An exploratory study into select technical key performance Indicators and estimated transfer fees in the 2nd division of the German Bundesliga

Date
2023-03
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: In light of the escalating transfer fees in professional football, an inflated market and clubs having to adjust their budgets due to the ongoing restrictions resulting from the COVID-19 pandemic, accurate financial management of sporting organisations is required. This especially applies to football clubs playing outside Europe’s big five leagues, as well as those in the lower divisions of professional football. The overall aim of this MSc project was to shed light on a second division league which has previously received less attention than first divisions. More specifically, the current study assessed how a player’s performance in the season immediately before his transfer influenced his transfer fee. This was done through the lens of the Moneyball approach, which states that specific characteristics and/or skills are being undervalued by the labour market. As transfer fees can rise to millions of euros paired with clubs being under financial pressure, quantitative rather than intuitive decision-making is required. This study assessed the relationship between select technical key performance indicators (KPIs) and their relationship to estimated transfer fees in the 2nd division of the German Bundesliga, using publicly available data. The selected KPIs for defenders, midfielders and strikers were tackles and interceptions, tackles and assists, and goals and assists, respectively. A statistician was consulted for a sample size calculation and revealed a minimum of 12 participants per position (defender, midfielder, striker) were required. Statistical analysis was performed with the use of GraphPad Prism (Version 9.3.1, GraphPad Software, LLC, USA). Prior to statistical analyses, normality tests were conducted using the Kolmogorov-Smirnoff and Shapiro-Wilk tests. A Kruskal Wallis test was employed to assess the differences between the various player positions. Where significant differences were found, a Dunn’s post hoc test was performed. Associations between variables were tested for by means of Spearman correlations. All data are presented as median, and significance was set at p < 0.05. Analyses revealed that interceptions have no association with the transfer fees of defenders (rs=0.06, p=0.76). Tackles were found to be negatively associated with the transfer fees of defenders (rs=-0.41, p=0.03), while no association was found between tackles and the transfer fees of midfielders (rs= 0.17, p=0.24). Midfielders’ transfer fees displayed a positive association with assists (rs=0.44, p=0.00), while no association was found between the transfer fees of strikers and assists (rs=0.15, p=0.34). In addition, strikers’ transfer fees also did not display any association with goals (rs=0.22, p=0.16). As interceptions were found to be negatively associated with the transfer fees of defenders, it is possible that interceptions are being undervalued. However, as more successful teams tend to dominate ball possession, it is likely that players from less successful teams have more opportunity to intercept a ball, and, as less success reduces transfer fees, this would explain this association. Tackles, on the other hand, may display a negative association with transfer fees of defenders as tackles may be a sign of an error prone and/or tactically inapt player, as tackles may be used to recover from tactical and/or technical errors. Assists’ positive association with the transfer fees of midfielders is explained by the increased chances of success with an increased number of assists. Consequently, one would also expect such an association for strikers. However, previous studies have found that the transfer fees of strikers in higher divisions are associated with a publicity premium. Consequently, it is hypothesised, that due to the reduced media attention in the lower leagues, this publicity premium only exists to a much smaller degree and thereby reduces the extend to which assists influence strikers’ transfer fees. The same reasoning applies to how goals influence a strikers transfer fees, as goals also did not show an association with the transfer fees of strikers. As the current study did not control for players moving up or down a division, results need to be interpreted with caution. The findings support previous research with tackles displaying a negative association with the transfer fees of defenders and assists displaying a positive association with the transfer fees of midfielders. However, as opposed to most other leagues, goals are not the determining factor for the transfer fees of strikers transferring to or within the 2nd division of the Bundesliga. This implies that other characteristics and/or KPIs may be at play that impact the transfer fees of strikers significantly more than the selected ones, and future research is tasked with identifying these.
AFRIKAANSE OPSOMMING: Gesien in die lig van voortdurende stygende oordragkostes in professionele sokker, 'n opgeblaasde mark en klubs wat vorseer word om begrootings aan te pas as gevolg van voortgesette beperkinge weens die heersende COVID-19 pandemie, vereis vaardige finasiele bestuur van sport vereenigings. Dit is veral van toepassing op sokker klubs buite die Europese groot vyf ligas, sowel as die laer afdelings van professionele sokker. Daarom is die huidige studie op die 2de liga van die Duitse Bundesliga uitgevoer. Die bevordering van analise tegnologië in professionele sokker lei tot die toeneemde beskikbaarheid van groot hoeveelhede data. Klubbestuurders moet dus bewus wees van hoe spelerdata hul oordragfooie beïnvloed. Deur gebruik van beskikbare publieke data, het hierdie studie die verhouding tussen sleutel-prestasie aanwysers (KPIs) en die verhouding tot beraamde oordragfooie in die 2de divisie van die Duitse Bundesliga analisier. Die geselekteerde sleutel-prestasie aanwysers vir verdedigers, middelveldspelers en doelskieters was ‘tackles‘ en ‘interceptions‘, ‘tackles‘ en ondersteunings (assists) en doelskoppe (goals) en ondersteunings (assists), onderskeidelik. 'n Statistikus is geraadpleeg vir 'n monster grootte berekening en het bevind dat 'n minimum van 12 deelnemers per posisie (verdediger, middelveldspeler, doelskieter) benodig word. Statistiese analise is uitgevoer met die gebruik van “GraphPad Prism“ (weergawe 9.3.1, GraphPad Software, LLC, USA). Alvoorens daar met statistiese analise begin is, is normaliteitstoetse gedoen met behulp van die “Kolmogorov-Smirnoff en Shapiro-Wilk toetse“. 'n “Kruskal Wallis-toets“ is gebruik om die verskille tussen die verskillende spelerposisies te bepaal. Waar beduidende verskille bevind is, is 'n "Dunn's post hoc-toets" uitgevoer. Assosiases tussen veranderlikes was getoets deur Spearman korrelasies. Alle data word as mediaan aangebied, en betekenis is ingestel op p < 0.05. Analise het aan die lig gebring dat ‘interceptions‘ geen verband het met die oordragfooie van verdedigers nie (rs=0.06, p=0.76). ‘Tackles‘ is gevind om negatief geassosieer te word met die oordragfooie van verdedigers (rs=-0.41, p=0.03), terwyl geen assosiasie tussen ‘tackles‘ en die oordragfooie van middelveldspelers gevind is nie (rs=0.17, p=0.24). Middelveldspelers se oordragfooie het 'n positiewe assosiasie met ondersteunings (assists) getoon (rs= 0.44, p = 0.00) terwyl geen assosiasie tussen die oordragfooie van doelskieters en ondersteunings (assists) gevind is nie (rs=0.15, p=0.34). Daarbenewens het doelskieters se oordragfooie ook geen assosiasie met doele vertoon nie (rs=0.22, p=0.16). Bevindings het getoon dat oordragfooie van verdedigers negatief met ‘interceptions’ geassossieer word, heel moontlik omdat hulle onderwaardeer word. Omdat meer sukselvolle spanne gereeld meer besit van die bal het, is dit waarskynlik dat spelers van minder suksesvolle spanne meer geleentheid kry om ‘n bal te onderskep, wat suksesvolle oordragfooie verminder. Dit is hoe hierdie assosiase verduidelik word. ‘Tackles’ wys ‘n negatieve assososiase met oordragfooie van verdedigers op omdat ‘tackles’ ‘n teken van groot hoveelhede voute van ‘n speler kan wees omdat ‘tackles’ dikwels gebruik word om van taktiese/tegniese foute te herstel. Die positieve assosiasies tussen ‘assists’ en die oordragfooie van middelveldspelers kan deur die grooter wen-waarskynlikeid met meer assists verduidelik word. Dus, die selfde assosiasie sou vir doelskieters verwag word, maar vorige studies het bevind dat oordragfooie van doelskieters in hoër ligas met ‘n publisiteits premie gepaard gaan. Dus, deur die mindere media aandag in die laer ligas, bestaan die publisieteits premie op ‘n baie kleiner graad en dus verminder die grootte van hoe ‘assists’ die oordragfooie van doelskieters verander. Die selfde redenasie kan gebruik word om te verduidelik, hoekom doele en die oordragfooie van doelskieters geen assosiase opwys nie. Die huidige studie het nie gekyk watter invloed spelers, wat van ‘n laer na ‘n hoër liga beweeg, op die oordragfooie het nie. Dus moet die bevindings van hierdie studie krities benader word. Met ‘tackles‘ wat 'n negatiewe assosiasie op die oordragfooie van verdedigers vertoon en ondersteunings (assists) se positiewe vertoonings op die oordragfooie van middelveldspelers, word die bevindings ondersteun deur vorige navorsing. In teenstelling met die meeste ander ligas, is doele egter nie die bepalende faktor vir die oordragfooie van doelskieters nie wat na of binne die 2de liga van die Bundesliga oorgedra word. Dit impliseer dat ander eienskappe en/of KPI's op die spel kan wees wat die oordragfooie van doelskieters aansienlik meer beïnvloed as die geselekteerde faktore. Toekomstige navorsing is getaak om hierdie verskille te identifiseer.
Description
Thesis (M Sport Sc)--Stellenbosch University, 2023.
Keywords
Citation