Analysis of home advantage in sports

master's thesis

Timur Kulenović (Author), Jure Demšar (Mentor)

Abstract

Home advantage is a phenomenon observed in sports where teams tend to perform better when playing in their home environment compared to playing away from home. In this work, we addressed home advantage in professional basketball and football leagues, focusing on five top-level leagues in each sport. Our goal was to quantify home advantage and compare it between the leagues. An important part of our work was the collection and careful curation of datasets, which we made publicly available for broader research applications. A novel metric was proposed to measure home advantage at the individual game level, differentiating it from conventional season-based metrics. To quantify home advantage and the uncertainty surrounding it, we utilized Bayesian hierarchical models. Our results suggest that in basketball the ABA league displayed the highest degree of home advantage, while NBA had the lowest degree of home advantage. In football, four of the leagues yielded similar results, whereas La Liga showed significantly higher levels of home advantage. Seasonal trends were also explored. For some leagues, we observed lower home advantage in the seasons affected by COVID-19. Additionally, the study investigated the connection between potential factors and home advantage, for this purpose we utilized the proposed home advantage metric. The crowd attendance is mainly positively correlated and the referee bias is often negatively correlated with home advantage but the results differ between the leagues. Nevertheless, we highlight the existence of unexplored factors that could substantially contribute to home advantage.

Keywords

home advantage;football;basketball;Bayesian statistics;data science;computer science;master's thesis;

Data

Language:	English
Year of publishing:	2023
Typology:	2.09 - Master's Thesis
Organization:	UL FRI - Faculty of Computer and Information Science
Publisher:	[T. Kulenović]
UDC:	004:796(043.2)
COBISS:	177579523
Views:	38
Downloads:	5
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	Slovenian
Secondary title:	Analiza prednosti domačega igrišča v športu
Secondary abstract:	Prednost domačega igrišča v športu je pojav v športu, pri katerem ekipe dosegajo boljše rezultate na domačem prizorišču v primerjavi z rezultati, ki jih dosegajo na gostujočem prizorišču. V delu smo naslovili prednost domačega igrišča v košarki in nogometu, pri čemer smo se osredotočili na pet profesionalnih lig v vsakem športu. Naš cilj je bil kvantificirati prednost domačega igrišča in jo primerjati med ligami. V delu smo zbrali in obdelali množico podatkov ter ustvarili podatkovno zbirko, ki je javno dostopna za širšo uporabo. Za potrebe analize smo predstavili novo metriko, ki meri prednost domačega igrišča na ravni tekme, kar jo razlikuje od obstoječih metrik, ki prednost domačega igrišča merijo na ravni skupine tekem oz. sezone. Za namen kvantifikacije prednosti domačega igrišča in oceno negotovosti smo uporabili Bayesovske hierarhične modele. Rezultati kažejo, da je med izbranimi košarkarskimi ligami prednost domačega igrišča največja v ligi ABA, najnižja pa v ligi NBA. V nogometu smo ugotovili, da z višjo prednostjo domačega igrišča izstopa španska liga, v ostalih štiri ligah pa so rezultati podobni. Raziskali smo tudi sezonske trende, kjer pri nekaterih ligah opazimo nižjo prednost domačega igrišča v času epidemije COVID-19. Preučili smo tudi povezavo med potencialnimi dejavniki ter prednostjo domačega igrišča, kjer smo uporabili novoustvarjeno metriko. Tu se rezultati nekoliko razlikujejo med ligami, v glavnem pa je vpliv gledalcev pozitivno koreliran, sodniška prisotnost pa negativno korelirana s prednostjo domačega igrišča. Kljub temu izpostavljamo obstoj dejavnikov, ki jih v raziskavo nismo vključili, vendar verjetno pomembno prispevajo k prednosti domačega igrišča.
Secondary keywords:	prednosti domačega igrišča;nogomet;košarka;podatkovne vede;magisteriji;Bayesova statistika;Šport;Računalništvo;Univerzitetna in visokošolska dela;
Type (COBISS):	Master's thesis/paper
Study programme:	1000471
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. v Ljubljani, Fak. za računalništvo in informatiko
Pages:	VIII, 65 str.
ID:	21492992