Live Demo

---

What is CricScope?

CricScope is a luxury-grade IPL match intelligence dashboard that computes real-time win probabilities using machine learning — trained on historical ball-by-ball delivery data spanning 2008–2020.

Built with a fintech-inspired dark UI featuring glassmorphism cards, gold gradients, and a premium serif + mono type system. Every design decision was intentional: this is not a student project — it's a production-grade sports analytics product.

GSSoC '26 & NSoC 2026 — Project Admin: Arnav Singh

---

📚 Documentation

• Contributing Guidelines
• Security Policy
• Code of Conduct
• License

---

Architecture

┌─────────────────────────────────────────────────────────────────────┐
│                         CricScope Pipeline                         │
├─────────────────────────────────────────────────────────────────────┤
│                                                                     │
│   matches.csv ────┐                                                 │
│                   ├──► Merge on match_id ──► Filter: Inning 2       │
│   deliveries.csv ─┘                │                                │
│                                    │                                │
│                      ┌─────────────▼───────────────┐                │
│                      │      Feature Engineering      │               │
│                      │                               │               │
│                      │  current_score  (cumsum)      │               │
│                      │  runs_left      target-score  │               │
│                      │  balls_left     120-ball no.  │               │
│                      │  wickets        10-dismissed  │               │
│                      │  CRR            score/over    │               │
│                      │  RRR            runs*6/balls  │               │
│                      └─────────────┬───────────────┘                │
│                                    │                                │
│                      ┌─────────────▼───────────────┐                │
│                      │       Sklearn Pipeline        │               │
│                      │                               │               │
│                      │  ColumnTransformer            │               │
│                      │    OneHotEncoder              │               │
│                      │      batting_team             │               │
│                      │      bowling_team             │               │
│                      │      city                     │               │
│                      │    passthrough                │               │
│                      │      numeric features         │               │
│                      │                               │               │
│                      │  LogisticRegression           │               │
│                      │    max_iter = 1000            │               │
│                      └─────────────┬───────────────┘                │
│                                    │                                │
│                      ┌─────────────▼───────────────┐                │
│                      │ predict_proba() → [0, 1]    │                │
│                      │ Confidence: High/Mod/Close  │                │
│                      └─────────────────────────────┘                │
└─────────────────────────────────────────────────────────────────────┘

Model Evaluation Metrics

The prediction engine uses a Logistic Regression classifier trained on IPL ball-by-ball match data spanning IPL seasons from 2008–2020.

Dataset Split

Split Type	Ratio
Training Data	80%
Testing Data	20%

---

Model Highlights

• Real-time win probability prediction
• Ball-by-ball dynamic recalculation
• Match-state feature engineering
• OneHotEncoded categorical preprocessing
• Optimized scikit-learn pipeline
• Context-aware chase prediction logic

---

Detailed evaluation metrics and expanded cross-validation results will be added in future model benchmarking updates.

---

Tech Stack

Layer	Technology	Purpose
Frontend	Streamlit + Custom CSS	UI rendering & layout
Styling	Glassmorphism + Premium CSS	Luxury UI/UX
ML Pipeline	scikit-learn	Preprocessing + Logistic Regression
Data	Pandas + NumPy	Feature engineering
Deployment	Streamlit Cloud	Live hosting

---

Project Structure

cricscope/
│
├── assets/
│   ├── dashboard.png
│   ├── prediction-page.png
│   ├── analytics.png
│   
├── app.py
├── matches.csv
├── deliveries.csv
├── requirements.txt
└── README.md
└── demo_.gif

---

Getting Started

Prerequisites

• Python 3.9+
• IPL Dataset

Dataset: https://www.kaggle.com/datasets/patrickb1912/ipl-complete-dataset-20082020

---

1 — Clone Repository

git clone https://github.com/Arnav-Singh-5080/cricscope.git
cd cricscope

---

2 — Create Virtual Environment

python -m venv venv

# Linux / macOS
source venv/bin/activate

# Windows
venv\Scripts\activate

---

3 — Install Dependencies

pip install -r requirements.txt

---

4 — Add Dataset

cricscope/
├── app.py
├── matches.csv
└── deliveries.csv

---

5 — Run Application

streamlit run app.py

Visit:

http://localhost:8501

---

Open Source Programs 2026

CricScope is an officially selected project under:

• GirlScript Summer of Code 2026 (GSSoC '26)
• Nexus Summer of Code 2026 (NSoC 2026)

Contributors are evaluated on:

• Code quality
• UI consistency
• ML innovation
• Documentation quality
• Feature implementation

This repository welcomes:

• Beginners
• Open-source contributors
• UI/UX developers
• ML engineers
• Competitive programmers

---

Why Contribute?

• Beginner-friendly contribution environment
• Real-world ML engineering exposure
• Production-grade UI/UX experience
• Active mentoring & PR reviews
• Portfolio-worthy open-source project
• Industry-style collaboration workflow

---

Contribution Guide

How to Contribute

# 1. Fork the repository

# 2. Clone your fork
git clone https://github.com/<your-username>/cricscope.git

# 3. Move into project
cd cricscope

# 4. Create feature branch
git checkout -b feature/your-feature-name

# 5. Make changes

# 6. Commit
git add .
git commit -m "feat: describe your changes"

# 7. Push
git push origin feature/your-feature-name

# 8. Open Pull Request

---

Contribution Guidelines

• Maintain the premium dark luxury aesthetic
• Follow the existing project structure
• Keep pull requests focused
• Test locally before opening PR
• Write meaningful commit messages
• Respect UI consistency

---

🔐 Security Policy

Please review the Security Policy for vulnerability reporting guidelines and supported versions.

---

🤝 Code of Conduct

Please read our Code of Conduct before contributing to the project.

Good First Issues

Area	Task	Difficulty
UI	Animated win probability graph	Medium
UI	Mobile responsiveness	Medium
UI	Team stat pills	Easy
ML	IPL 2021–2024 integration	Easy
ML	SHAP interaction visualizations	Medium
ML	Cross-validation metrics	Medium
Feature	Match report PDF export	Hard
Feature	Head-to-head analytics	Medium
Docs	Add screenshots & GIF demo	Easy

---

Project Admin

Arnav Singh

Project Admin — GSSoC '26 & NSoC 2026

 

 

---

Support the Project

If this project helped you, consider giving it a ⭐ on GitHub.

It helps more contributors discover the project and motivates future development.