Q/A

What do you mean by `optimal`?

All instances are solved to optimality using mathmetical modeling. Data is provided by FPL Review, problems are modeled using sasoptpy package and finally solved using open-source CBC solver. You can see the modeling source code here. Feel free to clone the repository and play with the model.

I found a better squad, how about it?

You can find another squad with exactly the same objective value, but cannot get anything strictly better. Optimization models are solved to exact optimality.

How often / when do you update this page?

This page is automatically built everyday using GitHub Actions, and updated 4 times (4:00, 10:00, 16:00, 22:00 GMT) and whenever there is a code change.

I have a suggestion or improvement to your model. How can I contribute?

Feel free to submit a pull request on GitHub, or open an issue.

Are you using these solutions for your team?

You bet I do! I solve an optimization model every week and use the forecast for next 8 gameweeks, but it is a bigger model. I use SAS Optimization for every decision, from transfers to chip usage for my team Optimized FC (ID: 2221044). I tweet regularly on optimization, sports analytics, and FPL on Twitter.

I would like to use optimization for FPL, too! Where can I start?

Depends on where you feel more comfortable working at: Solver add-in is available on Microsoft Excel and several open-source and commercial solvers have interfaces to programming languages. I maintain the Python interface for SAS Optimization that I have mentioned (sasoptpy) and would be happy to answer your questions on Twitter.