For this one, I decided to track the Glicko-2 rating for every team in both the Tier 1 and Tier 2 leagues. I expected a sloppy mess of spaghetti with no real rhyme or reason to it. Instead, three teams stood out: Arsenal, Chelsea, and Manchester City. They were clearly better than all the other teams and had been for many years. Then, I noticed that Manchester United climbed through Tier 2 in a single year and quickly matched the three established teams in Tier 1. That gave me something to work with.
Potential issues, shortcuts taken, Future Workâ˘
After spending some time trying to explain Glicko-2, I’ve decided to leave that explanation to the expert and the Wikipedians. The short version is that maybe the rating deviation should increase when a new season starts or when the roster changes or when a team moves to a different tier.
Data
The data for this plot comes from The English Women’s Football (EWF) Database. I used the version that’s in the Plotly Figure Friday repo. Plotly’s version goes through the end of the 2023-2024 season.
While writing up this blog post, I took a look at some of the other visualizations linked in the EWF repo. It looks like they largely agree that Arsenal, Chelsea, and Manchester City dominate the league. I don’t know much about English football, so that corroboration gives me some confidence that this Glicko-2 approach reflects reality.
Code
The code for this plot lives in my fork of the Plotly Figure Friday repo.