Football's Next Competitive Edge: IMPECT

Jens Hegeler

co-found a company called Impact together with a teammate of mine at the time, Stefan Reinhardt and then two other friends who were from school, Lukas and Matthias. And yeah, founded that company and then a year ago, roughly joined Catapult after the acquisition. I'm now, the title says I'm head of commercial for global scouting, which basically means that I'm looking at the commercial side of the Impact product vertical within Catapult.

Richard Graves 01:21

In business terms, things have moved pretty quickly, haven't they, from the origin when you founded Impex itself to where you find yourselves now as part of the Catapult organisation. Going back to the beginning, you were still playing professional football, I think I'm right in saying at the time. So what was it you and your three other co-founders saw in data analytics systems that were being used then that you thought left a gap for what you developed?

Jens Hegeler 01:52

Well, I think we joined the topic by coincidence basically. So Stefan, my teammate at the time, he was attending a debate at the University of Sports in Cologne, which was about the use of data at that time. So that's now 10 or 11 years ago. And he was actually invited to a panel discussion because Professor von der University was presenting his new book, which I think had somewhat the name. the truth about football basically. And it was about his research that he had done around football analytics in football. And was invited to participate in this panel discussion and discuss how data is communicated to us players. And the interesting outcome of this panel discussion was that Stefan learned on this occasion that none of the known metrics had a significant correlation to the end result in football. So it didn't matter if you had more ball possession than the opponent, it didn't matter if you were running more than the opponent because those

those metrics wouldn't correlate to winning a football game. And that was basically the starting point. So Stefan went back from this event and the next morning we trained together and he was like, you don't believe what happened yesterday to me. I was on that event and I've learned all the metrics that we see after every game in our dressing room. And we had a sheet of paper in our dressing room at the time. They don't mean anything because like we're not more like to win a game of football just by having the ball more than the opponent. And yeah, that was starting point for us.

to look into it and try to understand why there are actions in football that are more valuable than others and how can we make that measurable. And yeah, I a closer look at the topic.

Richard Graves 03:23

Well, let's get into that in a little bit more detail because to the untrained eye, what was it that you and Stefan felt that coaches, analysts, people working in recruitment were missing when assessing individual performances?

Jens Hegeler 03:38

I think what we figured out quite quickly that in football all of the metrics that were used and were created were just looking at the quantity of action rather than the quality. So you were measuring the number of passes, the number of touches, but not necessarily the impact that those actions had. And that was the first observation. The second one that was interesting, a of metrics that... that were used in football were coming from other sports, so weren't really designed for that sport, so passes into the final third is a metric that is common in ice hockey, but not necessarily in football, because there are no thirds on the football pitch, there's no real advantage being in the final third. There's an advantage being in the box, because you might get a penalty if you receive a foul, but in the final third it's not necessarily a good thing. So that was like...

The first observation that a lot of metrics have been used from other sports and not necessarily fit to football and a lot of metrics look at the quantity of action rather than the quality. And that what we've tried to not to solve but we would try to find better ways of measuring the quality of action. And that's what we then later called PECIC which was our initial kind of view on data that we need to collect.

Richard Graves 04:51

I'm glad you brought up packing because it sounds, I think, a little bit strange in a football context. People might listen to this. Hang on, we're talking about football. What do you mean when you're saying packing? But this is one of the key concepts of the Impex system, isn't it?

Jens Hegeler 05:06

It is, yeah. mean, in simple terms, we're trying to look at is the impact that each action has towards the opponent's formation or towards your own formation. Packing the name is just a name that we decided to give our, this concept or this approach. The reason for that is that the name didn't exist and we wanted to protect the idea basically to a And back in the days, you had that game of Pac-Man where like,

You eat balls, I'm not sure if you've played that as a video game back in the days, long time ago. And the idea was like, okay, we're eating opponents with an action when we're taking opponents out of a game with a successful action. So the concept of packing is in any given situation, we're looking at where are you able with this action to eliminate, to bypass opponents and get them out of play? Or did you eliminate your own teammates when they gave the ball away and try to...

Richard Graves 05:40

Back in the day, yes absolutely.

Jens Hegeler 06:05

try to make that measurable because it has an effect obviously if you're breaking lines, if you're eliminating opponents with a pass because you're more likely than to score a goal in the next action because you've created a better situation for your team.

Richard Graves 06:18

So in other words, the concept of packing isn't just about an individual completing passes or moving up and down the football pitch. It's the impact of those movements, the impact and effectiveness of those passes that gives it context in a game time situation.

Jens Hegeler 06:35

Yeah, exactly. So it looks at every action. So not like at some moments we're looking at it some not. Like in every action we're looking, are you eliminating opponents or not with this action? And it's a holistic approach. It's not just looking at passes or dribbles. It's looking at every action that's taking place. It could be a ball loss, could be a ball win, could be an action in possession where you, with a pass, with a dribble, with a cross.

So it's an holistic approach that I'm trying to understand if you were able to beat the opponent with this action and how you did that. That's basically your choice. It's your weapon of choice. If you dribble past opponents, if you cross past opponents, if you play Dijkners or if you just win the ball in good areas. But in the end, need to eliminate opponents because that's what will make it more likely that you're going to score a goal.

Richard Graves 07:24

The concept of packing has now led to something termed the packing expected threats. How was that built on the original concept and what does that tell us?

Jens Hegeler 07:35

It's actually like an evolvement of it. So pecking itself is looking at how you're eliminating opponents and it doesn't really matter where that's happening on the pitch and when it's happening if it's in your own half in a counter situation or in the final third. While pecking expected threat is maybe the more advanced version of it. We're not just looking at how you're eliminating opponents but also...

Where is it happening on the pitch? Which opponents are you eliminating? Is it the opponent strikers? Is it the opponent defenders? Because there's a difference in the value that you're creating for your team if you're obviously eliminating opponent defenders because it's harder than maybe eliminating opponent strikers, which might happen more often in a game. the expected threat model is basically more advanced version of that trying to bring more factors into consideration and it's trying to put like a more precise value, how much you're

increase in the chance of scoring with an action.

Richard Graves 08:28

And this appears, I know you're working with organisations right across Europe, but certainly in Germany, professional football in Germany, you seem to have tapped into something here that's proving popular, receptive with clubs, and that in itself has led to a measurable success with teams like Leverkusen and Leipzig amongst others.

Jens Hegeler 08:50

For us it was both, Stefan and both German. So was obviously the most obvious place to start with and showcase our idea and our data. I think that worked out quite well for us. Teams are successful for various reasons. One of them is using data eventually in their decision-making process, maybe in scouting or in analysis. It's not the only reason for being successful, just one of them. I think in our case, what has helped that the idea of packing itself is quite easy to understand for decision makers, it's easy to understand for coaches, it's easy to understand for the owners of football teams, because the concept is in theory you could count it with your eyes if you wanted to. And that is also something that is very important when you're creating.

metrics that these metrics that you're creating that they are understandable for the people that have to use those and that their behavior that you want to change. So I think the idea of picking has a bit of an advantage there that it's a metric that is easy to understand for people that are working in the industry and therefore was quite useful for them.

Richard Graves 10:06

Yeah, and any concepts that gives people working in professional sports that 1%, that extra edge, perhaps a little bit of information that their rivals won't have is going to appeal. And that's why this no doubt is gaining traction. If we can talk in terms of a practical example, and perhaps you can explain actually how impact works, given a scenario. If I'm working at a football club, for example, Jens, and say I'm looking for a midfield, a one who can

receive the ball under pressure, is able to turn up field, progress with possession of the football and then is able to break the opposition line, be it with a dribble, a pass, whatever it might be. How does impact work and what information does it give you to inform a decision over whether or not this player might be for you?

Jens Hegeler 10:59

Well, I guess first it starts with, do you really need this kind of midfielder? Like, do you need this kind of player starting to analyze your own team and your own performance? And is that really the kind of profile that you need? And we'll try to inform that by showcasing what the data looks for your own team. Then if that is a problem in your team, that you need this kind of midfielder, then obviously through our data we try to...

try to help you find those players. And there different ways how you can access our data. You can either just access the raw data and get it into your own system. If you're a very advanced team in the Premier League and you have the resources to in-house work with our data, then you can obviously do that. The other part is you can use it through our software where we have different features in place that help you then to create those profiles, screen markets, run those profiles across different competitions. We'll try to help the user understand.

What is a performance in a smaller league worse compared to a performance in a bigger league? So a typical example of a player outperforming the Dutch Eredivisie, what would that mean for a Premier League team? These numbers, so we try to help him with that and in the end it's... The idea is that it should feel a bit like Football Manager where you as a user can obviously define those criteria based on the data and then you get feedback based on data which players could fit quite well to what you're looking for and help you narrow down that search a little bit. And then since we've most recently launched our video scouting solutions you can then obviously also do the eye test and obviously have a look at the player yourself and get just a view on the player and have the...

footage on top of that.

Richard Graves 12:48

You're taking me back to my younger days now. I was very successful at times with all sorts of clubs on football manager. Unfortunately, never developed the skills in the real world to progress a professional career in football management. But as is always the case with any data and analytics systems, context is key, isn't it? I know you touched upon it there. How do you separate, for example, if a team dominates possession, 75 %

possession of the game, it stands to reason that an individual will see more of the ball and will have more opportunities to progress possession forward, break lines, be incisive with their passing. Equally, it might be that you play for a team that often doesn't have the majority of possession, but it turns out that you are more incisive with the possession you have, more productive with it. Does the Impex systems give that analysis and does it put it into that context?

Jens Hegeler 13:44

You're right, mean in general the more context you have the better it is, the more precise is your feedback on a player. You're right, in some cases it makes sense to try to look at the team context. It's always the trickiest part, like how much of the performance is down to just the quality of the player and how much of it is down to the quality of the team. So if like the overall assumption is in a good team it's easy for everybody to shine.

I don't necessarily believe that because in a big team there are lot of big players and lot of big players also demand a lot of actions and demand lot of situations with the ball so it's not necessarily easy for a central... Defender at Main City to create high numbers because you will have a holding midfielder who wants a ball all the time and wants to be like the the progressive part of the of the build-up. You might have a left back or another center back next to you who wants to be very progressive. just being in a better team doesn't necessarily mean it's easier for you to create higher numbers. But yeah, in some cases it makes sense to adjust those numbers and

like try to delete the team context a little bit and look at the ratio of a metrics in relation to what were the numbers that the team achieved in this metric to get a clearer picture of those players. But usually if you're a ball playing center back, this will always shine through, like regardless in which team you're playing, even if you're playing a team that doesn't build up from the back, it will still show that you are the one that...

in those few moments where you're building up are the one that is building up. So there are some elements in a player prover that usually always shine through regardless of the team contact basically.

Richard Graves 15:29

Is there ever a concern, you've played football at a high professional level, obviously you've helped develop this system, is there ever a concern that in modern football there are occasions when we become too reliant on data? Is there still room for human judgment and the old-fashioned eyeball test?

Jens Hegeler 15:48

100%. I mean, there are things that we don't measure good enough with data. So I don't think it's going to change. So I think data will always be like, or can be obviously a bit like a red flag. can show. Basically on the other side, okay, this player is something that the eye test says like is a guy we should go for, but the data really flex them as a risky signing here. So I think their data can be very supportive in protecting you from doing like huge mistakes because the player might look better on a video than the outcome in reality is. But at the same time, there are things that we were just not really.

measuring good enough where it's important to watch a player and get a sense for his qualities on video. And that's why, obviously, in our case, we've added this element now because we think it's needed in the scouting workflow and in the decision-making workflow. And I don't think this will go away, so to say. So the eye test needs... The player always needs to fit the eye test, I guess.

Richard Graves 17:01

And obviously since the acquisition by Catapult, it works for Catapult aligning what Impec does with its own systems, but it opened doors, a lot more doors in theory I'd imagine for Impec, potentially new clients, new systems to work alongside. How do the two mesh and what is it now enabling you to do that you weren't before? I know back in the summer there's a new video scouting system that Impec's brought out as well.

Jens Hegeler 17:28

Yeah, mean, various advantages that came through that acquisition. The most obvious one, you're part of a bigger team now and you can leverage the knowledge of that team. the local knowledge in different markets, having people on the ground that can support teams, which is huge benefit for us, that we have now people that speak the language, are in that market, that are present and can help the football clubs get the best out of other products. That's obviously huge advantage. Then obviously, Kanepold has always been strong with the products that they already had, with the ProVideo Suite, with the wearables side of the...

of the product range. So trying to connect the dots here and making those products being better to each other and helping the clubs to get a more rounded picture of their performance is super interesting and also super interesting for us to help the company to achieve that. And then if we want to develop new parts... like we've done in summer with the video scouting solution. It's also easier within a bigger organization where there's just more power in place, more knowledge, more expertise to create that. I think, as you mentioned, the video scouting pieces isn't a good example of that. Being able to acquire rights on scales from other rights holders, making sure that we get the software in place in time, connecting with the current product that we already had within Impact and Cutout Pult.

was quite a big task and we're actually quite happy with the outcome of it. And that was definitely only possible because we are part of a bigger company now.

Richard Graves 19:10

Yeah, let's be honest. Users across any sphere want to have systems in place, tech in place that talks to each other, that marries, that meshes to give them the greatest overall all-around picture. And we started this conversation by saying when you and Stefan first spoke about going forward with this idea, it was because you felt traditional data and analytics systems weren't giving you footballing answers in the correct context. So when you look at

the systems that Catapult already has in its portfolio, be it working with Perch or the Vector systems, and you align it with the tech and information that you've got. Is this another step on the road from your perspective to giving that all-round football picture?

Jens Hegeler 19:54

Yeah, 100%. I guess, I mean, that's also the aim in the end, that we can provide this well-rounded picture. You just said it, obviously, with Purge measuring maybe indoors or in the gym, what is happening with the World Cup, what is happening outside on the pitch. I think we're already in a position where we can provide this information and it's now about connecting those and helping clubs to fully understand those different insights that these products generate.

That was also very intriguing to us. I Catapult always, like we always considered Impact as like Nestle's company because we've played ourselves and we felt like it's coming from football, for football was a bit like our slogan. And with the feeling that Catapult has a similar view on it, it's also like... an athletes company because there are like a thousand athletes around the world that wear catapult products. And so we feel, felt like it's, very like-minded in that sense in trying to help football organizations to better understand the performance of their athletes. And may it be for recruitment, may it be for performance analysis. And I think we're on a good way here to, yeah, to getting a better picture of that.

Richard Graves 21:05

It stands to reason then that when a door such as this opens, it opens opportunity as well, doesn't it? So if we throw ahead in the next three to five years, for example, what do see the future looking like now? Because I've already said that you appear to have come a long way very quickly in what is it 10 to 12 years or so. What are the opportunities as you see them now? What's the direction you see it taking?

Jens Hegeler 21:30

That's a very hard question to answer because obviously if I would know how the future would look like then would obviously we would be very very in a very elite position to create products for that. I think what won't change maybe the other way around so what would not change in the next three to five years I think clubs will will rely on data to make that decision. I think this part is growing more and it's not going away. And as Catapult being in a position as a company there to provide a lot of insights, to measure a lot of data is a very powerful position to be in. Obviously, for us, it's about better understanding the insight that we can create from the data that we're collecting made on the performance side, but also on the scouting side.

help clubs to gain more insights out of the data and out of the tons of data that we're collecting.

Richard Graves 22:37

Will AI have a role in the coming years? the tech that you give us now tells us exactly what an individual athlete is doing, how they're moving, where the potential red flags might be so you can beat them in advance if you like. Artificial intelligence is the great unknown frontier, isn't it? And it's a question that you find many in the sports science world.

Jens Hegeler 22:39

ffff...

Richard Graves 23:04

are excised about if for no other reason than we don't actually know the answer.

Jens Hegeler 23:10

Yeah, it's a big obviously will be game changing and what I just mentioned, think, of both sits a lot of data and obviously having AI in place help us leverage that data in better way is a powerful powerful position. I think it will be will be key. So we're looking at it quite excited, to be honest, because we we hope it helps us to understand and connect the dots a little bit better for us as well. And

Yeah, I think that's it for now because it's so quickly changing and improving that I don't want to make a prediction here how the looks like in three years time because I might be outdated in three months already with my assumption here. But yeah, I think...

Richard Graves 24:00

Such is the rapid pace of technology and knowledge, isn't it? You're absolutely right. Jens Hegel, it's been a pleasure talking to you. Before we let you go, those that have been drawn in by our conversation, where can they stay across what you're doing and the progress of impact in the Catapult family as it is now along social media?

Jens Hegeler 24:21

I mean, the first point of contact would always be the Catapult website. I think that's where we'll try to inform about latest developments, product updates, in general what we're doing here at Impact and Catapult. I think that would always be my first point of contact. I mean, I'm always reachable via LinkedIn or these social media platforms. So, yeah, happy to connect with anyone who likes to connect.

Richard Graves 24:46

Well once again, thank you very much for your time and wish you the best of luck going forward.

Jens Hegeler 24:49

Thanks Richard, it was a pleasure.