'Spain LIGA BBVA']['away_team_api_id'], df[df['league_id'] == 'Spain LIGA BBVA']['away_team_goal'],  color= '#FF7F50', label= 'away', edgecolor= 'white')
plt.barh(df[df['league_id'] == 'Spain LIGA BBVA']['home_team_api_id'], df[df['league_id'] == 'Spain LIGA BBVA']['home_team_goal'],  color= '#FFD700', label= 'home', edgecolor= 'white')
plotting_data('Goals summation', 'Clubs', 'Goals summation for each club')

Q4¶

what's the players coordinates for the most scoring club in Spain league?

In [43]:
#function make scatter plot for club coordinates
def scatter_players_coordinates(club, category, color, label):
    plt.scatter(df[df['home_team_api_id'] == club][[f'{category}_player_X{player_num}' for player_num in range(1, 12)]],
                df[df['home_team_api_id'] == club][[f'{category}_player_Y{player_num}' for player_num in range(1, 12)]],
                edgecolor= 'black', color= color, label= label)
In [46]:
#plottin spain Real Madrid players coordinates
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates('Real Madrid CF', 'home', '#FFFF00', 'Real Madrid home')
scatter_players_coordinates('Real Madrid CF', 'away', '#dec3c3', 'Real Madrid away')
plotting_data('Coordinates', 'Players', 'Real Madrid coordinates')

Q5¶

Is there a relation between players coordinates and chance to win?

In [50]:
#plotting spain Real Madrid players coordinates
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates('Real Madrid CF', 'home', '#FFFF00', 'Real Madrid home')
scatter_players_coordinates('Real Madrid CF', 'away', '#dec3c3', 'Real Madrid away')
#annotate important coordinate points[(2, 8), (3, 5), (8, 8), (7, 5)]
[plt.annotate('most important coordinates', xy= point, arrowprops= {'facecolor' : 'red', 'shrink' : 0.002}, xytext= (8, 9)) for point in [(2, 8), (3, 5), (8, 8), (7, 5)]]
plotting_data('Coordinates', 'Players', 'Real Madrid coordinates')

According to last plot -> If we connect these four points will make trapezium shape, if this shape concentrated with coordinates emphasizly these four points [(2, 8), (3, 5), (8, 8), (7, 5)] that make chance to score more goals increase

Let's prove the relation with any another club coordinate (any another club will be less goals than Real Madrid look at Q3 above)

Let's compare any club players coordinte with most scoring club in all data frame

In [51]:
#plotting Recreativo players coordinates
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates('RC Recreativo', 'home', '#051e3e', 'Recreativo home')
scatter_players_coordinates('RC Recreativo', 'away', '#b3cde0', 'Recreativo away')
#annotate important coordinate points[(2, 8), (3, 5), (8, 8), (7, 5)]
[plt.annotate('most important coordinates', xy= point, arrowprops= {'facecolor' : 'red', 'shrink' : 0.002}, xytext= (8, 9)) for point in [(2, 8), (3, 5), (8, 8), (7, 5)]]
plotting_data('Coordinates', 'Players', 'Recreativo coordinates')

This plot shows that 'Recreativo' club has no coordinates in four important points [(2, 8), (3, 5), (8, 8), (7, 5)], and low concentration coordinate in this trapezium shape so it has low number of scoring goals

Let's make another comparison for the closest club to 'Real Madrid; in goals is 'Barcelona' with deference 2 goals according to Q3

In [57]:
#plotting of Barcelona
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates('FC Barcelona', 'home', '#fe4a49', 'Barcelona home')
scatter_players_coordinates('FC Barcelona', 'away', '#f4f4f8', 'Barcelona away')
plotting_data('Coordinates', 'Players', 'Barcelona coordinates')
#plotting of Real Madrid
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates('Real Madrid CF', 'home', '#FFFF00', 'Real Madrid home')
scatter_players_coordinates('Real Madrid CF', 'away', '#dec3c3', 'Real Madrid away')
plotting_data('Coordinates', 'Players', 'Real Madrid coordinates')
#plotting to compare
fig, mm = plt.subplots(figsize =(8,6))
#Barcelona coordinate
scatter_players_coordinates('FC Barcelona', 'home', '#fe4a49', 'Barcelona home')
scatter_players_coordinates('FC Barcelona', 'away', '#f4f4f8', 'Barcelona away')
#Real Madrid coordinate
scatter_players_coordinates('Real Madrid CF', 'home', '#FFFF00', 'Real Madrid home')
scatter_players_coordinates('Real Madrid CF', 'away', '#dec3c3', 'Real Madrid away')
#annotate important coordinate points[(2, 8), (3, 5), (8, 8), (7, 5)]
[plt.annotate('most important coordinates', xy= point, arrowprops= {'facecolor' : 'red', 'shrink' : 0.002}, xytext= (8, 9)) for point in [(2, 8), (3, 5), (8, 8), (7, 5)]]
plotting_data('Coordinates', 'Players', 'Comparing between Real Madrid and Barcelona coordinates')

According to last comparison we conclude that

1- the concentration of coordinates for 'Real Madrid' is more than 'Barcelona' by 2 points ('Real Madrid' = 22 point, 'Barcelona' = 20 point) in trapezium[(2, 8), (3, 5), (8, 8), (7, 5)]
2- 'Real Madrid' take control at [(3, 5), (7, 5)] points while 'Barcelona' not

that make 'Real Madrid' more scorer than 'Barcelona'

Another comparison

Q6¶

Is there a relation between numbers of players and scoring more gools?

In [15]:
#function makes horizontal bar for players numbers
def barh_players_numbers(club, category, color, label):
    plt.barh([f'player_{player_num}' for player_num in range(1, 12)], 
         [df[df['home_team_api_id'] == club][f'{category}_player_{player_num}'].iloc[0]  for player_num in range(1, 12)], 
         color= color, label= label)
In [129]:
#let's make array for Real Madrid to get mean of it's players numbers
real_array = np.array([df[df['home_team_api_id'] == 'Real Madrid CF'][f'home_player_{player_num}'].iloc[0]  for player_num in range(1, 12)] + 
                [df[df['home_team_api_id'] == 'Real Madrid CF'][f'away_player_{player_num}'].iloc[0]  for player_num in range(1, 12)]).mean()
#let's make array for Barcelona to get mean of it's players numbers
barcelona_array = np.array([df[df['home_team_api_id'] == 'FC Barcelona'][f'home_player_{player_num}'].iloc[0]  for player_num in range(1, 12)] + 
                [df[df['home_team_api_id'] == 'FC Barcelona'][f'away_player_{player_num}'].iloc[0]  for player_num in range(1, 12)]).mean()
In [136]:
#plotting of Real Madrid
fig, mm = plt.subplots(figsize =(8,6))
barh_players_numbers('Real Madrid CF', 'home', '#FFFF00', 'Real Madrid home')
barh_players_numbers('Real Madrid CF', 'away', '#dec3c3', 'Real Madrid away')
plotting_data('Coordinate', 'Players', 'Real Madrid players numbers')
#plotting of Barcelona
fig, mm = plt.subplots(figsize =(8,6))
barh_players_numbers('FC Barcelona', 'home', '#fe4a49', 'Barcelona home')
barh_players_numbers('FC Barcelona', 'away', '#f4f4f8', 'Barcelona away')
plotting_data('Coordinate', 'Players', 'Barcelona players numbers')
#comparison plotting 
fig, mm = plt.subplots(figsize =(8,6))
plt.barh('Real Madrid', real_array, color = '#dec3c5', label= 'Real  mean')
plt.barh('Barcelona', barcelona_array, color= '#f5f4f8', label= 'Barcelona  mean')
plotting_data('Coordinate', 'Players', 'comaparison between players numbers')

According to last plot we can take players numbers as a second relation if the coordinate of home and away club congruent

Let's see that at Match has id (20562) two clubs are same at coordinate

In [126]:
df[df['id'] == 20562]
Out[126]:
id country_id league_id season stage date match_api_id home_team_api_id away_team_api_id home_team_goal away_team_goal home_player_X1 home_player_X2 home_player_X3 home_player_X4 home_player_X5 home_player_X6 home_player_X7 home_player_X8 home_player_X9 home_player_X10 home_player_X11 away_player_X1 away_player_X2 away_player_X3 away_player_X4 away_player_X5 away_player_X6 away_player_X7 away_player_X8 away_player_X9 away_player_X10 away_player_X11 home_player_Y1 home_player_Y2 home_player_Y3 home_player_Y4 home_player_Y5 home_player_Y6 home_player_Y7 home_player_Y8 home_player_Y9 home_player_Y10 home_player_Y11 away_player_Y1 away_player_Y2 away_player_Y3 away_player_Y4 away_player_Y5 away_player_Y6 away_player_Y7 away_player_Y8 away_player_Y9 away_player_Y10 away_player_Y11 home_player_1 home_player_2 home_player_3 home_player_4 home_player_5 home_player_6 home_player_7 home_player_8 home_player_9 home_player_10 home_player_11 away_player_1 away_player_2 away_player_3 away_player_4 away_player_5 away_player_6 away_player_7 away_player_8 away_player_9 away_player_10 away_player_11 goal shoton shotoff foulcommit card cross corner possession
20561 20562 Scotland Scotland Premier League 2011/2012 37 06/05/2012 0:00 1189737 Heart of Midlothian St. Johnstone FC 2 0 1.0 2.0 4.0 6.0 8.0 2.0 4.0 6.0 8.0 4.0 6.0 1.0 2.0 4.0 6.0 8.0 2.0 4.0 6.0 8.0 4.0 6.0 1.0 3.0 3.0 3.0 3.0 7.0 7.0 7.0 7.0 10.0 10.0 1.0 3.0 3.0 3.0 3.0 7.0 7.0 7.0 7.0 10.0 10.0 115675.0 94467.0 39223.0 37307.0 32891.0 75373.0 32553.0 32681.0 32593.0 34945.0 24449.0 14100.0 32725.0 35968.0 32484.0 23186.0 17528.0 23127.0 47575.0 32559.0 56852.0 195585.0 2 no data no data no foul no cards no cross no corners no data

We see that home team score more goals, so let's show the coordinates and players numbers

In [137]:
#plotting home team coordinates
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates(df[df['id'] == 20562]['home_team_api_id'].iloc[0], 'home', '#dec3c3', 'home team')
plotting_data('coordinates', 'players', 'home team coordinate')

fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates(df[df['id'] == 20562]['away_team_api_id'].iloc[0], 'away', '#f4f4f8', 'away team')
plotting_data('coordinates', 'players', 'home team coordinate')

Same coordinate so let's see second relation (players numbers mean)

In [138]:
#make array for home team to get players number mean
home_array = np.array([df[df['home_team_api_id'] == 'Real Madrid CF'][f'home_player_{player_num}'].iloc[0]  for player_num in range(1, 12)]).mean()
#let's make array for Barcelona to get mean of it's players numbers
away_array = np.array([df[df['home_team_api_id'] == 'FC Barcelona'][f'away_player_{player_num}'].iloc[0]  for player_num in range(1, 12)]).mean()
In [139]:
#plot comparison between players numbers to see the relation is right or not
fig, mm = plt.subplots(figsize =(8,6))
plt.barh(df[df['id'] == 20562]['home_team_api_id'].iloc[0], home_array, color = '#dec3c5', label= 'home')
plt.barh(df[df['id'] == 20562]['away_team_api_id'].iloc[0], away_array, color= '#f5f4f8', label= 'away')
plotting_data('Coordinate', 'Players', 'comaparison between players numbers')

We conclude that from last plots:

1- If players numbers of team deacrease that make more ability to score more goals

In [17]:
df.sample()
Out[17]:
id country_id league_id season stage date match_api_id home_team_api_id away_team_api_id home_team_goal away_team_goal home_player_X1 home_player_X2 home_player_X3 home_player_X4 home_player_X5 home_player_X6 home_player_X7 home_player_X8 home_player_X9 home_player_X10 home_player_X11 away_player_X1 away_player_X2 away_player_X3 away_player_X4 away_player_X5 away_player_X6 away_player_X7 away_player_X8 away_player_X9 away_player_X10 away_player_X11 home_player_Y1 home_player_Y2 home_player_Y3 home_player_Y4 home_player_Y5 home_player_Y6 home_player_Y7 home_player_Y8 home_player_Y9 home_player_Y10 home_player_Y11 away_player_Y1 away_player_Y2 away_player_Y3 away_player_Y4 away_player_Y5 away_player_Y6 away_player_Y7 away_player_Y8 away_player_Y9 away_player_Y10 away_player_Y11 home_player_1 home_player_2 home_player_3 home_player_4 home_player_5 home_player_6 home_player_7 home_player_8 home_player_9 home_player_10 home_player_11 away_player_1 away_player_2 away_player_3 away_player_4 away_player_5 away_player_6 away_player_7 away_player_8 away_player_9 away_player_10 away_player_11 goal shoton shotoff foulcommit card cross corner possession
15437 15438 Netherlands Netherlands Eredivisie 2015/2016 11 31/10/2015 0:00 1983366 De Graafschap PSV 3 6 1.0 2.0 4.0 6.0 8.0 2.0 4.0 6.0 8.0 4.0 6.0 1.0 2.0 4.0 6.0 8.0 3.0 5.0 7.0 3.0 5.0 7.0 1.0 3.0 3.0 3.0 3.0 7.0 7.0 7.0 7.0 10.0 10.0 1.0 3.0 3.0 3.0 3.0 7.0 7.0 7.0 10.0 10.0 10.0 467024.0 209418.0 586678.0 240788.0 166065.0 186923.0 298933.0 179208.0 32593.0 149211.0 248154.0 150236.0 207642.0 184821.0 70812.0 389845.0 178765.0 467495.0 31045.0 158581.0 150201.0 289441.0 <goal><value><comment>n</comment><stats><goals... <shoton><value><stats><blocked>1</blocked></st... <shotoff><value><stats><shotoff>1</shotoff></s... <foulcommit><value><stats><foulscommitted>1</f... <card><value><comment>y</comment><stats><ycard... <cross><value><stats><crosses>1</crosses></sta... <corner><value><stats><corners>1</corners></st... <possession><value><comment>29</comment><stats...
In [86]:
df[df['id'] == 9312]
fig, CL = plt.subplots(figsize =(8,6))
plt.scatter(df[df['id'] == 15438][['home_player_X1', 'home_player_X2', 'home_player_X3', 'home_player_X4', 'home_player_X5', 'home_player_X6', 'home_player_X7', 'home_player_X8', 'home_player_X9', 'home_player_X10', 'home_player_X11']],
            df[df['id'] == 15438][['home_player_Y1', 'home_player_Y2', 'home_player_Y3', 'home_player_Y4', 'home_player_Y5', 'home_player_Y6', 'home_player_Y7', 'home_player_Y8', 'home_player_Y9', 'home_player_Y10', 'home_player_Y11']],
           edgecolor= 'black', color= 'g')
CL.set_xlabel('Coordinate')
CL.set_ylabel('Players')
plt.title('home coordinates')
fig, CL = plt.subplots(figsize =(8,6))
plt.scatter(df[df['id'] == 15438][['away_player_X1', 'away_player_X2', 'away_player_X3', 'away_player_X4', 'away_player_X5', 'away_player_X6', 'away_player_X7', 'away_player_X8', 'away_player_X9', 'away_player_X10', 'away_player_X11']],
            df[df['id'] == 15438][['away_player_Y1', 'away_player_Y2', 'away_player_Y3', 'away_player_Y4', 'away_player_Y5', 'away_player_Y6', 'away_player_Y7', 'away_player_Y8', 'away_player_Y9', 'away_player_Y10', 'away_player_Y11']],
           edgecolor= 'black', color= 'r')
CL.set_xlabel('Coordinate')
CL.set_ylabel('Players')
plt.title('away coordinates')
Out[86]:
Text(0.5, 1.0, 'away coordinates')
In [87]:
df[df['id'] == 9312]
fig, CL = plt.subplots(figsize =(8,6))
plt.scatter(df[df['id'] == 15438][['home_player_X1', 'home_player_X2', 'home_player_X3', 'home_player_X4', 'home_player_X5', 'home_player_X6', 'home_player_X7', 'home_player_X8', 'home_player_X9', 'home_player_X10', 'home_player_X11']],
            df[df['id'] == 15438][['home_player_Y1', 'home_player_Y2', 'home_player_Y3', 'home_player_Y4', 'home_player_Y5', 'home_player_Y6', 'home_player_Y7', 'home_player_Y8', 'home_player_Y9', 'home_player_Y10', 'home_player_Y11']],
           edgecolor= 'black', color= 'g')
CL.set_xlabel('Coordinate')
CL.set_ylabel('Players')
plt.title('home coordinates')
fig, CL = plt.subplots(figsize =(8,6))
plt.scatter(df[df['id'] == 15438][['away_player_X1', 'away_player_X2', 'away_player_X3', 'away_player_X4', 'away_player_X5', 'away_player_X6', 'away_player_X7', 'away_player_X8', 'away_player_X9', 'away_player_X10', 'away_player_X11']],
            df[df['id'] == 15438][['away_player_Y1', 'away_player_Y2', 'away_player_Y3', 'away_player_Y4', 'away_player_Y5', 'away_player_Y6', 'away_player_Y7', 'away_player_Y8', 'away_player_Y9', 'away_player_Y10', 'away_player_Y11']],
           edgecolor= 'black', color= 'r')
CL.set_xlabel('Coordinate')
CL.set_ylabel('Players')
plt.title('away coordinates')
Out[87]:
Text(0.5, 1.0, 'away coordinates')

Conclusion¶

>

We have clean our data and make plots to see the relation between scoring more goals and player coordinate and number to see the coorelation key to score more goals

First let's answer Question we have asked¶

Q1 -> ['Belgium Jupiler League', 'England Premier League', 'France Ligue 1', 'Germany 1. Bundesliga', 'Italy Serie A', 'Netherlands Eredivisie', 'Poland Ekstraklasa', 'Portugal Liga ZON Sagres', 'Scotland Premier League', 'Spain LIGA BBVA', 'Switzerland Super League']e'],

Q2 -> 'Spain LIGA BBVA'
Q3 -> 'Real Madrid CF'
Q4

In [140]:
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates('Real Madrid CF', 'home', '#FFFF00', 'Real Madrid home')
scatter_players_coordinates('Real Madrid CF', 'away', '#dec3c3', 'Real Madrid away')
plotting_data('Coordinates', 'Players', 'Real Madrid coordinates')

Q5 -> the relation between players coordinates and scoring goals is directly proportional at four points [(2, 8), (3, 5), (8, 8), (7, 5)] which make trapezium shape

In [141]:
#plotting spain Real Madrid players coordinates
fig, mm = plt.subplots(figsize =(8,6))
scatter_players_coordinates('Real Madrid CF', 'home', '#FFFF00', 'Real Madrid home')
scatter_players_coordinates('Real Madrid CF', 'away', '#dec3c3', 'Real Madrid away')
#annotate important coordinate points[(2, 8), (3, 5), (8, 8), (7, 5)]
[plt.annotate('most important coordinates', xy= point, arrowprops= {'facecolor' : 'red', 'shrink' : 0.002}, xytext= (8, 9)) for point in [(2, 8), (3, 5), (8, 8), (7, 5)]]
plotting_data('Coordinates', 'Players', 'Real Madrid coordinates')

Q6 -> the relation between players numbers and scoring goals is inversly proportional , you can measure it with the mean of each team players number

Second, How to increase your team goals?¶

According to last analysis you must:
1- concentrate players concentration at middle of stadium to get more control to match than another team, to make that you must be more concentrated than your rival at [(2, 8), (3, 5), (8, 8), (7, 5)] trapezium, atleast with one more point and focus on two points [(3, 5), (7, 5)] becouse they play as a defense and play maker for attackers.
2- decrease your players numbers in match as possible to be able to score more goals

Third, limitation¶

Let's be more obvious, this analysis may has a limitaion at analysis of player numbers that make it perhaps less useful, However this limitaion in analysis will not make a high distroy at all, best wishes