Load Packages¶

In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from scipy.stats import f_oneway
from scipy.stats import shapiro
from scipy.stats import levene
from scipy.stats import pearsonr
import statsmodels.api as sm
from sklearn.preprocessing import LabelEncoder
from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score, confusion_matrix

Load Dataset¶

In [2]:
data = pd.read_excel('Strikers_performance.xlsx')
data.head()
Out[2]:
Striker_ID Nationality Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball Hold-up Play Aerial Duels Won Defensive Contribution Big Game Performance Consistency Penalty Success Rate Impact on Team Performance Off-field Conduct
0 1 Spain Left-footed No 17.483571 10.778533 34.795488 0.677836 0.166241 0.757061 50.921924 71.806409 15.682532 30.412215 6.152481 0.820314 0.922727 8.570370 11.451388
1 2 France Left-footed Yes 14.308678 13.728250 31.472436 0.544881 0.192774 0.796818 61.396150 53.726866 19.843983 26.474913 6.093172 0.803321 0.678984 3.444638 8.243689
2 3 Germany Left-footed No 18.238443 3.804297 25.417413 0.518180 0.160379 0.666869 65.863945 60.452227 20.090084 24.164116 3.408714 0.766540 0.843858 8.429491 9.506835
3 4 France Right-footed No 22.615149 9.688908 20.471443 0.599663 0.184602 0.638776 88.876877 60.511979 22.363152 44.129989 6.339820 0.611798 0.662997 6.532552 8.199653
4 5 France Left-footed Yes 13.829233 6.048072 29.887563 0.582982 0.105319 0.591485 75.565531 54.982158 13.165708 37.859323 8.465658 0.701638 0.906538 8.414915 6.665333

Data Cleaning¶

Missing values¶

In [3]:
data.isnull().sum()
Out[3]:
Striker_ID                    0
Nationality                   0
Footedness                    0
Marital Status                0
Goals Scored                  0
Assists                       0
Shots on Target               0
Shot Accuracy                 0
Conversion Rate               0
Dribbling Success             0
Movement off the Ball         6
Hold-up Play                  0
Aerial Duels Won              0
Defensive Contribution        0
Big Game Performance          2
Consistency                   0
Penalty Success Rate          5
Impact on Team Performance    0
Off-field Conduct             0
dtype: int64
In [4]:
from sklearn.impute import SimpleImputer

imputer = SimpleImputer(strategy='median')  
data[['Movement off the Ball', 
      'Big Game Performance', 
      'Penalty Success Rate']] = imputer.fit_transform(data[['Movement off the Ball', 
                                                             'Big Game Performance', 
                                                             'Penalty Success Rate']])
data.isnull().sum()
Out[4]:
Striker_ID                    0
Nationality                   0
Footedness                    0
Marital Status                0
Goals Scored                  0
Assists                       0
Shots on Target               0
Shot Accuracy                 0
Conversion Rate               0
Dribbling Success             0
Movement off the Ball         0
Hold-up Play                  0
Aerial Duels Won              0
Defensive Contribution        0
Big Game Performance          0
Consistency                   0
Penalty Success Rate          0
Impact on Team Performance    0
Off-field Conduct             0
dtype: int64

Data Types¶

In [5]:
data.dtypes
Out[5]:
Striker_ID                      int64
Nationality                    object
Footedness                     object
Marital Status                 object
Goals Scored                  float64
Assists                       float64
Shots on Target               float64
Shot Accuracy                 float64
Conversion Rate               float64
Dribbling Success             float64
Movement off the Ball         float64
Hold-up Play                  float64
Aerial Duels Won              float64
Defensive Contribution        float64
Big Game Performance          float64
Consistency                   float64
Penalty Success Rate          float64
Impact on Team Performance    float64
Off-field Conduct             float64
dtype: object
In [6]:
variables = ['Goals Scored', 'Assists', 
             'Shots on Target', 
             'Movement off the Ball', 
             'Hold-up Play', 
             'Aerial Duels Won', 
             'Defensive Contribution', 
             'Big Game Performance', 
             'Impact on Team Performance', 
             'Off-field Conduct']

for var in variables:
    data[var] = data[var].astype('int')
    
data.dtypes
Out[6]:
Striker_ID                      int64
Nationality                    object
Footedness                     object
Marital Status                 object
Goals Scored                    int64
Assists                         int64
Shots on Target                 int64
Shot Accuracy                 float64
Conversion Rate               float64
Dribbling Success             float64
Movement off the Ball           int64
Hold-up Play                    int64
Aerial Duels Won                int64
Defensive Contribution          int64
Big Game Performance            int64
Consistency                   float64
Penalty Success Rate          float64
Impact on Team Performance      int64
Off-field Conduct               int64
dtype: object
In [7]:
data.head()
Out[7]:
Striker_ID Nationality Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball Hold-up Play Aerial Duels Won Defensive Contribution Big Game Performance Consistency Penalty Success Rate Impact on Team Performance Off-field Conduct
0 1 Spain Left-footed No 17 10 34 0.677836 0.166241 0.757061 50 71 15 30 6 0.820314 0.922727 8 11
1 2 France Left-footed Yes 14 13 31 0.544881 0.192774 0.796818 61 53 19 26 6 0.803321 0.678984 3 8
2 3 Germany Left-footed No 18 3 25 0.518180 0.160379 0.666869 65 60 20 24 3 0.766540 0.843858 8 9
3 4 France Right-footed No 22 9 20 0.599663 0.184602 0.638776 88 60 22 44 6 0.611798 0.662997 6 8
4 5 France Left-footed Yes 13 6 29 0.582982 0.105319 0.591485 75 54 13 37 8 0.701638 0.906538 8 6

Exploratory Data Analysis¶

Perform descriptive analysis¶

In [8]:
round(data.describe(), 2)
Out[8]:
Striker_ID Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball Hold-up Play Aerial Duels Won Defensive Contribution Big Game Performance Consistency Penalty Success Rate Impact on Team Performance Off-field Conduct
count 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00 500.00
mean 250.50 14.52 7.60 25.26 0.60 0.20 0.70 69.28 59.33 19.04 39.47 6.43 0.74 0.80 6.52 7.57
std 144.48 4.91 2.96 7.08 0.10 0.05 0.10 10.33 10.15 4.95 9.90 1.98 0.10 0.10 2.11 2.09
min 1.00 0.00 0.00 4.00 0.31 0.05 0.40 40.00 35.00 4.00 8.00 2.00 0.46 0.53 0.00 0.00
25% 125.75 11.00 6.00 20.00 0.54 0.17 0.64 62.00 52.00 16.00 33.00 5.00 0.68 0.72 5.00 6.00
50% 250.50 15.00 8.00 25.00 0.60 0.20 0.70 69.00 60.00 19.00 39.50 6.00 0.75 0.80 6.50 8.00
75% 375.25 18.00 9.00 30.00 0.67 0.23 0.76 76.00 66.00 22.00 46.00 8.00 0.81 0.87 8.00 9.00
max 500.00 34.00 15.00 43.00 0.92 0.36 1.00 98.00 92.00 34.00 71.00 12.00 1.00 1.00 13.00 13.00

Perform percentage analysis¶

In [9]:
freq_Footedness = data['Footedness'].value_counts()
perc_Footedness = freq_Footedness/len(data['Footedness'])*100
perc_Footedness
Out[9]:
Right-footed    53.4
Left-footed     46.6
Name: Footedness, dtype: float64
In [10]:
plt.figure(figsize=(12, 6))
perc_Footedness.plot(kind='pie', autopct='%1.2f%%')
plt.title('Percentage of strikers by their footedness')
plt.ylabel('')
plt.show()
No description has been provided for this image

Which nationality strikers have the highest average number of goals scored?¶

In [11]:
goals_by_nationality = data.groupby('Nationality')['Goals Scored'].mean().sort_values(ascending=False)
round(goals_by_nationality)
Out[11]:
Nationality
Brazil     15.0
Spain      15.0
France     14.0
Germany    14.0
England    14.0
Name: Goals Scored, dtype: float64

What is the average conversion rate for players based on their footedness?¶

In [12]:
conversion_rate_by_footedness = data.groupby('Footedness')['Conversion Rate'].mean()
conversion_rate_by_footedness
Out[12]:
Footedness
Left-footed     0.198086
Right-footed    0.200592
Name: Conversion Rate, dtype: float64

What is the distribution of players' footedness across different nationalities?¶

In [13]:
footedness_by_nationality = pd.crosstab(data['Nationality'], data['Footedness'])
footedness_by_nationality
Out[13]:
Footedness Left-footed Right-footed
Nationality
Brazil 42 53
England 50 59
France 42 51
Germany 44 47
Spain 55 57
In [14]:
plt.figure(figsize=(12, 6))
sns.countplot(x='Nationality', hue='Footedness', data=data)
plt.title('Tistribution of players footedness across different nationalities')
plt.xlabel('Nationality')
plt.ylabel('Count')
plt.show()
No description has been provided for this image

Create a correlation matrix with a heatmap¶

In [15]:
num_variables = data.select_dtypes(include = ['number']).columns

correl_matrix = round(data[num_variables].corr(), 3)
correl_matrix
Out[15]:
Striker_ID Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball Hold-up Play Aerial Duels Won Defensive Contribution Big Game Performance Consistency Penalty Success Rate Impact on Team Performance Off-field Conduct
Striker_ID 1.000 0.025 0.066 0.003 0.018 0.047 -0.031 0.060 -0.036 0.016 0.012 -0.052 -0.017 -0.006 -0.027 0.103
Goals Scored 0.025 1.000 -0.071 -0.059 0.066 -0.011 0.034 0.016 0.003 -0.047 0.014 -0.004 0.045 -0.021 0.103 -0.053
Assists 0.066 -0.071 1.000 0.072 -0.020 0.028 0.050 0.011 -0.046 0.001 -0.011 0.026 -0.007 -0.081 -0.022 -0.015
Shots on Target 0.003 -0.059 0.072 1.000 -0.021 -0.044 0.010 -0.062 -0.102 -0.068 0.011 0.039 0.049 0.029 0.025 0.055
Shot Accuracy 0.018 0.066 -0.020 -0.021 1.000 -0.088 0.023 -0.018 -0.044 0.012 0.027 -0.019 0.005 0.017 0.041 0.084
Conversion Rate 0.047 -0.011 0.028 -0.044 -0.088 1.000 -0.020 0.038 0.059 0.059 -0.002 0.037 0.031 -0.049 -0.031 -0.004
Dribbling Success -0.031 0.034 0.050 0.010 0.023 -0.020 1.000 -0.056 0.002 -0.072 0.015 -0.059 -0.030 -0.054 -0.012 -0.060
Movement off the Ball 0.060 0.016 0.011 -0.062 -0.018 0.038 -0.056 1.000 -0.057 0.013 -0.006 -0.075 -0.030 -0.071 -0.046 -0.035
Hold-up Play -0.036 0.003 -0.046 -0.102 -0.044 0.059 0.002 -0.057 1.000 0.080 0.029 -0.031 0.147 0.068 -0.046 0.035
Aerial Duels Won 0.016 -0.047 0.001 -0.068 0.012 0.059 -0.072 0.013 0.080 1.000 -0.025 -0.013 0.047 -0.002 0.043 0.004
Defensive Contribution 0.012 0.014 -0.011 0.011 0.027 -0.002 0.015 -0.006 0.029 -0.025 1.000 -0.055 -0.062 0.044 -0.019 0.007
Big Game Performance -0.052 -0.004 0.026 0.039 -0.019 0.037 -0.059 -0.075 -0.031 -0.013 -0.055 1.000 -0.001 0.005 -0.002 0.053
Consistency -0.017 0.045 -0.007 0.049 0.005 0.031 -0.030 -0.030 0.147 0.047 -0.062 -0.001 1.000 0.024 -0.035 0.099
Penalty Success Rate -0.006 -0.021 -0.081 0.029 0.017 -0.049 -0.054 -0.071 0.068 -0.002 0.044 0.005 0.024 1.000 0.048 -0.010
Impact on Team Performance -0.027 0.103 -0.022 0.025 0.041 -0.031 -0.012 -0.046 -0.046 0.043 -0.019 -0.002 -0.035 0.048 1.000 0.006
Off-field Conduct 0.103 -0.053 -0.015 0.055 0.084 -0.004 -0.060 -0.035 0.035 0.004 0.007 0.053 0.099 -0.010 0.006 1.000
In [16]:
plt.figure(figsize=(18, 10))
sns.heatmap(correl_matrix, annot=True)
plt.title('Heatmap of Correlation Matrix')
plt.show()
No description has been provided for this image

Statistical Test¶

Find whether there is any significant difference in consistency rates among strikers from various nationality¶

In [17]:
# Normality test
stat, p_value = shapiro(data['Consistency'])
print('P value: ', round(p_value, 3))

P value:  0.451
In [18]:
# Filtering data
Spain = data.query('Nationality == "Spain"')['Consistency']
France = data.query('Nationality == "France"')['Consistency']
Germany = data.query('Nationality == "Germany"')['Consistency']
Brazil = data.query('Nationality == "Brazil"')['Consistency']
England = data.query('Nationality == "England"')['Consistency']
In [19]:
# Levene test for statistics
stats, p_value = levene(Spain, France, Germany, Brazil, England)
print("P value: ", round(p_value, 3))

P value:  0.808
In [20]:
# One way ANOVA

Test_stat, p_value = f_oneway(Spain, France, Germany, Brazil, England)
print("P value: ", round(p_value, 2))

P value:  0.19

Check if there is any significant correlation between strikers' Hold-up play and consistency rate¶

In [21]:
# Normality test
stat, p_value = shapiro(data['Hold-up Play'])
print('P value: ', round(p_value, 3))

P value:  0.151
In [22]:
# Linearity test
plt.figure(figsize = (10, 6))
sns.regplot(x = 'Hold-up Play', y = 'Consistency', data = data)
plt.title('Linearity between Hold-up Play and Consistency')
plt.xlabel('Hold-up Play')
plt.ylabel('Consistency')
plt.show()
No description has been provided for this image
In [23]:
# Pearson correlation
HU_play = data['Hold-up Play']
Consistency = data['Consistency']

corr, p_value = pearsonr(HU_play, Consistency)
print("Correlation coefficient: ", round(corr, 3))
print("P value: ", round(p_value, 3))

Correlation coefficient:  0.147
P value:  0.001

Check if strikers' hold-up play significantly influences their consistency rate¶

In [24]:
x = data['Hold-up Play']
y = data['Consistency']

x_constant = sm.add_constant(x)
model = sm.OLS(y, x_constant).fit()

print(model.summary())

                            OLS Regression Results                            
==============================================================================
Dep. Variable:            Consistency   R-squared:                       0.021
Model:                            OLS   Adj. R-squared:                  0.020
Method:                 Least Squares   F-statistic:                     10.93
Date:                Thu, 18 Apr 2024   Prob (F-statistic):            0.00101
Time:                        17:19:02   Log-Likelihood:                 429.97
No. Observations:                 500   AIC:                            -855.9
Df Residuals:                     498   BIC:                            -847.5
Df Model:                           1                                         
Covariance Type:            nonrobust                                         
================================================================================
                   coef    std err          t      P>|t|      [0.025      0.975]
--------------------------------------------------------------------------------
const            0.6548      0.027     24.031      0.000       0.601       0.708
Hold-up Play     0.0015      0.000      3.306      0.001       0.001       0.002
==============================================================================
Omnibus:                        1.708   Durbin-Watson:                   2.135
Prob(Omnibus):                  0.426   Jarque-Bera (JB):                1.744
Skew:                          -0.100   Prob(JB):                        0.418
Kurtosis:                       2.791   Cond. No.                         358.
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.

Feature Engineering¶

Create a new feature - Total contribution score¶

In [25]:
data['Total contribution score'] = (data['Goals Scored'] + data['Assists'] + data['Shots on Target'] + data['Dribbling Success'] + data['Aerial Duels Won'] + data['Defensive Contribution'] + data['Big Game Performance'] + data['Consistency'])
data.head()
Out[25]:
Striker_ID Nationality Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball Hold-up Play Aerial Duels Won Defensive Contribution Big Game Performance Consistency Penalty Success Rate Impact on Team Performance Off-field Conduct Total contribution score
0 1 Spain Left-footed No 17 10 34 0.677836 0.166241 0.757061 50 71 15 30 6 0.820314 0.922727 8 11 113.577376
1 2 France Left-footed Yes 14 13 31 0.544881 0.192774 0.796818 61 53 19 26 6 0.803321 0.678984 3 8 110.600139
2 3 Germany Left-footed No 18 3 25 0.518180 0.160379 0.666869 65 60 20 24 3 0.766540 0.843858 8 9 94.433410
3 4 France Right-footed No 22 9 20 0.599663 0.184602 0.638776 88 60 22 44 6 0.611798 0.662997 6 8 124.250575
4 5 France Left-footed Yes 13 6 29 0.582982 0.105319 0.591485 75 54 13 37 8 0.701638 0.906538 8 6 107.293123

Encode the Footedness and marital status by LabelEncoder¶

In [26]:
encoder = LabelEncoder()
data['Footedness'] = encoder.fit_transform(data['Footedness'])
data['Marital Status'] = encoder.fit_transform(data['Marital Status'])
data.head()
Out[26]:
Striker_ID Nationality Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball Hold-up Play Aerial Duels Won Defensive Contribution Big Game Performance Consistency Penalty Success Rate Impact on Team Performance Off-field Conduct Total contribution score
0 1 Spain 0 0 17 10 34 0.677836 0.166241 0.757061 50 71 15 30 6 0.820314 0.922727 8 11 113.577376
1 2 France 0 1 14 13 31 0.544881 0.192774 0.796818 61 53 19 26 6 0.803321 0.678984 3 8 110.600139
2 3 Germany 0 0 18 3 25 0.518180 0.160379 0.666869 65 60 20 24 3 0.766540 0.843858 8 9 94.433410
3 4 France 1 0 22 9 20 0.599663 0.184602 0.638776 88 60 22 44 6 0.611798 0.662997 6 8 124.250575
4 5 France 0 1 13 6 29 0.582982 0.105319 0.591485 75 54 13 37 8 0.701638 0.906538 8 6 107.293123

Create the dummies for Nationality and add with the data¶

In [27]:
dummies = pd.get_dummies(data['Nationality'])
processed_df = pd.concat([data, dummies], axis = 1)
processed_df = processed_df.drop('Nationality', axis = 1)
processed_df.head()
Out[27]:
Striker_ID Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball ... Consistency Penalty Success Rate Impact on Team Performance Off-field Conduct Total contribution score Brazil England France Germany Spain
0 1 0 0 17 10 34 0.677836 0.166241 0.757061 50 ... 0.820314 0.922727 8 11 113.577376 0 0 0 0 1
1 2 0 1 14 13 31 0.544881 0.192774 0.796818 61 ... 0.803321 0.678984 3 8 110.600139 0 0 1 0 0
2 3 0 0 18 3 25 0.518180 0.160379 0.666869 65 ... 0.766540 0.843858 8 9 94.433410 0 0 0 1 0
3 4 1 0 22 9 20 0.599663 0.184602 0.638776 88 ... 0.611798 0.662997 6 8 124.250575 0 0 1 0 0
4 5 0 1 13 6 29 0.582982 0.105319 0.591485 75 ... 0.701638 0.906538 8 6 107.293123 0 0 1 0 0

5 rows × 24 columns

Cluster Analysis¶

Perform KMeans clsutering¶

In [28]:
# Selecting features
x = processed_df.drop('Striker_ID', axis = 1)

# Calculating WCSS score
wcss = []

for i in range(1, 15):
    kmeans = KMeans(n_clusters = i, init = 'k-means++')
    kmeans.fit(x)
    wcss_score = kmeans.inertia_
    wcss.append(wcss_score)
In [29]:
# Plotting elbow chart
plt.figure(figsize = (12, 6))
plt.plot(range(1, 15), wcss, marker = 'o')
plt.title('Elbow methods')
plt.xlabel('Number of Clusters (k)')
plt.ylabel('WCSS')
plt.show()
No description has been provided for this image
In [30]:
# Building KMeans with k = 2
final_km = KMeans(n_clusters = 2)
final_km.fit(x)

# Generating labels
labels = final_km.labels_
labels
Out[30]:
array([1, 1, 1, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0,
       1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 1, 1,
       1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 1, 0, 1, 1, 0, 1,
       1, 1, 0, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0,
       1, 1, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 1, 0,
       1, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0,
       0, 0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1,
       0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1,
       1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 1, 0, 0, 0,
       1, 0, 1, 1, 0, 0, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0,
       0, 0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 0, 0, 0, 1, 0, 1, 0, 1, 0,
       1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1,
       1, 1, 1, 1, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1,
       1, 0, 1, 1, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0,
       0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 1, 0, 1, 0, 1, 1, 0, 0, 0, 1,
       0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 1, 0,
       0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 0, 0, 1, 1,
       0, 1, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1,
       1, 1, 0, 0, 1, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1,
       0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1,
       0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1,
       0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 0, 0, 1,
       0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1], dtype=int32)
In [31]:
# Adding labels
processed_df['Clusters'] = labels
processed_df.head()
Out[31]:
Striker_ID Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball ... Penalty Success Rate Impact on Team Performance Off-field Conduct Total contribution score Brazil England France Germany Spain Clusters
0 1 0 0 17 10 34 0.677836 0.166241 0.757061 50 ... 0.922727 8 11 113.577376 0 0 0 0 1 1
1 2 0 1 14 13 31 0.544881 0.192774 0.796818 61 ... 0.678984 3 8 110.600139 0 0 1 0 0 1
2 3 0 0 18 3 25 0.518180 0.160379 0.666869 65 ... 0.843858 8 9 94.433410 0 0 0 1 0 1
3 4 1 0 22 9 20 0.599663 0.184602 0.638776 88 ... 0.662997 6 8 124.250575 0 0 1 0 0 0
4 5 0 1 13 6 29 0.582982 0.105319 0.591485 75 ... 0.906538 8 6 107.293123 0 0 1 0 0 1

5 rows × 25 columns

In [32]:
# Checking clusters
round(processed_df.groupby('Clusters')['Total contribution score'].mean(), 2)
Out[32]:
Clusters
0    123.39
1    101.90
Name: Total contribution score, dtype: float64
In [33]:
# Assigning meaningfull names
mapping = {0:'Best strikers', 1:'Regular strikers'}
processed_df['Strikers types'] = processed_df['Clusters'].map(mapping)
In [34]:
# Deleting the Clusters variable
processed_df = processed_df.drop('Clusters', axis = 1)
processed_df.head()
Out[34]:
Striker_ID Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball ... Penalty Success Rate Impact on Team Performance Off-field Conduct Total contribution score Brazil England France Germany Spain Strikers types
0 1 0 0 17 10 34 0.677836 0.166241 0.757061 50 ... 0.922727 8 11 113.577376 0 0 0 0 1 Regular strikers
1 2 0 1 14 13 31 0.544881 0.192774 0.796818 61 ... 0.678984 3 8 110.600139 0 0 1 0 0 Regular strikers
2 3 0 0 18 3 25 0.518180 0.160379 0.666869 65 ... 0.843858 8 9 94.433410 0 0 0 1 0 Regular strikers
3 4 1 0 22 9 20 0.599663 0.184602 0.638776 88 ... 0.662997 6 8 124.250575 0 0 1 0 0 Best strikers
4 5 0 1 13 6 29 0.582982 0.105319 0.591485 75 ... 0.906538 8 6 107.293123 0 0 1 0 0 Regular strikers

5 rows × 25 columns

Data Preprocessing for ML¶

New feature mapping¶

In [35]:
mapping = {'Best strikers':1, 'Regular strikers': 0}
processed_df['Strikers types'] = processed_df['Strikers types'].map(mapping)
processed_df.head()
Out[35]:
Striker_ID Footedness Marital Status Goals Scored Assists Shots on Target Shot Accuracy Conversion Rate Dribbling Success Movement off the Ball ... Penalty Success Rate Impact on Team Performance Off-field Conduct Total contribution score Brazil England France Germany Spain Strikers types
0 1 0 0 17 10 34 0.677836 0.166241 0.757061 50 ... 0.922727 8 11 113.577376 0 0 0 0 1 0
1 2 0 1 14 13 31 0.544881 0.192774 0.796818 61 ... 0.678984 3 8 110.600139 0 0 1 0 0 0
2 3 0 0 18 3 25 0.518180 0.160379 0.666869 65 ... 0.843858 8 9 94.433410 0 0 0 1 0 0
3 4 1 0 22 9 20 0.599663 0.184602 0.638776 88 ... 0.662997 6 8 124.250575 0 0 1 0 0 1
4 5 0 1 13 6 29 0.582982 0.105319 0.591485 75 ... 0.906538 8 6 107.293123 0 0 1 0 0 0

5 rows × 25 columns

Selecting features¶

In [36]:
x = processed_df.drop(['Striker_ID', 'Strikers types'], axis = 1)
y = processed_df['Strikers types']

Scaling features¶

In [37]:
scaler = StandardScaler()
scaled_x = scaler.fit_transform(x)
scaled_x
Out[37]:
array([[-1.07047781, -1.03252879,  0.5050467 , ..., -0.47801802,
        -0.47169258,  1.86125917],
       [-1.07047781,  0.968496  , -0.10638998, ...,  2.09197134,
        -0.47169258, -0.53727069],
       [-1.07047781, -1.03252879,  0.70885893, ..., -0.47801802,
         2.12002488, -0.53727069],
       ...,
       [-1.07047781,  0.968496  , -0.10638998, ..., -0.47801802,
        -0.47169258, -0.53727069],
       [ 0.93416229,  0.968496  , -0.9216389 , ..., -0.47801802,
        -0.47169258, -0.53727069],
       [-1.07047781, -1.03252879, -1.32926335, ..., -0.47801802,
        -0.47169258, -0.53727069]])

Train test split¶

In [38]:
from sklearn.model_selection import train_test_split

x_train, x_test, y_train, y_test = train_test_split(scaled_x, y, test_size = 0.2, random_state = 42)

Predictive Classification Analytics¶

Build a logistic regression machine learning model to predict strikers type¶

In [39]:
# Model training
lgr_model = LogisticRegression()
lgr_model.fit(x_train, y_train)

#Prediction
y_lgr_pred = lgr_model.predict(x_test)

# Evaluation
accuracy_lgr = accuracy_score(y_test, y_lgr_pred)
print(accuracy_lgr*100,'%')

97.0 %
In [40]:
# Creating confusion matrix
conf_matrix_lgr = confusion_matrix(y_test, y_lgr_pred)

# Plotting confusion matrix
plt.figure(figsize = (12, 6))
sns.heatmap(conf_matrix_lgr, annot = True, fmt = "d", cmap = "Blues")
plt.title('Confusion Matrix for LGR model')
plt.xlabel('Predicted')
plt.ylabel('Actual')
plt.show()
No description has been provided for this image

Thank you!¶