In this short article, I will look at how we can use event data or x-y data to analyse a single match. Many conclusions can be drawn from a single match, but are they the right conclusions? And isn’t it more prudent to look at a series of games? All valid questions, but single match analysis can prove to be of worth and I will demonstrate this.
I’m looking at the game Celtic vs Real Betis and I will use event data to illustrate certain trends within the game, and provide a conclusion based on those visualisations. I will focus on shots taken, pass maps, pass networks and heat maps.
In the image above we see a shot map. In this image, we can see every shot taken by a particular team – Celtic in this example – and where they were taken. Celtic have had 8 shots in total, with 3 of those shots resulting in a goal and 5 didn’t go in. Now this is all quite descriptive, but we can also make the conclusion that Celtic did shoot more from the right side within the penalty area as 6 of the 8 shots were from that area. As an analyst you could pose the question: why is that? Is that where the most opportunities presented themselves or is it because the strongest shooters were in those positions? Important is to compare it with the game plan: where do you want to shoot and where do you want to finish?
In the images above you can see 3 different pass maps. In the first map you can see every pass made by Celticon the opposition’s half. It can show how well the passing is in that area, and where the most passes have been conducted. The yellow arrows mean successful passes, while the red arrow mean unsuccessful passes.
If you look at the second and third pass maps, you can see the passes made from the half-spaces. Half spaces are a sign of creativity in passing and movement. When playing passes from the half-spaces, you can have multiple options and that’s important for an analyst to look at it when translating the data. As you can see, Celtic loves to operate on the opposition’s half more from half-spaces than the right side, which could mean that that flank is utilised more and more players would invert.
It adds to the analysis when you can have actual game data and show were the errors were, just like looking at success rates of passes.
In the image above you see a pass network. This is the network of average positions and pass frequencies by the Celtic players in the game against Real Betis. This is automated to calculate the aforementioned metrics until the first sub. In this case it’s already in the 28th minute, but the network still stands. You can see where everyone is situated on average and which side/players get the most passes and where the passes go to.
It doesn’t show the formations as such, but does give details about the average shape of the team. In this example we see that Celtic does use the left flank more than the right flank, and we have also seen this in the pass maps above. The two players are rather isolated and we can cross-reference this with the information provided when we look at the video of the game.
Obviously there are many more types of event data and visualisations you can use, but in this article I wanted to illustrate how some of the event data can be used relevantly in the single match analysis.