This project aims in deriving useful insights from greek football league 2020-2021 by mean statistics on aggregated data, as well as on individual team level. In addition it aims on evaluating bookmaker's (B365) accuracy.

Data were found on football-data.uk.com as a csv file. The site also provides a txt explaining the abbreviations in the csv columns

The project main parts are the following:

• Find bookmaker's (B365) accuracy on the whole league as well as per outcome (win,loss,draw)
• Find variables mean statistics grouped by bookmaker being right or wrong
• Find bookmaker's accurracy per outcome on individual team level
• Find mean match statistics per outcome
• Find most important variables on winning with DecisionTree feature importance and Lasso
• Find mean statistics of a specified team grouped by outcome
• Find mean statistics of a specified team with a specified opponent grouped by outcome
• Find team's most important variables on winning

0:Draw, 1:Win, 2:Loss