Soccer Graph Analysis Pipeline

This README provides a simple step-by-step guide for running the soccer graph analysis scripts in this repository.

1. Data to Graph (if necessary)

Convert raw match data CSV to a graph JSON file.

python generate_soccer_graph.py <raw_data_csv>

• The output file is automatically named as SoccerGraphData/<raw_data_csv_base>_graph.json.

2. Analyze Subgraphs

Analyze subgraphs of the generated graph.

python analyze_graph_subsets.py <graph_json_file> <max_k>

3. Compute Node Metrics

Compute node metrics for each team/node.

python compute_node_metrics.py <graph_json_file> <subset_analysis_csv> [upset_rate_csv]

• OPTIONAL PREREQ: Calculate upset scores (if you want to include upset rates in node metrics):

python soccer_upset_rate.py <raw_data_csv> <graph_json_file>

4. Plot Rating vs Intransitivity Centrality

Plot HHD rating vs IC score for each node.

python rating_icscore.py <node_metrics_csv>

5. OPTIONAL: Plot Upset vs Intransitivity Centrality

Plot upset rate vs IC score for each node.

python upset_icscore.py <node_metrics_csv>

6. OPTIONAL: Plots Ranking vs IC and Ranking vs Upset

• PREREQ: Store real ranking in a CSV file (e.g., real_ranking_EPL.csv)

python real_ranking_plots.py <node_metrics_csv> <real_ranking_csv> <real_rank_col>

---

• All scripts require command-line arguments as shown above.
• Output files are automatically named based on input files.
• For more details, run each script without arguments to see usage instructions.