Home Advantage and Referee Bias: A Comparative Analysis of League Level Differences

Abstract

Home advantage (HA) and referee bias are two important phenomena in football (soccer) and are frequently investigated by researchers. This study focused on analyzing HA and referee bias at different league levels. The study included 632 matches played in Turkey’s top two leagues during the 2022–2023 season. Both leagues showed statistically significant game outcome differences between home and away teams regarding points, goals, shots, shots on target, passes, accurate passes, and ball possessions, indicating a HA. However, regarding referee bias, this study found that only Level 1 teams significantly differed in penalties and that this difference was not evident after analyzing for Bayes factors. Among Level 2 teams, as there was no statistical difference between home and away teams on any parameter indicating referee bias, and all Bayes factors indices were between 3.92–11.15, the null hypothesis was supported. Therefore, the results revealed a HA at both league levels, but a referee bias was not present in either league. This study emphasized that the HA phenomenon is clear and still evident in elite football and sub-elite leagues.

Keywords

referee decisions match analysis performance analysis crowd noise spectators game location

Introduction

Home advantage (HA) has been demonstrated in many sports, including football (Wolaver & Magee, 2022). One of the key important advantages of playing at home is the presence of fans (Pollard, 2008). It is widely believed that these spectators boost the performance of home team players while potentially hampering the performance of away team athletes (Işın & Gómez-Ruano, 2023). In 2019, due to the COVID-19 pandemic precautions, sports competitions were sometimes played behind closed doors around the world (Alonso et al., 2022). This led to a reduction in HA effects and highlighted the benefits of home team fan presence and crowd noise (Bilalić et al., 2021; Sors et al., 2023). In addition, team ability is also seen as a key factor that may affect HA. Ramchandani et al. (2021) conducted a study using a 24-year dataset to analyze the relationship between HA and team ability in the English Football Leagues. They found that high-ability teams had significantly more HA than medium and low-ability teams.

Another factor contributing to the HA has been home team familiarity with the home stadium. As the home team has more match experience in their own stadium than the away team, the home team tends to perform better (Wolaver & Magee, 2022). One aspect of this stadium familiarity factor is the type of playing surface (e.g. artificial turf) (Pollard, 2008). In fact, Barnett and Hilditch (1993) found that home teams using artificial turf had a greater HA than those using natural grass. In addition, travel fatigue has been found to cause a decrease in the away team’s performance compared to the home team, further supporting the HA (Pollard, 2008).

Finally, another phenomenon that has increased the HA is referee bias. This phenomenon refers to the tendency of referees to make decisions that favor the home team, as also influenced by crowd noise (Dohmen & Sauermann, 2016). Many researchers have mentioned referee bias in favor of the home team in football (Boyko et al., 2007; Goumas, 2014; Işın & Gómez-Ruano, 2023; Sors et al., 2021). Investigators have noted that referees added more extra time when the home team was losing (Lago-Peñas & Gómez-López, 2016), and that referees show more yellow cards and red cards to the away team (Sors et al., 2021), suggesting that referees consciously or unconsciously made decisions that favor the home team (Dohmen & Sauermann, 2016).

Studies have been conducted to investigate the differences in HA across competitive levels in football. Most of these studies have concentrated on professional men’s football (Pollard & Gómez, 2014a), with only a few examining HA in women’s football (Leite & Pollard, 2020; Pollard & Gómez, 2014a). Pollard et al. (2017) reported that HA was higher in men’s than in women’s competitions. In the study by Pollard and Gómez (2014b), regional differences were found to affect HA in different national football leagues. However, studies focusing on different leagues within a country (Leite & Pollard, 2018; Sánchez et al., 2009) have shown varied results regarding the impact of HA on league level.

There is a need for further understanding of HA in non-elite leagues to determine whether there is a similarity in HA and referee bias among leagues within the same country. In addition, given the specific characteristics of the leagues, the results of the Turkish Leagues may differ from those of others (Seckin & Pollard, 2008). Furthermore, more information on HA and referee bias in Turkish leagues needs to be available (Işın & Gómez-Ruano, 2023). Both to fill this gap and because of the inconsistent results at different league levels, I focused on HA and referee bias in the top two leagues in Turkey. Therefore, I investigated whether there is a similar HA and referee bias in both leagues and how any league difference might affect the match performances of home and away teams. I expected these research findings to provide an opportunity to further research there is a league difference in whether the referees are biased in favor of the home team in matches with spectators present.

Method

Match Samples and Data Collection

This present study was approved by the Akdeniz University Social and Humanities Scientific Research and Publication Ethics Committee (2023–15/373), and it was designed in accordance with the Declaration of Helsinki.

During the 2022–2023 season, I analyzed 313 games in the TFF Super League (referred to as Level 1) and 319 games in the TFF First League (referred to as Level 2). For the 2022–2023 season, there were initially supposed to be 19 teams in each league. However, due to an earthquake that hit multiple Turkish provinces in February, two teams from each league had to withdraw. I considered the matches played before withdrawal by these teams, but since the matches played after the withdrawal were forfeit matches, matches after the teams’ withdrawals were not included in data analysis.

I obtained from InStat Scout (InStat®) various data points for goals, shots, shots on target, passes, accurate passes, and ball possession to assess HA and fouls, yellow cards, red cards, offsides and penalties to assess referee bias. Previous studies have shown InStat® to have high reliability (α = .85–.93) (Kubayi & Larkin, 2020; Silva & Marcelino, 2023). To quantify home advantage, I employed Pollard’s (1986) formula: Home advantage = (H/(H + A)) * 100%, which discerns the presence of HA. A score exceeding 50% signifies HA, in accordance with the findings of Lago-Peñas et al. (2016).

Statistical Analysis

The data underwent testing with Kolmogorov–Smirnov tests (p > .05) to determine whether data distributions met assumptions of normality. I then used a two-sample t test and calculated effect sizes (ES). ES was categorized as small effect (d = .20), medium effect (d = .50) and large effect (d = .80) according to Cohen’s d (1998). I also calculated Bayes factors (BF01) to assess evidence for or against the null hypothesis. The null hypothesis is considered to be supported by anecdotal, substantial, strong, very strong and decisive evidence when the Bayes factors (BF₀₁) are 1–3, 3–10, 10–30, 30–100, >100 respectively (Jarosz & Wiley, 2014). I conducted all statistical analyses using the Jamovi statistical software (version 2.3.18.0). The data were reported as means (and standard deviations).

Results

Among Level 1 teams, the analysis revealed that home teams had a clear advantage over away teams on all parameters indicating HA (t (624) range = 2.41–5.71, p < .05; ES range = .19, small - .46, medium). This was supported by Bayes factors, meaning that the alternative hypothesis was confirmed (BF₀₁ range = .01–.67). However, regarding potential referee bias, no parameters except for penalties (t (624) = 2.25; p = .025; ES = .18, small) showed any sign of bias. While there was some evidence of bias in favor of home teams on penalties called, the BF01 was close to 1 (BF = .96), which supports both hypotheses equally. Therefore, there was no evidence (all BF₀₁ > 1) of referee bias in favor of home teams in the Level 1 league (Table 1).

Table 1.

Analyses of Home Advantage and Referee Bias According to Various Performance Indicators of Level 1 Home and away Teams in Turkey.

Variables^a	Descriptive statistics		Independent Samples t test
	Home Team n:313	Away Team n:313	Independent Samples t test
	M (SD)	M (SD)	t	p	ES	BF₀₁ (H0/H1)^b
Points	1.58 (1.33)	1.18 (1.30)	3.82	<.001	.31	.01	—
Goals	1.69 (1.36)	1.35 (1.23)	3.27	.001	.26	.06	—
Shots	13.19 (5.08)	10.95 (4.73)	5.71	<.001	.46	.00	—
Shots on target	5.41 (2.83)	4.27 (2.48)	5.39	<.001	.43	.00	—
Passes	493.27 (108.10)	471.77 (107.13)	2.50	.013	.20	.54	—
Accurate passes	411.25 (105.79)	391.06 (104.12)	2.41	.016	.19	.67	—
Ball possession	51.18 (9.50)	48.82 (9.50)	3.10	.002	.25	.11	—
Fouls	12.94 (3.75)	12.84 (3.80)	.33	.743	.03	10.65	Strong
Yellow cards	2.17 (1.46)	2.22 (1.43)	−.47	.638	−.04	10.07	Strong
Red cards	.12 (.35)	.15 (.36)	−1.13	.260	−.09	6.04	Substantial
Offsides	2.18 (1.76)	1.98 (1.63)	1.46	.145	.12	3.96	Substantial
Penalties	.26 (.51)	.18 (.41)	2.25	.025	.18	.96	—

Note: M: Mean, SD: Standard Deviation, ES: Cohen’s d effect size, BF₀₁: Bayes Factor. p-values legend: *p < .05; **p < .001, ***p < .001; ns: not significant (p > .05).

^aAll data presented on a per-match basis.

^bCalculated considering a one-sided null hypothesis.

There was also a HA among Level 2 teams on all HA parameters (t (636) range = 2.41–5.71, p < .05; ES range = .18, small - .44, medium). All HA parameters except goals were supported by Bayes factors analyses, meaning that the alternative hypothesis was confirmed (BF₀₁ range = .00–.05). In addition, there was no evidence of referee bias against the away teams based on the parameters analyzed, and, for this analysis, the null hypothesis was supported by all Bayes factors (all BF₀₁ > 1.00) (Table 2).

Table 2.

Analysis of Home Advantage and Referee Bias According to Various Performance Indicators of Level 2 Home and away Teams in Turkey.

Variables^a	Descriptive statistics		Independent Samples t test
	Home Team n:319	Away Team n:319	Independent Samples t test
	M (SD)	M (SD)	t	p	ES	BF₀₁ (H0/H1)^b
Points	1.55 (1.32)	1.20 (1.29)	3.36	<.001	.27	.05	—
Goals	1.42 (1.25)	1.22 (1.08)	2.24	.026	.18	.99	—
Shots	11.75 (4.53)	9.82 (4.30)	5.54	<.001	.44	.00	—
Shots on target	4.60 (2.50)	3.88 (2.31)	3.79	<.001	.30	.01	—
Passes	452.47 (105.79)	422.12 (105.71)	3.63	<.001	.29	.02	—
Accurate passes	370.27 (101.31)	341.06 (102.11)	3.63	<.001	.29	.02	—
Ball possession	51.66 (9.38)	48.03 (9.33)	4.89	<.001	.39	.00	—
Fouls	13.57 (4.11)	13.28 (3.73)	.95	.343	.08	7.30	Substantial
Yellow cards	2.08 (1.32)	2.24 (1.36)	−1.47	.141	−.12	3.92	Substantial
Red cards	.12 (.34)	.14 (.39)	−.76	.448	−.06	8.54	Substantial
Offsides	1.83 (1.63)	1.67 (1.51)	1.31	.190	.10	4.88	Substantial
Penalties	.22 (.43)	.21 (.45)	.18	.857	.01	11.15	Strong

Note: M: Mean, SD: Standard Deviation, ES: Cohen’s d effect size, BF₀₁: Bayes Factor. p-values legend: *p < .05; **p < .001, ***p < .001; ns: not significant (p > .05).

^aAll data presented on a per-match basis.

^bCalculated considering a one-sided null hypothesis.

Figure 1 shows the comparison of the percentage of HA between both leagues. The analysis shows that both leagues had similar advantages for home teams.

Figure 1.

Variation of HA Data Points According to Turkish Football League Levels.

Discussion

Generally, past investigators have found a HA for player performance in football, and in some cases, referees were also found to be influenced by crowd noise and made biased decisions in favor of the home team (Goumas, 2014). However, there have been inconsistent results regarding the impact of HA on league level in different leagues within a country (Leite & Pollard, 2018; Sánchez et al., 2009), and this phenomenon needs to be better understood in Turkish leagues. For these reasons, I investigated HA and referee bias at different league levels in football in Turkey, and these data revealed that, while HA exists irrespective of league level, there was no indication of referee bias at either league level. Thus, neither HA nor referee bias are dependent on league level in Turkish football.

I analyzed HA using some of the team match data and the HA score (Level 1: 57% vs. Level 2: 56%). Based on the HA score, I found that the level of competition did not affect HA in either league. However, these values were lower than the mean HA value (approximately 60%) reported in another recent HA study in European national leagues (Dellagrana et al., 2023; Lago-Peñas et al., 2016; Pollard & Gómez, 2014a). In Portuguese football, HA was affected by the level of competition. There was more evidence of HA in the amateur and semi-pro leagues (60.36% and 60.46%, respectively) than at the professional level (58.31%) (Almeida & Volossovitch, 2017). Similarly, Leite and Pollard (2018) found a greater HA at Level 2 compared to Level 1 competitions. Since these researchers investigated different leagues, their results may have been influenced by factors such as differences in player quality and spectator numbers or enthusiasm. Nevertheless, the common finding in these studies was the presence of HA at all levels of football.

Seckin and Pollard (2008) analyzed HA and referee bias in the Turkish league over 12 seasons, and they concluded that home teams scored more points and had significantly more shots on target than away teams. More recently, Işın and Gómez-Ruano (2023) revealed that playing at home is a significant advantage for home teams in the Turkish Super League, and this advantage diminished in matches played without spectators. Therefore, spectators are said to play the role of a “12th man” for home teams. Similarly, in the present study, home teams were more successful than away teams, indicating the presence of HA in both leagues, consistent with findings from Leite and Pollard (2018) and Almeida and Volossovitch (2017). Interestingly, Leite and Pollard (2018) pointed out that HA was higher for Level 2 than for Level 1 teams, and they offered explanatory theories such as the idea that Level 2 teams can create a more intimidating atmosphere in favor of the home team, despite a lower level of spectator support. Second, Level 2 teams may travel with different comfort than Level 1 teams. In Turkey, Level 1 teams are from a relatively closer region, supporting the idea that spectators, rather than travel fatigue, create an intimidating atmosphere favoring the home team. On the contrary, Level 2 teams have lower attendance numbers than Level 1 teams, diminishing the impact of the crowd. As Leite and Pollard (2018) suggested, Level 2 teams may need more travel comfort. Compared to Level 1, Level 2 teams are from many regions of Turkey, resulting in more travel time for away teams and, perhaps, more crowd enthusiasm for home teams.

Işın and Gómez-Ruano (2023) also studied referee bias in the Turkish Super League between the 2014–15 and 2021–22 seasons. They revealed that away teams tended to receive more yellow cards, fewer penalties, and were caught offside more often in matches with spectators compared to home teams. However, this trend did not occur in matches without spectators, and the referee bias was then eliminated. Interestingly, I found no evidence of referee bias in matches with spectators at either league level, which differs from previous research. The most likely reason for this could be the Video Assistant Referee (VAR) system. As the VAR system was used in both leagues, this offered the opportunity to re-evaluate the wrong decisions made by the referees. This system may have diminished possible referee bias by correcting referee’s wrong decisions, whether these decisions were or were not affected by audience pressure. Nonetheless, the VAR system also resulted in players being less aggressive (Carlos et al., 2019; Han et al., 2020), leading to fewer needs for referee interventions (Işın, 2023).

Limitations and Directions for Further Research

While this study investigated HA and referee bias in the top two leagues in Turkey, due to a major earthquake, two teams from both leagues (Level 1: Hatayspor and Gaziantep FK; Level 2: Yeni Malatyaspor, and Adanaspor) withdrew from their leagues in February. Although these teams’ forfeited games were not included in the data analysis, all the matches that should have been played in the regular season were completed, and this factor may have affected the study’s findings. In addition, devastating earthquake effects may have affected crowd behavior during this period and altered the impact of fans on HA and referee bias. Additionally, this study focused only on HA and referee bias in a single season because VAR was used in the last season in the Level 2 league, meaning that including previous seasons would have resulted in a biased approach, as VAR is known to affect the accuracy of referee decisions. Therefore, future investigators should compare HA and referee bias in the seasons before and after the introduction of VAR, and this study should be replicated in a period when there is no natural disaster that might affect crowd behavior during matches.

Conclusion

In this study, I aimed to analyze HA and referee bias in different league levels of Turkish football I found a HA in both leagues but no HA referee bias in either league. These data offer practical applications for coaches and football players in understanding HA and referee bias in Turkish football. Since playing away seems to negatively affect team performances and playing at home positively affects athletes’ play, away players and coaches should be prepared for matches by considering crowd noise, unfamiliarity with the field or travel fatigue. Data from this study confirm that the HA phenomenon is evident in both elite and sub-elite football leagues.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Institutional Review Board Statement

The study was conducted in accordance with the Helsinki Declaration and approved by the Akdeniz University Social and Humanities Scientific Research and Publication Ethics Committee (2023-15/373).

ORCID iD

Ali Işın

Author Biography

Ali Işın holds a PhD in sport sciences and is an Assistant professor at the Faculty of Sport Sciences, Akdeniz University, where he focuses on sports injuries, injury prevention, relative age effect in sport, and performance analysis.

References

Almeida

C. H.

Volossovitch

(2017). Home advantage in Portuguese football: Effects of level of competition and mid-term trends. International Journal of Performance Analysis in Sport, 17(3), 244–255. https://doi.org/10.1080/24748668.2017.1331574

Alonso

Lorenzo

Ribas

, et al. (2022). Impact of COVID-19 pandemic on HOME advantage in different European professional basketball leagues. Perceptual and Motor Skills, 129(2), 328–342. https://doi.org/10.1177/00315125211072483

Barnett

Hilditch

(1993). The effect of an artificial pitch surface on home team performance in football (soccer). Journal of the Royal Statistical Society - Series A: Statistics in Society, 156(1), 39–50. https://doi.org/10.2307/2982859

Bilalić

Gula

Vaci

(2021). Home advantage mediated (HAM) by referee bias and team performance during covid. Scientific Reports, 11(1), 21558. https://doi.org/10.1038/s41598-021-00784-8

Boyko

R. H.

Boyko

A. R.

Boyko

M. G.

(2007). Referee bias contributes to home advantage in English Premiership football. Journal of Sports Sciences, 25(11), 1185–1194. https://doi.org/10.1080/02640410601038576

Carlos

L. P.

Ezequiel

Anton

(2019). How does video assistant referee (VAR) modify the game in elite soccer? International Journal of Performance Analysis in Sport, 19(4), 646–653. https://doi.org/10.1080/24748668.2019.1646521

Cohen

(1998). Statistical power analysis for the behavioral sciences. Routledge Academic.

Dellagrana

R. A.

Nunes

R. F.

Silva

R. L.

(2023). The importance of crowd support and team quality to home advantage in Brazilian soccer league first division. Perceptual and Motor Skills, 130(3), 1255–1268. https://doi.org/10.1177/00315125231169876

Dohmen

Sauermann

(2016). Referee bias. Journal of Economic Surveys, 30(4), 679–695. https://doi.org/10.1111/joes.12106

10.

Goumas

(2014). Home advantage and referee bias in European football. European Journal of Sport Science, 14(sup1), S243–S249. https://doi.org/10.1080/17461391.2012.686062

11.

Han

Chen

Lago-Peñas

, et al. (2020). The influence of the video assistant referee on the Chinese Super League. International Journal of Sports Science and Coaching, 15(5–6), 662–668. https://doi.org/10.1177/1747954120938984

12.

Işın

(2023). Video yardımcı hakem uygulamasının hakem kararları üzerine etkileri: Tanımlayıcı araştırma. Turkiye Klinikleri Journal of Sports Sciences, 15(2), 237–242. https://doi.org/10.5336/sportsci.2022-94757

13.

Işın

Gómez-Ruano

M. Á.

(2023). How the 12th man influences football matches: The role of fans and referees in the home advantage phenomenon. Perceptual and Motor Skills, 130(5), 2177–2188. https://doi.org/10.1177/00315125231191683

14.

Jarosz

A. F.

Wiley

(2014). What are the odds? A practical guide to computing and reporting Bayes factors. The Journal of Problem Solving, 7(1), 2. https://doi.org/10.7771/1932-6246.1167

15.

Kubayi

Larkin

(2020). Technical performance of soccer teams according to match outcome at the 2019 FIFA Women’s World Cup. International Journal of Performance Analysis in Sport, 20(5), 908–916. https://doi.org/10.1080/24748668.2020.1809320

16.

Lago-Peñas

Gómez-López

(2016). The influence of referee bias on extra time in elite soccer matches. Perceptual and Motor Skills, 122(2), 666–677. https://doi.org/10.1177/0031512516633342

17.

Lago-Peñas

Gómez-Ruano

Megías-Navarro

, et al. (2016). Home advantage in football: Examining the effect of scoring first on match outcome in the five major European leagues. International Journal of Performance Analysis in Sport, 16(2), 411–421. https://doi.org/10.1080/24748668.2016.11868897

18.

Leite

Pollard

(2018). International comparison of differences in home advantage between level 1 and level 2 of domestic football leagues. German Journal of Exercise and Sport Research, 48(2), 271–277. https://doi.org/10.1007/s12662-018-0507-2

19.

Leite

Pollard

(2020). Comparison of home advantage between level 1 and level 2 in women’s football leagues. Journal of Anthropology of Sport and Physical Education, 4(4), 9–13. https://doi.org/10.26773/jaspe.201002

20.

Pollard

(1986). Home advantage in soccer: A retrospective analysis. Journal of Sports Sciences, 4(3), 237–248. https://doi.org/10.1080/02640418608732122

21.

Pollard

(2008). Home advantage in football: A current review of an unsolved puzzle. The Open Sports Sciences Journal, 1(1). https://doi.org/10.2174/1875399x00801010012

22.

Pollard

Gómez

M. A.

(2014a). Comparison of home advantage in men's and women’s football leagues in Europe. European Journal of Sport Science, 14(sup1), S77–S83. https://doi.org/10.1080/17461391.2011.651490

23.

Pollard

Gómez

M. A.

(2014b). Components of home advantage in 157 national soccer leagues worldwide. International Journal of Sport and Exercise Psychology, 12(3), 218–233. https://doi.org/10.1080/24748668.2017.1304031

24.

Pollard

Prieto

Gómez

M. Á.

(2017). Global differences in home advantage by country, sport and sex. International Journal of Performance Analysis in Sport, 17(4), 586–599. https://doi.org/10.1080/24748668.2017.1372164

25.

Ramchandani

Millar

Wilson

(2021). The relationship between team ability and home advantage in the English football league system. German Journal of Exercise and Sport Research, 51(3), 354–361. https://doi.org/10.1007/s12662-021-00721-x

26.

Sánchez

P. A.

García-Calvo

Leo

F. M.

, et al. (2009). An analysis of home advantage in the top two Spanish professional football leagues. Perceptual and Motor Skills, 108(3), 789–797. https://doi.org/10.2466/PMS.108.3.789-797

27.

Seckin

Pollard

(2008). Home advantage in Turkish professional soccer. Perceptual and Motor Skills, 107(1), 51–54. https://doi.org/10.2466/pms.107.1.51-54

28.

Silva

Marcelino

(2023). Inter-operator reliability of InStat Scout in female football games. Science and Sports, 38(1), 42–46. https://doi.org/10.1016/j.scispo.2021.07.015

29.

Sors

Grassi

Agostini

, et al. (2021). The sound of silence in association football: Home advantage and referee bias decrease in matches played without spectators. European Journal of Sport Science, 21(12), 1597–1605. https://doi.org/10.1080/17461391.2020.1845814

30.

Sors

Grassi

Agostini

, et al. (2023). The influence of spectators on home advantage and referee bias in national teams matches: Insights from UEFA nations league. International Journal of Sport and Exercise Psychology, 21(2), 290–305. https://doi.org/10.1080/1612197x.2022.2044367

31.

Wolaver

A. M.

Magee

(2022). Ghost games: Crowds, referee bias, and home advantage in European football leagues. Journal of Sport Behavior, 45(3).