Analytics in Football - Official Thread

[Philby]

A good read.

Introducing 1v1s

The NBA has tipped off, meaning I’m watching a lot more basketball recently.

talatabont.substack.com

 

[Cloudy O'Ranieri]

A good read.

Introducing 1v1s

The NBA has tipped off, meaning I’m watching a lot more basketball recently.

talatabont.substack.com

Ok, this thing's broken. It tells us that United and Rashford are good at something!

 

[Cloudy O'Ranieri]

Anybody have access to the following, and willing to pass it on, please?

xG, xA and a battle for hearts and minds – a friendly debate between a fan and The Athletic’s data expert

Duncan Alexander and a football fan discuss data, where it is used well, where it could be used better and why it's sometimes confusing

www.nytimes.com

 

[Philby]

Anybody have access to the following, and willing to pass it on, please?

xG, xA and a battle for hearts and minds – a friendly debate between a fan and The Athletic’s data expert

Duncan Alexander and a football fan discuss data, where it is used well, where it could be used better and why it's sometimes confusing

www.nytimes.com

Does this work for you?

https://archive.ph/hLyOm

 

[Cloudy O'Ranieri]

Does this work for you?

https://archive.ph/hLyOm

Seems to be working alright, thanks.

 

[Philby]

Will give this a listen.

He was head of data analytics @ Man Utd for a couple of years until INEOS arrived.

Manchester United make historic appointment to improve trophy chances

An experienced data scientist has been brought into the club to help on-field performances

www.manchestereveningnews.co.uk

 

[Philby]

A good read. Interesting to City have the top 2 players in this analysis.

Haven't checked, but perhaps they don't necessarily overlap on the pitch anywhere near as much as guys who typically pop at the top of these lists. So are job-sharing to some degree.

Digging Deeper Into Ball Carrying

In this article, we explore how Hudl Statsbomb data can be leveraged to identify valuable traits in ball carrying players.

www.hudl.com

 

[Philby]

Encouraging, but not surprising, to see Wirtz appear near the top of these playmaker charts

Contextual Scouting: The Modern Playmaker

Trym Sorum uses SkillCorner Game Intelligence to profile the Modern Playmaker - attacking midfielders that unlock opposing defences in the Finish phase.

skillcorner.com

 

[Cloudy O'Ranieri]

Interesting development: Fotmob are doing xG for League of Ireland Premier Division. Only started last week so won't be too comprehensive for this season, as a whole, but hopefully will continue for a long time to come.

In a totally related matter, how would you be at automating the collection of per shot xG data from that site? @Philby

I've tried to get AI to help me (reluctantly for environmental reasons), but I don't think my prompts have been precise enough to convey just how tick I am with this scripting stuff. It seems to think I can work Python, when I can't.

 

[Philby]

Scraping from sites like FotMob would likely be fairly achievable using Python alright.

Obviously scraping shot-by-shot, game-by-game data would be a bigger undertaking than scraping averages (e.g. player X averages 0.25 xG per 90 across season 2024/25), because you'd need to build a list of games and scrape shot data game-by-game.

What's the use case for granular shot-by-shot data?

What have you tried so far scraping-wise?

(FYI I don't think it'll be achieveable to craft a prompt for an AI tool, e.g. ChatGPT, which will actually scrape the data directly and return it within the ChatGPT window. But if the thought is to create a script in Python (or other technology) then there are ways to use ChatGPT to help you interatively develop versions of that script. But you'd need somewher eto run that script - e.g. on your local machine or even Google Colab)