The PL Predictions is a weekly, online game that I manage. For more information on how the game is played and managed click here. This season I have decided to track some statistics for the game as a way to develop my Microsoft Excel skills and explore some Data Visualization tools such as Flourish and Google Data Studio. 

​

The main dataset is currently at approximately 300 unique data points, 1 per matchweek per player. This will increase to nearly 1000 by the end of the Premier League season. For each matchweek per player I track: the ten predictions and whether they were predicted correctly or not; the banker* prediction and whether it was correct or not; the matchweek score; the Head-to-Head result; the amount of draws predicted, as well as whether they were correct or incorrect; the position in the table as well as many other columns. I also track data on a team-by-team basis such as how many times a team has been predicted overall, how many times a team was chosen as a banker* prediction and how many times a team has been predicted to draw. 

​

*Banker system is explained here