xG_Football

1. We build models to find the expected goal metric, one of the most important tools in football data analytics.

2. The dataset used here is acquired from Kaggle: https://www.kaggle.com/datasets/albaclosatarres/football-shots-data/data

3. Instead of doing a train-test split on the same dataset, we decide to use our own data. Since we are not able to get data in this particular fashion, we acted clever and asked ChatGPT to generate shot data for four hypothetical players Akash, Binod, Imesh and Heshan.

4. Now, we train our models on the Kaggle dataset and use them on this real-hypothetical data.

5. To make it look better, we plot the shots taken on the football pitch where one can visualize the quality of each shot taken by these 4 players.

6. Further changes will be updated soon, and a lot of room for improvement for the correct prediction of True goals.