Before we start using data to scout the opposition, it’s important we understand why almost all professional teams around the world are using data.
It’s not just professional teams, either. Youth academies, colleges and semi-pro teams are also looking to gain an extra competitive edge wherever possible, so naturally, data analysis is an area they’re exploring and using.
All humans have subconscious bias, and while at one end of the spectrum some ‘traditionalists’ may doubt the value and impact of data, at the other end it’s possible some people over-value it’s impact. General consensus within the highest levels of the industry is for data to be well integrated into most processes within football clubs, while not losing sight of subjectivity and the art of coaching and working with human beings.
The quality and understanding of data is improving at a fast pace, and with visual tracking, AI and machine learning models, we’re seeing more and more brilliant use-cases for the implementation of insights provided by data. Arguably however, the largest impact is still found by knowing what questions to ask, how to use insights effectively, and crucially, knowing what information data might be missing.
Here’s what our professional contributors had to say about the benefits of using data:
Earlier we touched on the potential pitfalls of using data. What we find most often at the APFA is people thinking they understand how to use data, and drawing conclusions from it when most, more experienced analysts would correct these insights within seconds of hearing them.
Here’s an example from a popular book on scouting ‘The Nowhere Men’ – which is an excellent read, and highly recommended by the APFA!
“The data is sold on it’s objectivity, it’s fact. But how is it interpreted? An example – you play the ball back to your full back and he’s been told we’re going to clip it into the channel, whenever possible. So he does that. And the opposition full back comes across and heads it out for a throw-in. Now in the data world, that’s an incomplete pass because the opposition won it. Now in the real world, the real football world, we’ve just won a throw-in in the opposition’s final third. So the full back has done exactly what we wanted him to do. Well done, son. That’s where the people using this information have got to be intelligent enough to understand what they’re looking at and to understand it’s limitations. The role in many ways is about information. How you get information, how you manage it, and how you use it.”
James Smith – Head of Technical Scouting at Everton
Any experienced analyst could easily explain the problems with the assumptions, and how modern day player recruitment with data first of all is unlikely to look at Passing % in the first place when evaluating a player, but second, how advanced modelling can account for a defender successfully moving his team up the pitch and winning a throw.
The issue here is when a scout think they know how to use stats and makes assumptions based on an incomplete knowledge of the subject. James Smith makes an excellent point about the importance of the people using data understanding how it can get you into trouble if used incorrectly.
Here’s our contributors discussing some of the potential traps people can fall into when using data incorrectly:
The first task for this module is to briefly outline the risks and rewards of using data at your club if you currently work for one, or in the role you’d ideally like to be working in. If you’ve ever experienced a situation which was helped or hindered by data, please describe this too!
Please add these thoughts to the comments section at the bottom of this module and feel free to read some responses from others who’ve taken this course.
Below is an example from Oliver Gage, who was an opposition analyst for Sheffield Wednesday and Houston Dynamo earlier in his career.
“I once presented a report to a coach on the opposition we were about to face. He was pretty old school and to be honest I was really struggling to get him to buy in to the concept of using data. The report outlined that the opposition striker had a higher % of headed shots than any other forward in the league.
I was confident this meant we would be facing a lot of crosses in the upcoming match. I was hoping this would impact training and perhaps some time would be spent defending crosses in the build up to the game.
Unfortunately the coach had already watched their two previous games, and the the striker was suspended for them. They didn’t cross much at all, as the replacement forward was small and quick – the total opposite. He decided we should work on our high defensive line instead.
Ultimately the big striker started the game. We drew 2-2 and he scored a header in the second half. You never know what might have happened in football, but I like to think we might have been better prepared if the report was taken more seriously.”
Oliver Gage – APFA CEO & Former Professional Analyst
Once you’ve added your comments to the section below, please move on to the next module.
I believe that from both a players perspective, as well as the perspective of a coach data is extremely useful.
By analyzing the data we can predict how a team is going to set up and how they want to play. Does the data say the team likes to create overloads in wide channels, or do they like to attack centrally and narrow? From the goal kick, do they play out the back or do they look to beat the press by playing long and trying to win second balls? Are they a team that like to press? Or do they sit in a mid/low block and let you have the ball.
However, football is a game full of uncertainty and the reality is we never really know how a game is going to be played, we can only do our best to prepare for how our opposition might play. Focusing “too much” on the data may deter your side from focusing on what they do best and possibly playing a different style only because the data said so.
I´m currently in the role of a player but whenever I saw data from my team or myself it helped me to understand where to improve in certain areas or situations. I think Data puts a game into perspective and gives it an objective touch in addition to personal impressions of the game, whether that´s a coach or player. In terms of preparation for an opponent, having data available can be huge because it allows you to prepare in a certain direction. Of course, football is unpredictable at times but it´s unlikely for a team to play a totally different system and personnel within a season, and knowing the tendencies of an opponent can´t be a disadvantage, ever.
A risk would be to depend on data too heavily and ignore the strength of your own team. Data needs to be interpreted well and looking at a number or percentage never tells you the whole story. To pick up the example from earlier, a striker with a higher percentage of shots on goals than others in the league is one point but it also means that the service from the players around the striker arrives at a higher percentage as well. Therefore you have to as yourself how to stop service to the striker rather than just shut down the striker on his own.
I joined a collegiate program with an “old school” head coach who liked the idea of starting to use Data and brought me on as an opposition analyst. I wrote a report for our upcoming match and presented it to him and with almost every data set I was met with “well the data doesn’t always paint an accurate picture”. Essentially I had run into a situation where the data presented information that contradicted the way the head coach had wanted us to play. So do we follow the data? Or the way we want to play? Do we try to incorporate the two? It can be a difficult decision especially when you begin to clash minds but at the end, a difference of opinions or viewpoints is always more valuable than someone agreeing to just agree.
I was asked to scout our next opponent and see where their attacking opportunities were most likely to develop from so we could then come up with a strategy with the coach. When looking into the Wyscout data from their previous three matches, we noted that they had a higher number of line breaks down the left hand side which lead to a higher number of crosses from that side into the box.
Having a look at the passing network from the left hand side, the oppositions attacking midfielder, striker and left winger had significantly more passes between each other. Upon inspection of the video, their line breaks came in the form of two variations: Attacking midfielder receiving with time and space and slotting the ball through to the winger running in behind OR the winger dropping deeper to drag the opposition fullback out of line, and then combining with the midfielder to release their striker in behind.
So with the data combined with the video, I presented to the coach and we worked on two potential options: Limit the space in behind by dropping off earlier OR making sure their attacking midfielder was tightly marked since the majority of the balls through came from his boot. We also anticipated that if their midfielder was tightly marked and couldn’t find the time to play in behind, they would start playing more direct balls from deeper areas to release their winger.
The game ended up in a draw but we found that we were able to limit their attacking threat down the left hand side.