FIFA Player Value Analysis

Exploratory data analysis and preprocessing pipeline for the FIFA Player Performance and Market Value dataset, building toward ML-based market value classification.

Problem Statement

This project explores, cleans, and prepares the FIFA Player Performance and Market Value dataset for future machine learning tasks — specifically predicting and categorizing player market value into three tiers (Low / Medium / High) based on a player's performance metrics, club affiliation, and biographical attributes.

Dataset

Source: Kaggle — FIFA Player Performance and Market Value Analytics

The dataset is not included in this repository (excluded via .gitignore). Download it from Kaggle and place it in the project root before running the notebook.

Approach

The notebook implements a full EDA and preprocessing pipeline:

Step	Details
Data Exploration	Summary statistics, missing value audit, distribution checks
Encoding	Binary encoding, ordinal encoding, one-hot encoding for categorical features
Normalization	StandardScaler applied to continuous features
Binning	Equal-width binning to discretize the market value target into Low / Medium / High
Feature Engineering	Derived features such as goals_per_90
Feature Selection	Pearson correlation analysis for numeric features; chi-square testing for categorical features
Sampling	Stratified sampling to address class imbalance
Visualizations	Histograms, boxplots, scatterplots, and a full correlation heatmap

Key Insights

• Position has no significant effect on market value — one-way ANOVA found no statistically significant difference in market value across player positions (p > 0.05).
• Interaction effect detected — injury_prone and transfer_risk_level together have a meaningful interaction effect on market value, beyond either variable alone.
• Counter-intuitive finding — 255 players rated 85+ overall fall into the Low Value category, suggesting overall rating alone is not a reliable proxy for market value.
• Scouting recommendation — Target players aged 22–26 from lower-value clubs who combine a high overall_rating with a low transfer_risk score for maximum value acquisition.

Running Locally

1. Clone the repository

   git clone https://github.com/aliahmedd4/fifa-player-value-analysis.git
   cd fifa-player-value-analysis

2. Install dependencies

   pip install -r requirements.txt

3. Download the dataset

   • Go to the Kaggle dataset page
   • Download fifa_player_performance_market_value.csv
   • Place it in the project root (same folder as the notebook)

4. Launch the notebook

   jupyter notebook fifa_value_analysis.ipynb

Project Structure

fifa-player-value-analysis/
├── fifa_value_analysis.ipynb  # Main analysis notebook
├── requirements.txt      # Python dependencies
├── README.md
└── .gitignore

Contributors

• Ali Sherif
• Ahmed Rashad
• Asser Ehab
• Ali Ahmed