Abstract:Cricket, with its intricate strategies and deep history, increasingly captivates a global audience. The Indian Premier League (IPL), epitomizing Twenty20 cricket, showcases talent in a format that lasts just a few hours as opposed to the longer forms of the game. Renowned for its fusion of technology and fan engagement, the IPL stands as the world's most popular cricket league. This study concentrates on Dream11, India's leading fantasy cricket league for IPL, where participants craft virtual teams based on real player performances to compete internationally. Building a winning fantasy team requires navigating various complex factors including player form and match conditions. Traditionally, this has been approached through operations research and machine learning. This research introduces the FanCric framework, an advanced multi-agent system leveraging Large Language Models (LLMs) and a robust orchestration framework to enhance fantasy team selection in cricket. FanCric employs both structured and unstructured data to surpass traditional methods by incorporating sophisticated AI technologies. The analysis involved scrutinizing approximately 12.7 million unique entries from a Dream11 contest, evaluating FanCric's efficacy against the collective wisdom of crowds and a simpler Prompt Engineering approach. Ablation studies further assessed the impact of generating varying numbers of teams. The exploratory findings are promising, indicating that further investigation into FanCric's capabilities is warranted to fully realize its potential in enhancing strategic decision-making using LLMs in fantasy sports and business in general.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to use advanced multi - agent systems (multi - agent system) and large language models (LLMs) to optimize the selection of fantasy cricket teams, especially in the context of the Indian Premier League (IPL). Specifically, the research aims to improve the ability to build winning fantasy cricket teams on the Dream11 platform by introducing the FanCric framework. ### Main problems 1. **How to optimize the selection of fantasy cricket teams**: Traditional fantasy cricket team selection relies on operations research and machine - learning methods, but these methods may not be able to make full use of structured and unstructured data. The FanCric framework aims to surpass traditional methods and provide better team - selection strategies by combining LLMs and multi - agent systems. 2. **Evaluating the effectiveness of the FanCric framework**: The research evaluates the performance of the FanCric framework by comparing it with the "Wisdom of Crowds" and simple Prompt Engineering methods. The "Wisdom of Crowds" refers to the average of the choices of a large number of participants, while Prompt Engineering is to generate teams using predefined prompts. 3. **Exploring the contribution of multi - agent systems**: The research also explores the effectiveness of each agent in the FanCric framework (such as Researcher, Career Profiler, Form Assessor, etc.) in generating relevant and actionable information. 4. **The impact of the number of generated teams on performance**: Through Ablation Studies, the research analyzes the impact of generating different numbers of teams on the final performance. ### Research background - **Fantasy cricket matches**: On the Dream11 platform, participants create virtual teams based on the performance of actual players and compete on an international scale. - **Complex factors**: Building a successful fantasy team requires considering multiple complex factors, including player form, match conditions, home - away dynamics, and strategic team combinations. - **Application of advanced technologies**: The FanCric framework uses LLMs and the LangGraph orchestration framework to process structured and unstructured data to improve the accuracy of team selection. ### Research objectives - **Verifying the effectiveness of the FanCric framework**: Verify through experiments whether the FanCric framework can be significantly better than traditional methods, especially in complex decision - making environments. - **Exploring the application of AI technology in sports analysis**: Demonstrate the potential of multi - agent systems and LLMs in sports analysis and broader business decision - making. Through the answers to these questions, the research hopes to provide a new and efficient method in the field of fantasy sports and also provide a reference for the application of AI technology in other fields.

FanCric : Multi-Agentic Framework for Crafting Fantasy 11 Cricket Teams

Data Science Approach to predict the winning Fantasy Cricket Team Dream 11 Fantasy Sports

Time Series Modeling for Dream Team in Fantasy Premier League

Innovative ranking strategy for IPL team formation

Efficient Feature Representations for Cricket Data Analysis using Deep Learning based Multi-Modal Fusion Model

Deep Artificial Intelligence for Fantasy Football Language Understanding

Analysis and Prediction of Cricket Match Using Machine Learning

Large Scale Diverse Combinatorial Optimization: ESPN Fantasy Football Player Trades

Predicting Outcome of Indian Premier League (IPL) Matches Using Machine Learning

CAMP: A Context-Aware Cricket Players Performance Metric

CRICTRS: Embeddings based Statistical and Semi Supervised Cricket Team Recommendation System

Decision support system for policy-making: Quantifying skill and chance in daily fantasy sports

The Aware Cricket Ground

An Integrated Framework for Team Formation and Winner Prediction in the FIRST Robotics Competition: Model, Algorithm, and Analysis

Indian Premier League (IPL), Cricket, Online Social Media

Prediction of IPL Match Outcome Using Machine Learning Techniques

Identification of skill in an online game: The case of Fantasy Premier League

A study on Machine Learning Approaches for Player Performance and Match Results Prediction

Popularity Measuring and Prediction Mining of IPL Team Using Machine Learning

Efficient Feature Interactions with Transformers: Improving User Spending Propensity Predictions in Gaming

Using Machine Learning to Predict the Outcome of English County twenty over Cricket Matches