UFC Betting Data / Machine Learning Project

I’m looking for a Python/data science freelancer to help build a data pipeline and prediction model for UFC betting. The initial focus will be gathering, cleaning, and structuring historical UFC fight data and historical betting odds. If the engagement goes well, there is an opportunity to continue into feature engineering, machine learning, backtesting, model calibration, and bet-sizing. The ultimate goal is a system that estimates a fighter’s probability of winning and compares that probability against sportsbook odds to identify potentially mispriced bets. Initial Scope * Gather historical UFC fight and fighter data from reliable sources * Gather historical betting odds where available * Clean, normalize, and merge the datasets * Build point-in-time features using only information that would have been available before each fight * Prevent look-ahead/data leakage * Create a repeatable process for adding future UFC events and updating the dataset * Document data sources, assumptions, and methodology Potential later work includes below + reccomendations from the freelancer: * Logistic regression / Elo baseline models * XGBoost, LightGBM, or similar ML models * Probability calibration * Walk-forward historical backtesting * Comparison against no-vig sportsbook implied probabilities * Betting thresholds and bankroll sizing * Model diagnostics and performance tracking Important Requirement: Transferable System I am not looking for a black-box model that only the freelancer can operate. I have some prior programming experience but am rusty with Python, so clear code and practical documentation and handoff are important. All code, datasets, scripts, notebooks, documentation, and model files created during the project must be transferred to me. The system should be organized and documented so that I can understand the workflow, run the model myself, update it with new fight data, and continue improving it after the engagement. Ideal Candidate Experience with Python, pandas, data scraping/API integration, machine learning, time-series or sports-model backtesting, and probability calibration is preferred. Experience with sports betting or UFC data is a plus but not required. When applying, please briefly describe relevant projects you have completed, your proposed approach to the initial data-gathering phase, and an estimated cost or number of hours for that phase.