Why Fortuna Düsseldorf crashed II
Matthias Werner
I turn disconnected processes into data-backed growth engines | RevOps | BizOps | Data & Analytics
Or: Would have been Onisiwo and Mehmedi the better choice?
Within the scope of the previous article, it was tried to investigate the reasons for the significant crash of Fortuna Düsseldorf after playing a very solid campaign in 2018/2019 Bundesliga. While it seems their playing style did not change too much, their key personnel certainly did.
As already suspected, Dodi Lukebakio and Benito Raman had a significant impact on the 2018/2019 season’s performances. Both scored 10 goals each and further collected 4 respectively 3 assists. This results in a share of approx. 42% of the whole team’s goals and 20% of all assists. The graph below shows some of the team’s metrics and the share that Lukebakio and Raman have contributed.
The fact that these two players have been of major importance for Düsseldorf is hardly questionable, but how do the two compare to their peers, the other forwards of the 2018/2019 Bundesliga season. To approach this question, the following scatter plot shows the ratio of assists (A) to expected assists (xA) as well as the ratio of non-penalty goals (npG) to non-penalty expected goals (npxG). Fortuna Düsseldorf players are coloured red, their peers grey. The larger the dots, the more goals the respective forward scored in absolute terms. In order to strengthen the meaningfulness, only players who have completed at least 30% of the possible minutes of play were considered. The data shows that Benito Raman quite strongly outperformed the expectations when it comes to goals and assists. Dodi Lukebakio, however, overachieved the expectations regarding assists but also slightly underachieved in terms of goals, similar to his teammate Rouwen Hennings.
Naturally, the curse of good performance also hit Fortuna, as other, wealthier clubs, considered the two as transfer targets for the summer window. Finally, Raman was sold for reported € 6.5M to Schalke and Lukebakio joined - by a detour - Hertha BSC Berlin. As replacements for their former life insurance, those responsible decided to sign Nana Ampomah, a left winger from Belgian side Waasland-Beveren, and Dawid Kownacki, who already was on loan from Sampdoria during the second half of the 2018/2019 season and seemed to perform very well (4 goals and a 2,86 npG/npxG ratio). Both, Ampomah and Kownacki, cost Fortuna a reported € 11.5M in transfer fees. One may argue that Kownacki heavily outperformed during his loan so that regression to his mean performance could be expected sooner than later. Indeed, looking at the same scatter plot for the 2019/2020 Bundesliga campaign reveals that Kownacki did not score any goal, although 1.9 would have been expected. Also, he did not deliver any assist, while the xA ranges at 0.7. Another new signing, Erik Thommy, performed well over the expectations but on lower absolute levels than Raman and Lukebakio did (2 goals, 4 assists). Meanwhile, Ampomah did not even reach the threshold of 30% playing time to be considered in this plot, maybe related to the injury he suffered from until week 4 of the season.
Finally, for clubs with more limited financial resources, it is crucial to find exactly those (hidden) opportunities, to outthink the competition instead of outspending.
Just looking at the data, it appears Fortuna Düsseldorf has been a little unlucky at the transfer market and the replacements performed not as well as they were supposed to do. This, of course, raises the question of what might have been a better substitute. The upper graph already shows that Karim Onisiwo performed comparable to Benito Raman and Admir Mehmedi even better in terms of npG/npxG ratio. Looking at such ratios always brings up the question of whether the overperformance was achieved by actual skill or mere luck. In such cases, it is always recommended to look at multiple seasons. It is interesting that Raman, Lukebakio, Onisiwo, and Mehmedi all have been able to outperform their npxG and xA in the 2019/2020 campaign again (until week 21), which indicates that they all are skilled above-average.
So, would Fortuna have been better off with signing Onisiwo and Mehmedi instead of Ampomah and Kownacki? At least the data indicate that this could be the case. During the transfer window, clubs do not only need to account for their demands regarding skilled players but also have to consider their economic limitations. Onisiwo, during that time, had a market value of € 3.5M, which rose to € 4.5M by today. With contracting Ampomah for reported € 4.0M, Düsseldorf invested indeed in a fairly similar range. Mehmedi’s market value after June 2019 was estimated at around € 10M, a little more than Fortuna paid for Kownacki (€ 7.5M) but not too far away to be considered as completely unrealistic - at least from the mere economic perspective. In the end, the question if Fortuna Düsseldorf would have performed better with other signings remains a thought experiment - but still, it’s interesting to see what honest data may tell about it. Onisiwo and Mehmedi most likely would not have been everybody’s first consideration in such a discussion - and that is the core idea of this article. Data often helps to think out of the box and to discover opportunities others do not detect. Finally, for clubs with more limited financial resources, it is crucial to find exactly those opportunities, to outthink the competition instead of outspending. Of course, to have a more profound picture one could expand the analysis over more seasons or add some more sophisticated statistical approaches. Nevertheless, this article primarily wants to simply draw attention to the possibilities deliberate data utilization can add to professional football clubs.
Thank you very much for reading. If you have any kind of feedback please do not hesitate to contact me. I am always happy to jump into a chat about football and/or data ??
______________________________________________
Market values taken from https://transfermarkt.de
Sales Director @ QueBIT
5 年Matthias, this is actually very cool. Hope you can extend this to the NBA and NHL in the future! ;)