Pogoń Szczecin Performance & Recruitment Dashboard (2025/26)

An advanced analytics and scouting dashboard designed for performance analysis and recruitment support for Pogoń Szczecin during the 2025/26 season. This project integrates match performance metrics from FotMob with market valuations, contract details, and representing agency data from Transfermarkt.

---

🚀 Key Features

1. Multi-Source Ingestion & Event Scraping (Phase 1):

   • Real-time extraction of 37 performance metrics from FotMob and Transfermarkt.
   • Event data scraper capturing spatial coordinate charts: Sofascore /average-positions via Playwright Chromium and FotMob /shotmap from next-data props.

2. ETL Pipeline & SQLite Database (Phase 2):

   • Normalization of count statistics to per-90 minutes rates.
   • Calculation of positional percentiles (GK, DF, MF, FW) across the entire Ekstraklasa league to eliminate role-based bias.
   • Storage in an optimized relational SQLite database (db/scouting.db).

3. Scouting Similarity Engine (Phase 3):

   • Cosine similarity matching using z-standardized vectors against dynamic candidate pools.
   • Scouting filters by age, market value, and position group with zero-variance protection.

4. Visualizations & Interactive Dashboard (Phase 4):

   • Interactive Plotly Radar charts displaying raw metrics alongside percentile levels.
   • Football Manager-style Plotly Scatter matrices for league-wide quadrant placement.
   • Season Shotmaps (xG-scaled circles on half-pitch layouts) and computed Passing Networks.

5. Integration Testing & Case Study (Phase 5):

   • Pytest pipeline checking data flow from SQLite querying to similarity matching.
   • Case study on Thomas Thomasberg's vertical 4-4-2 tactics.

---

📂 Project Structure

• data/raw/ – Raw JSON data files.
• db/ – Local SQLite database files (scouting.db).
• docs/ – Technical documentation, project schedule (WBS), and data acquisition analysis.
• src/ – Python source modules:
  • fetch_data.py – Multi-source scrapers and squad data aggregator.
  • etl.py – SQLite schema management and data normalization.
  • similarity.py – Cosine similarity calculation engine.
  • fetch_events.py – SQL query interfaces for shot maps and average positions.
  • viz.py – Matplotlib/Plotly tactical visualization charts.
  • scrape_and_insert_sofascore.py – Playwright scraper for average player positions.
  • scrape_fotmob_shots.py – Scraper for shot coordinates and xG.
• tests/ – Unit tests verifying ETL, matching logic, and viz outputs.

---

🛠️ Installation & Setup

Prerequisites

• Python 3.13 or newer
• Virtual environment tool (venv)

Environment Configuration

1. Clone the repository:

   git clone https://github.com/SirSail/pogon-analytics.git
   cd pogon-analytics

2. Create and activate a virtual environment:

   python -m venv .venv
   # Windows (PowerShell):
   .\.venv\Scripts\Activate.ps1

3. Install dependencies:

   pip install -r requirements.txt

Running the App

To run the Streamlit dashboard locally, execute:

streamlit run app.py

Running QA Tests

$env:PYTHONPATH="."
.\.venv\Scripts\pytest -v