Hi, I'm David Bates

I’m a Physics graduate from Imperial College London with a passion for sports analytics, machine learning, and using data to solve real-world problems. I love applying statistical modelling to sport, where numbers can uncover trends, challenge assumptions, and improve decision-making.

Projects

• 🏈 A custom Elo rating system that tracks and predicts NFL team performance, adjusting for factors like home-field advantage and margin of victory.
• 🎯 DeepCue, a machine-learning model that forecasts snooker outcomes by blending historical performance, matchup dynamics, and engineered features.

Climate & Data Analysis

My Master’s thesis focused on analysing satellite climate and emissions datasets, applying statistical inference to identify environmental trends and changes over time. This project combined data wrangling, statistical modelling, and clear visual storytelling to make complex patterns accessible to both scientific and general audiences.

Presenting my Master's project poster on climate and emissions data analysis.

Beyond Sports

My work also spans data automation and large-scale dataset exploration. I’ve built automated reporting pipelines and analytical workflows using Python, SQL, and Excel VBA, transforming raw information into clear and actionable insights.

Currently

I’m looking for opportunities at the intersection of data science, sport, finance, and technology. I’m keen to collaborate—whether that’s co-developing models, exploring new datasets, or simply exchanging ideas. If this resonates, let’s talk.

I am also working on an NFL Elo Rating and advanced match prediction model which I am posting about regularly on this website ([NFL Articles)](https://davidbates.me/nfl/archives/). After the start of the NFL season, I also plan to return to [DeepCue](https://davidbates.me/snooker/archives/), a snooker match prediction model I developped in 48 hours.

Recent Analysis

Here are my latest deep dives into the projects I’m building. They’re written to be approachable—step by step through the modelling choices, the trade-offs, and the takeaways.

Modelling NFL Quarterbacks: Choosing the Right Stats

This is the first post in my “Modelling NFL Quarterbacks” series, where I break down the process from picking stats to building ratings.

August 09, 2025 | 4 min read Read More

Launch of DeepCue

48-hour sprint to build a snooker match prediction engine using scraped data, feature engineering, and ML models.

May 02, 2025 | 1 min read Read More

Looking for Work

I’m currently open to data science and consulting roles, especially those involving:

• Sports analytics or predictive modelling
• Machine learning and data visualisation
• Complex problem solving with real-world impact

Feel free to get in touch via my About page or connect with me on LinkedIn.