Data Envelopment Analysis Journal > Vol 5 > Issue 2

Assessment of the Efficiency of Spanish Football Teams Through Profiling

Manuel Espitia-Escuer, Facultad de Economía y Empresa, Universidad de Zaragoza, Spain, Lucía Isabel García-Cebrián, Facultad de Economía y Empresa, Universidad de Zaragoza, Spain,
Suggested Citation
Manuel Espitia-Escuer and Lucía Isabel García-Cebrián (2021), "Assessment of the Efficiency of Spanish Football Teams Through Profiling", Data Envelopment Analysis Journal: Vol. 5: No. 2, pp 487-534.

Publication Date: 17 Aug 2021
© 2021 M. Espitia-Escuer and L. I. García-Cebrián
Performance measurement
Data envelopment analysisstochastic DEAprofilingmanagement sciencefootball teams


Download article
In this article:
1 Introduction 
2 Representative Variables of the Production Function in Football Teams 
3 DEA Method 
4 Results 
5 Conclusions 


The aim of this paper is to assess the efficiency of Spanish football teams that participated in the Spanish First Division between 2011 and 2016. We started by specifying the production function of football teams using the production process as a basis. Considering all the moves that can be made during a match, ordering them in the logical sequence that usually links them together and considering ball possession and non-possession as different phases lead to disaggregating the match into eight subdivisions whose efficiency is calculated using the data envelopment analysis (DEA) variant known as profiling. The representative input and output variables considered in these eight subdivisions are moves made during the matches. However, the actions football teams perform, irrespective of their type, are not the result of a standardised procedure. This has two consequences on the number of moves in the field of play: firstly, a minimal variation in playing conditions (both the team's and its opponent's) can alter the number; and, secondly, it is very difficult to control and arrive at a figure possibly established in advance. Since these circumstances can be interpreted as data imprecision, one of the stochastic DEA proposals has also been used in this paper as a calculation tool to verify the robustness of the results.

The results show the subdivisions in which the use of moves can be improved to increase the number of actions in the next stage. This knowledge could provide guidance for technical personnel for their training sessions.