Martin Eastwood

A Python code snippet demonstrating the new 'create_dixon_coles_grid' function in the penaltyblog library. The code shows how to initialize a probability grid using expected goals (lambdas) for the home and away teams, and then calculate home win probabilities, over/under 2.5 goal totals, and Asian Handicaps directly from that grid.

🚀 penaltyblog v1.9.0 is live! 🐍⚽️

New in this version:
✅ create_dixon_coles_grid(): Use lambdas from external ML models.
✅ goal_expectancy_extended(): Infer rho/lambdas from odds.
🛠️ Improved quarter lines (2.25/2.75) logic for totals market.
⚡️ More optimisations to FootballProbabilityGrid

GitHub - martineastwood/penaltyblog: ⚽ High-performance football analytics toolkit: build data pipelines, scrape data, model matches, rank teams, and bet smarter | Powered by pena.lt/y/blog 🚀 ⚽ High-performance football analytics toolkit: build data pipelines, scrape data, model matches, rank teams, and bet smarter | Powered by pena.lt/y/blog 🚀 - martineastwood/penaltyblog

Everything you need to get started is right here:

pip install penaltyblog --upgrade

💻 Repo: github.com/martineastwo...
📖 Docs: penaltyblog.readthedocs.io
📦 PyPI: pypi.org/project/pena...

A technical MCMC Trace Plot with two horizontal subplots labeled "attack_Arsenal" and "home_advantage." Each plot shows four overlapping colored lines (Chains 0 to 3) creating a dense, horizontal "fuzzy caterpillar" shape. This visualization indicates that the Bayesian model has successfully converged and is stable across 2,000 iterations.

Why build a custom MCMC sampler from scratch?

In my latest blog post, I dig into the "why" and the "how". From the decision to drop heavy dependencies like Stan for this use case, to using Cython for speed, and how to interpret those "fuzzy caterpillar" plots.

Read it here: pena.lt/y/2026/01/08...

What can you do with v1.8.0?

✅ Bayesian Dixon-Coles: Quantify uncertainty, not just point estimates
✅ Hierarchical Models: Learn league-wide variance to handle small sample sizes
✅ Built-in trace plots: Ensure model convergence
✅ Familiar API: Works seamlessly with existing scrapers and weights

A dark-themed code editor window showing Python code using the penaltyblog library. It demonstrates initializing a HierarchicalBayesianGoalModel with home/away goals and teams, fitting the model using MCMC sampling with parameters like 2000 samples and 4 chains, and finally predicting match probabilities for an Arsenal vs. Manchester City fixture

Bayesian goal models are back in penaltyblog v1.8.0 - and this time, they’re dependency-free! ⚽️📈

⚽ penaltyblog v1.7.1 is now available:

Added params_array and param_indices functions to goal models to make it easier to work with the model's parameters.

Thank you to Sebastian Velandia for this contribution!

pip install --upgrade penaltyblog

📝 A couple of important notes:

You'll need your own Opta subscription/credentials to use this (I provide the tools, not the data!) 🔑

As this is a v1 release of the integration, there's likely some rough edges. If you run into any, let me know - I'm happy to work with you to improve it. 🤝

Code example showing how the penaltyblog python package can now integrate directly with the StatsPerform Opta API to download data directly.

⚽ penaltyblog v1.7.0 is now available: Direct integration with the Opta (Stats Perform) API.

You can now stream matches & events lazily without downloading the JSON first. Includes helpers for human-readable filtering (no more memorising IDs).

pip install --upgrade penaltyblog

Have been doing Advent of Code in Nim and Kotlin this year and enjoying learning both. I can see Nim becoming one of my favourite languages!

Thanks Nils, it’s a good point and is on my todo list to dig further into how long it takes for something like FSAA to stabilise to something useful early on in a player’s career. Even with a fairly wide HDI at that stage, there’s still potentially benefits to its use

Not yet, but good idea!

New article: "Shrinkage, Uncertainty, and Son Heung-min: Using Bayesian Methods to Identify Finishing Ability"

which discusses using a Bayesian hierarchical approach to quantifying player finishing ability, with credible intervals to express uncertainty.

pena.lt/y/2025/10/01...

🎉 penaltyblog v1.6.1 is out!

✨ What's new:

- Python 3.14 support
- scipy 1.16+ compatibility
- Better numerical stability for Negative Binomial model
- New Colab notebook for implied probabilities example

pip install --upgrade penaltyblog

Absolutely. Also, people started sharing big (for then anyway) pre-trained networks that made it much easier to get started. I built many models in my day job back then by fine tuning BERT and ImageNet that I would have struggled to train from scratch without investing massively in compute.

We can also split Massey Ratings into attack & defence:

🔴 LFC: best attack in the league paired with a mid-table defence
⚪️ ARS: Elite at both ends. They have the #1 defence and the #3 attack
🌳 Forest: A disaster at both ends of the pitch

Massey Ratings for the English Premier League

Here's how the Premier League table really looks according to Massey Ratings, which account for strength of schedule.

📈 Arsenal (+1.5) & Man City (+1.3) are clear strongest overall
😬 Man Utd (-0.2) rank in the bottom half
📉 West Ham & Forest (-1.2) are worst teams by far

Thanks to everyone who suggested features and reported issues. Your input shapes the package's development.

Questions or feedback welcome at pena.lt/y/contact

Install: pip install penaltyblog
GitHub: github.com/martineastwo...

📚 Interactive Colab notebooks are available in the docs - experiment with real examples without any local setup.

I'll be steadily expanding these over the coming weeks to cover all functionality in the package.

Docs: penaltyblog.readthedocs.io

🔧 Improved implied odds module:

- New logarithmic overround removal method for better accuracy
- Structured results instead of raw arrays
- Better handling of edge cases

Making it easier to work with bookmaker probabilities in your analyses.

💰 Expanded betting utilities:

- Kelly Criterion for multiple outcomes
- Arbitrage opportunity detection
- Value bet identification
- Hedge bet calculations
- Odds format conversion (decimal/fractional/American)

All functions now return structured outputs for easier integration.

⚽ penaltyblog v1.6.0 is now available!

📊 MatchFlow updates:

- SQL-style joins for nested football data (left, right, outer, inner, anti)
- Cloud storage support: read/write directly to AWS S3, Google Cloud Storage, and Azure Blob
- Automatic type inference for join keys

pip install penaltyblog

🚀 New article on my blog walking through the latest updates to the penaltyblog python package for football (soccer) analytics & betting

✅ New interactive pitch plots
✅ 5-10× faster goal models
✅ New Flow query DSL

👉 pena.lt/y/2025/08/14...

📚 Docs: penaltyblog.readthedocs.io/en/latest/in...
💻 GitHub: github.com/martineastwo...
🐍 pip install penaltyblog

Feedback welcome, let me know what you build!

Image of code showing examples of penaltyblog Flow's new DSL

🔍 New: Flow Query DSL

Filter datasets with safe, Pythonic expressions:
- AST-parsed (no eval)
- Variables, regex, dates
- Access nested fields

Image of code demonstrating the the features in penaltyblog's goals models

📈 Goal models are now 5-10× faster

- Cython-powered analytical gradients for speed + stability
- Fine-tune with minimizer_options:

GIF demonstrating the interactive tooltips in penaltyblog's new Pitch plotting API

⚽ New: Pitch Plotting API

Build interactive football visualisations with:
- Multiple layouts & themes
- Scatter, heatmaps, arrows, comets
- Custom hover tooltips

🚀 penaltyblog v1.5.0 is here!

✅ Interactive pitch plots
✅ 5-10× faster goal models
✅ New Flow query DSL

What’s new 👇

📊 New article on my blog: How Accurate Are Soccer Odds? A Data Dive into 250 Million Betting Lines

🔍 How sharp are different bookmakers?
📈 How accurate are bookmaker's odds?
🎯 Are the odds well-calibrated?

➡️ pena.lt/y/2025/07/16...

There's also a new blog post here that explains more about the latest updates:

pena.lt/y/2025/06/10...

GitHub - martineastwood/penaltyblog: ⚽ High-performance football analytics toolkit: build data pipelines, scrape data, model matches, rank teams, and bet smarter | Powered by pena.lt/y/blog 🚀 ⚽ High-performance football analytics toolkit: build data pipelines, scrape data, model matches, rank teams, and bet smarter | Powered by pena.lt/y/blog 🚀 - martineastwood/penaltyblog

🚀 penaltyblog v1.4.0 is out!

Now includes a query plan optimiser for smarter Flow pipelines:

• Optional FlowOptimizer for smart rewrites (optimize=True)
• New .plot_plan() for pipeline viz
• .with_schema() for field validation
• Rolling- and time-based summaries

github.com/martineastwo...