Using a dataset from Kaggle dataset 'Football Data: Expected Goals and Other Metrics' and utilize SPSS to find which metrics are significant to improve goal-scoring output. The technique used in this analysis included Stepwise Multiple Linear Regression (including ANOVA, homoscedasticity & multicollinearity test) to establish a causal relationship between the dependent and independent variable. Then Factor Analysis technique (correlation metric, PCA, Varimax) to reduce the dimension of an independent variable

Dataset Link : https://www.kaggle.com/slehkyi/extended-football-stats-for-european-leagues-xg

Normality of the error term distribution

where it assumes there is linear relationship between independent variable and dependent variable

Constant variance of the error terms where it assume of error term are similar across the value of independent variable

Independence of the error term where it assumes that all independent variable are not highly correlated with each other

Top 3 predictors(variable) that is the most significant to the model

Determine whether there are any statistically significant differences between the means of three groups.

Y = 0.210+0.989X1-0.049X2+0.150X3

• Y=Goal scored against opponent (scored)
• X1=Expected Goal (xG)
• X2= Passes completed within an estimated 20 yards of goal (deep)
• X3= Number of goals missed in games (missed)