Our project is on FIFA 2019 player statistics. Football players are the most valuable asset for any team. Therefore, it is very important to predict the value of a player and find players who have the potential to become future stars.
The purpose of this project was to solve 2 problem statements.
- To predict the value of a player based on its playing attributes, using supervised learning.
- Utilize unsupervised learning to find players who are of low value currently but have high potential to become successful professionals in the future.
Expected Insight: The expected insight from the analysis was to understand the effect of players attributes on their value and find number of groups that are present to segregate players based on their playing attribute.
Business Objective Accomplished:
- Found top 3 playing attribute to determine player value.
- Model to be used by managers to estimate player value.
- Players with high potential to be recruited for next season.
I have also attached some graphs in the pdf file.
- First graph shows that the players who participated in FIFA 2019 can be clearly segregated into 4 clusters.
- Second graph highlights some important players of footfall with respect to their "potential" and "value" attributes.
- Third graph highlights players with similar "potential" as that of popular players but with lower "value", indicating the possibility that those players will become stars in future years.