As regular readers may know, the current set of tasks being carried out by Matz1lla involves Expected Goal data.
We have made great progress with this in recent weeks and, on our test servers, we now have tables and maps up and running. That gives me the opportunity to show work in progress and detail what is due to arrive in Update Six.
Effectively, we’re using Opta’s data to bring you both Expected Goals (xG) and Expected Assists (xA) for both players and teams.
If you are new to the concept you can view this excellent movie from Opta explaining how xG is generated.
A full explanation from Opta is also available here.
Meanwhile, the xA data is a little different to that currently available elsewhere.
The Expected Assists data you will get as part of the update will not be generated using the xG value of a shot, instead – as explained here – it assesses the likelihood that a pass will lead to a “primary assist”.
Opta’s full explanation is:
“Expected assists measures the likelihood that a pass will be a primary assist. The model is based on the finishing location of the pass, what type of pass it was and a variety of other factors. This model is not reliant on whether a shot was taken from this pass, so credits all passes, regardless of whether they result in a shot.”
By combining xG and xA, we are able to offer an Expected Goal Involvement (xGI) statistic for players, representing the number of goals they were expected to either score or assist over any given period.
The table below shows the top ten Defenders xGI for Gameweek 7 just gone.
So our Members will be getting access to three statistics – xG, xA and xGI, with the latter available for players only.
However, our team data also includes xGC, or Expected Goals Conceded, so that we can assess team defensive strength.
All this data is, as ever, available to slice by any of our filtered intervals – so you can view player and team data by Gameweek, by Matches and by Home and Away matches.
It’s also dropped into our comparison tools for players and teams and will be made available for adding to custom tables.
We’ve also mapped the data to provide a visual representation of xG.
So below you’ll see a graphic showing a comparison of Sergio Aguero and Gabriel Jesus and their xG so far. We’ve allowed for full tooltip information on any xG value, giving access to the shot type, the assister and the xG value assigned.
Effectively, coupled with our existing Opta data, this adds a whole new weapon to our armoury when analysing teams and players.
It allows for the assessment of players and teams that are perhaps either under or overperforming, therefore offering clues as to likely regression or improvement.
We will, of course, be using the data to create members’ articles, offering guidance on how to use it and what it can tell us.
We’re currently refining the work on the test server, fixing bugs and adding new methods of showing the data.
Next, we plan to add widgets to the members’ dashboard, showing the top ten under and overachieving players for xG for the current Gameweek, with the ability to quickly view any Gameweek and the season as a whole.
I would hope that we will be in a position to release the first iteration of xG as part of Update Six early next week.
