Fantasy Cricket XI Predictor

A machine learning-based web app that predicts fantasy cricket scores for players and generates an optimal playing XI using XGBoost.

---

Overview

This project takes match-level player statistics as input and predicts fantasy points. Based on predictions, it selects the best 11 players and assigns Captain and Vice-Captain roles.

---

Features

• Upload match-wise player dataset (CSV)
• Predict fantasy scores using trained XGBoost model
• Rank players automatically
• Generate Best XI (Top 11 players)
• Assign Captain and Vice-Captain
• Streamlit-based interactive UI
• Normalized ranking system (0–100 scale)

---

Project Structure

project/
│
├── app.py
├── data/
│   ├── fantasy_rank_model.pkl
│   ├── features.json
│
├── sample_match_big.csv
├── requirements.txt
├── README.md

---

Installation

1. Clone repository

git clone https://github.com/parinaB/fantasyPlayerPrediction.git
cd PlayerPredicts

2. Install dependencies

pip install -r requirements.txt

3. Run the app

streamlit run app.py

---

Input Format (CSV)

Your input CSV must contain:

player
match_id
avg_runs_last5
avg_wickets_last5
opp_bowling_strength
boundary_score
role_Batsman
role_Bowler
venue_avg_runs
opp_strength
last_runs
last_wickets

---

Model Details

• Algorithm: XGBoost Regressor
• Output: Predicted fantasy score
• Ranking: Min-max normalized (0–100 scale)

---

Output

• Top 11 Fantasy XI
• Captain and Vice-Captain
• Player ranking table
• Average and maximum predicted score

---

Notes

- Feature names must match features.json exactly
- Model expects engineered features only
- Ensure CSV format is correct before upload

---

Built for machine learning practice and fantasy sports analytics