Over the course of the last week, I have noticed some familiar discourse on social media around the use of data visualisations and football. Before I go into the nature of this narrative I should probably point out that I dont support one point of view over another but I do understand both.
There were some people posting that they were disillusioned with the sheer number of people on Twitter who were sharing scatter graphs with little or no context or insight. These are relatively easy to create and to throw up online and they can, at times, get good traction and interaction.
The problem comes when these scatter graphs are put up without any context or any attempt to gain insights from what the data is telling you. Data alone is rarely enough for you to gain insight into a player, a club or a league. You have to be able to either find the right combination of metrics to show what you are trying to prove or use a combination of formatting and text to try to add to the story.
At the same time, I understand why people would try to put these data visualisations up in order to start to get noticed and even to practice. I have done the same things when starting to use data in connection to football and there is nothing wrong with this, just try to think about what you are trying to say or get across and who you are trying to get the message across to!
Interestingly, these conversations were taking place on social media around the same time that I was having more focused work-related conversations, which I cant go into for obvious reasons, about how we can use data more effectively to give us the insights into data and to start to drill down into the kind of players who should be of interest to us as part of the recruitment process.
Data is an important part of the recruitment process but we have to remember that it is just part. There is little value as a recruitment analyst in inundating people with numbers, data or even visualisations if they are not relevant to the players or markets that you are looking to recruit from.
Recently I have started to see the value in finding ways to visualise a players output in direct relation to their team. Lets look at an example.
Heres a scatter graph that doesnt actually give us a lot of insight. The data set is all forwards in the top-5 European league that are 26 or under and that have played at least 750 minutes this season. The graph shows goals per 90 on the. vertical axis and xG per 90 on the horizontal axis and it largely tells us what we would expect it to. The likes of Lautaro Martinez, Diogo Jota, Dusan Vlahovic and Kylian Mbappe are very good forwards. Great, but what else? Not much, right?
Yes, we can drill down further towards the middle of the graph to look to see whether there are any interesting names there but in truth in visualisations like this will draw peoples eyes to the top-right area of the graph to see who the standout players are. But that doesnt mean that the metrics that have been used in this visualisation do not have merit when we are looking to assess forward players.
It was at this point that I real
