FC Cincinnati hired Alexander Schram as director of analytics and strategy last summer, hoping he could help the club in the scouting process to rebuild the roster under general manager Gerard Nijkamp.
The job has evolved into much more, and it’s clear Schram plays an important role with the growing club.
Schram worked with Nijkamp at his previous club, PEC Zwolle, as a consultant helping with analytics in the recruitment process there. After Cincinnati hired Nijkamp last May, the general manager recruited Schram to fill the open analytics job with FCC, beginning that position September 1, 2019.
Now, almost a year later, he’s busy helping the team prepare for upcoming matches, evaluating the current roster and helping the scouting department make decisions on targets for the summer transfer window (open until Oct. 29) and beyond. Schram took some time Monday to speak to me about his job and provide some insight on how FCC uses analytics and data.
Here’s what we discussed….
Question: How did you get into the field? And did you come from a soccer background? Or were you more just on the data side and decided to get into soccer with it?
Answer: Well, actually, I was always dreaming of becoming a professional player, but I soon realized I lacked the talent, so when I noticed that I had a talent with numbers, I soon decided that this is the area for me to go in to get into the sport. So, it’s something I knew when I was studying already that this is something I wanted to do, and after college (obtaining a Master’s degree in Econometrics at the University of Amsterdam in 2014), I got an opportunity to work at a Dutch consultancy firm — I am Dutch. I’m one of the Dutch invaders of FCC. And, I got a chance to work with a lot of clubs over there before joining FCC after Gerard Nijkamp joined the club.
Q: Did he recruit you or was it something you approached him about?
A: In The Netherlands with PEC Zwolle, I was there one day a week helping out with recruitment, using analytics in the recruitment process of that club. And that’s basically what I started doing here in the first place as well. So we did place a process of recruitment next to the already great scouting team we already have at the club. We also add analytics to evaluate players and see how they fit within the philosophy of the club.
Q: I assume you also do some data analysis to help evaluate the current roster? How would you describe the different aspects of your job?
A: I think my job is basically a 50-50 split between meetings with the soccer stakeholders within the club, including the coaches, and analyzing data on the other hand. When I started with the club, it was mostly recruitment, focusing on getting models and algorithms in place to evaluate players the right way. But now that that is in place and given the amount of time we got during the pandemic, we also worked a lot with the coaching staff, specifically Yoann Damet, to get reports going, not only about next oppositions or the next game we play against D.C. United. There are some reports on the way how they play, what are their strengths, what are their weaknesses. But also the players we got in during the previous transfer period, obviously there were reports on them before signing them, with strengths and weaknesses. Now that we have six games under our belt, we can also start looking into whether or not they are performing like the data was suggesting they would perform for us.
Q: How big is the data team you work with?
A: I am one of six sporting directors of the club, and basically the department is there to help out all the other departments. We currently have a team of four including myself. I run most data analytics together sometimes with some interns. Then we have Michael Goldman, who’s specialized in making sure we are salary cap compliant, and then we have our Head of Performance Vincent North and Performance Analyst Diego Martinez, who both do all the video analysis for the first team.
Q: With the summer transfer window open, has your job gone back to more focus on the player recruitment side? How often are you getting requests on evaluating potential player targets?
A: No, for sure, it is noticeable the market is open. We work in two ways. Obviously, there is the traditional way of players being offered to us through the networks of Gerard or Hunter (Freeman) and the coaches or through agents or whatever. And obviously, it’s a very busy period with names coming in every day. And, second, we use the data to evaluate leagues we don’t see a lot ourselves. So let’s say, for instance, the Slovakian League or the Czech Republic, we don’t watch those leagues a lot, so we first use data to basically go back from every player that’s in the league is a potential target to: ‘Here are the five best players for that position. We have to take a look at them in that league.’ And that’s something that’s more continuous throughout the year, that can be done every day of the year, whereas the other part is much busier now of course.
Q: Can you share some of the metrics you are looking at when evaluating players as potential targets?
A: That’s something, obviously, we have to be a little bit cautious of not giving too much away when other teams are able to read as well what we are doing. One that I already gave away to the MLS website is the one where we basically try to combine metrics into one (referred to as N’Golo Kante Index). So we look at the number of interceptions a player makes and the ratio between that and the number of turnovers he has, and that tells you something about how good a player is at both retaining the ball, but also winning it back, which for some positions is obviously very important in the way we want to play. Second, of course, one that everybody is looking at are expected goals (xG), specifically for forward players. That’s something we use a lot to evaluate forwards. We also have some models to incorporate both expected goals scored and expected goals conceded for all positions, including defenders and goalkeepers.
Q: Is there a metric you think is under used? Is there some sort of data/metric you would like to see become more of the regular soccer lexicon?
A: Not specifically one metric, but some stuff we are doing is much more focused on the predictive side of things, and what I see around a lot in the great soccer analytics community online is a lot of descriptive statistics, so looking back. And what we try to do is also build our models and algorithms in such a way that we can make valid predictions of the future, which is obviously a very, very important part of recruitment. So, not specifically one metric but I would like to see the community go to more predicted stats than descriptive stats because it speaks more.
Specifically, when recruiting in other leagues, to find the relationship between a performance in, let’s say, the Dutch Eredivisie, which we recruit a lot from, and MLS. To keep it as simple as possible, how much is 10 goals in the Eredivisie in MLS? How does a defensive performance in the Championship translate to performance in MLS?
Q: So, is there an algorithm designed to predict that or how do you come up with those?
A: Well, that’s where analytics comes into play and where you try to build your models in such a way that it takes into account what’s happened in the past to build towards the future, because obviously there are players moving from league to league for a long time now, so you can use this historic data to build such models.
Q: Do you get asked to research specific questions by the coaching staff or front office or do you have free rein to give input based on what you feel like is most important? What is the balance there as far as what they want versus what you feel like would help them?
A: Soccer is leading within the clubs. It’s a very dynamic game, so it’s hard to completely explain it using just data. There is a thing in this game, which is called soccer expertise, which I listen to a lot. So I would say it’s a 70-30 split between the soccer side telling me their ideas, their philosophy, how they want to play, what they want to see from a player in a position and me translating that into metrics that can tell us something about that, then 30 percent me coming up with metrics and then testing it with them, to see whether or not there’s something there.
And in terms of preparing for games or looking back at games, I think there’s a question that needs to be answered, a question from a coach that needs to be answered by me, instead of me providing answers to questions that are maybe not even there because the coaches deal with a lot of information already. So, I want to hear the right question first before giving answers.
Q: How much does the job change depending on who the coach is?
A: Well, actually not that much, and I think that’s a very good thing within this club is the organization has its philosophy and principles and it doesn’t so much depend on the head coach that is in charge. Obviously, Yoann is still part of the coaching staff, and he’s the one providing that consistency. So, I don’t see any major differences between the transition from Ron (Jans) to Jaap (Stam), but obviously there are also two Dutch coaches having not exactly the same philosophy but a similar way they want to approach things. Maybe it would have been different, but that’s something we don’t do as a club is change your future philosophy completely by signing a head coach from a total different culture.
Q: Where do you think soccer is relative to other sports with data and analytics? Do you study what goes on in other sports?
A: That’s a little tricky for me to answer, being Dutch and having this game as my main sport to follow. But what I know, what I hear from baseball, where — well, I don’t like to refer to the ‘Moneyball’ story too much but that story made analytics departments of other teams grow bigger also. I think that’s where soccer is right now with I think Liverpool leading the way and showing how successful you can be with having the proper analytics, especially in recruiting. I saw something today, if you look at the net spend of Premier League clubs over the last five seasons, that Liverpool is ranked 14th in the Premier League. And well, we all know how well their season has gone, proving that even in a sport as big as soccer is, you can still make a difference by being maybe smarter than the rest of them — I don’t know that smarter is the right word, but using analytics in the proper way. And to be completely fair, I don’t know anything about the other American sports. Another example is Brentford, a smaller club that is successfully using analytics as well.
One thing I can add, we are also looking a little bit into what the Cincinnati company Pro Football Focus is doing for NFL and college football teams and see how we can apply their approach maybe into this game. But again, I don’t know the sports, so I don’t know what’s good and what’s bad when you play football.
Q: Maybe a better question would be just how do you see data being used in MLS versus what you are used to in The Netherlands and what you saw throughout Europe? Is MLS, as a younger league, behind?
A: Well, in terms of analytics, I don’t think the MLS is behind many leagues to be honest. And I think that’s partially due to the culture that comes from the other sports in this country, where it’s part of the whole experience of watching a game is also having statistics about the game and metrics about the game available. So in a way, I don’t see MLS as being behind other leagues. And maybe it even helps that everything is relatively new. So one of the problems I encountered with some clubs over in Europe is that you have the club legend, the club icon, to put on their weight in conversations about the game purely because they were big players or coaches from the past. And so no, I think MLS from my experience, and I have to say, I see FC Cincinnati and I am in contact with some of the other directors of analytics but I cannot I cannot tell the story of the other club of course, but definitely the league is not behind compared to the rest of the world.
Q: How much of the soccer data that gets used by clubs is public vs. private? For example, some suspect MLB uses defensive metrics that aren’t public. Is there some of that in soccer too?
A: Well, I think this is maybe where a distinction between the data sources come into play with event data, so that’s the stuff happens on the ball, which is tracked by multiple companies. That’s also widely available for the public. But within the league, there are also optical tracking data, so capturing every player and its position and the ball 25 times a second, so you can imagine how big those data sets get. Those are not publicly available, but they can help us a lot when looking into stuff that happens off the ball, which is very important in this game, when our winger makes a good run forward, but he doesn’t get the ball, then this will not show up in event data, whereas stuff like that we can get from the optical tracking we get from Second Spectrum. The biggest part of data that is kept within the club itself is the optical tracking data.
Q: FCC is playing a Dutch style that you are familiar with, but what do you notice about the style of how the game is played here overall?
A: That’s something I also use some data for, to look into it not always trusting my own eyes, but I think it’s mostly the philosophy or the pace of the game. So, a lot more transitions, going back, going forward and going backward more often, which obviously asks for a different skill set now and again. At this point in time, we don’t have any teams doing completely different stuff like Red Bull with their pressing. The Dutch style, with the possession-based trying to build up from the back, leads to maybe a more technical game to watch, but
it’s a little less quick, so that’s a big difference between the two I would say. That’s also something you could measure using some simple metrics like number of passes per second, or number of touches per second, and stuff like that. That’s higher in the MLS. And of course, I have to add, coming into this league September 1 last year, I have seen maybe 10 rounds of games because of the pandemic. So I’m hoping to get to know the league even better in the coming and coming two weeks with the return to market play.
Q: For those that maybe share a similar background as you, where they realize they don’t have a future playing the game but enjoy the data side of it, are there any suggestions you would provide in terms of what they should try to learn?
A: First and foremost, learn how to code. If you want to work at FCC, you learn to code in Python or otherwise learn to code in ‘R,’ but you definitely have to be good at coding. Second, go online, go on Twitter and find the soccer analytics community. There is great stuff out there. There is great data available to play around with and to start up with. And, don’t be afraid to reach out to me or other counterparts at other clubs because we are all open to suggestions because this is a field that is still developing and any ideas are always welcome. It’s a little cheesy, but there’s no such thing as a bad idea when it comes to analytics until you try it.