Quantifying the performance of players in a football match

  IJETT-book-cover  International Journal of Engineering Trends and Technology (IJETT)          
© 2017 by IJETT Journal
Volume-45 Number-1
Year of Publication : 2017
Authors : Abhinav Yashkar
DOI :  10.14445/22315381/IJETT-V45P210


Abhinav Yashkar " Quantifying the performance of players in a football match ", International Journal of Engineering Trends and Technology (IJETT), V45(1),43-46 March 2017. ISSN:2231-5381. www.ijettjournal.org. published by seventh sense research group

It`s hard to evaluate the impact of a player’s performance on the team. The tradition method and rating systems involve looking at a few metrics which include goal scored, assists, key passes, tackles, intercept, etc. Often this methodology makes the goal scorer and the assist provider the most important players of the team, which might not always be the case. No wonder Ballon d`Or winners are forwards and not defenders. These numbers make sense when comparing similar metrics. However, when comparing a forward, whose primary job is to score and assist goals, with a defender, whose primary task is to clear the ball, tackle, it’s difficult. Football is a team sport; there is a complex interaction between the players. A winning goal might be a result of a threading pass by the midfielder and the winger making a diagonal run to take one of the central defenders with him creating space for the striker to score the goal. In such complicated scenario, it is tough to allocate the contribution of the goal. The paper discussed a statistical method, using regression and optimization, to qualitatively allocate the points contributed to the team by a particular player during a season. Thus, even thou the player who scored the winning goal might have secured 3 points for his team; his contribution to the team is not 3 points. The paper provides a methodology for distributing those 3 points to their rightful contributors.


[1] R. D. Baker and I. G. McHale, “Time Varying rating in association football,” Journal of the Royal Statistical Society, pp. 481–492, 2015.
[2] I.G. McHale, P.A. Scarf and D.E. Folker, “On the development of a soccer player performance rating system for the English Premier League” Interfaces, vol. 42, No. 4, pp. 339-351, July 2012.
[3] Theguardian article on Opta. [Online]. Available: https://www.theguardian.com/football/blog/2011/aug/10/opta-stats-premier-league
[4] J. Albert and J. Bennett, Curve Ball, New York: Springer, 2003.
[5] D. Hiscock, B. Dawson, J. Heasman, and P. Peeling, “Game movement and players performance in the Australian Football League,” International Journal of Performance Analysis in Sport, vol. 12, pp. 531–545, 2012.
[6] Whoscored homepage. [Online]. Available: https://www.whoscored.com/
[7] C. Anderson and D. Sally, The number game: Why everything you know about football is wrong, Penguin, 2013.
[8] I. McHale and P. Scarf, “Modelling soccer matches using bivariate discrete distribution with general dependence structure” working paper series, Salford Business School, Manchester, United Kingdom, Feb. 2007.
[9] English Premier League homepage. [Online]. Available: https://www.premierleague.com/
[10] Spanish LaLiga homepage. [Online]. Available: http://www.laliga.es/en
[11] Squawka homepage. [Online]. Available: http://www.squawka.com/home/

Football, Player Performance, Linear Programming Regression, Optimization, player rating