We compared several statistics to uncover the different appointment policies employed in different countries. In this section, we discuss these comparisons in detail.
Policy comparisons
Number of different referees
Most leagues use a different number of referees each season. This may have different reasons. For example, using a high number of referees has been suggested as a measure against corruption (Duvinage, 2012). This may have been the thought of the officials in Serie A which has been the subject of a recent judiciary investigation (‘Calciopoli’ in 2006; see, for example, Boeri and Severgnini (2011) for an account of Calciopoli) who are using significantly more referees than the other leagues in the Big Five. For example, Serie A used 45 different referees in the 2020-21 and 2021-22 seasons (Table 1). On the other hand, La Liga officials are consistently using 20 referees in all of the seasons (Table 1). Since there are 10 games in each round of La Liga this corresponds to a factor of 2 for the number of referees/games per round whereas other leagues operate with a slightly higher factor (shown in parentheses in Table 1). In general, a factor of 2 to 3 seems to be the choice of the officials in the highest league divisions to also account for the poor performance of referees and other events that may require resting referees now and then. EPL is the only league where this factor occasionally falls below 2 but not by much. Although not given here, the second-highest divisions generally use more referees. This is needed since the referees officiating in the highest divisions are selected from the pool of referees at the next level based on their performance. For example, Select 1 referees in the Premier League are promoted from Select 2 referees in the Championship. The Super League of Greece stands out in its usage of referees from other countries. In the 2017-18 season, they have solely used Greek referees. However, starting from the 2018-19 season, the Super League has been using referees from other countries in addition to Greek referees. For example, in the 2021-22 season, 13% of the games (23 out of 182 games) were called by non-Greek referees. In most cases, each invited referee called for one game, however, the number of such referees was quite high. Overall, referees from 24 different countries were utilized with 9 (2018-19), 10 (2019-20), 14 (2020-21), and 14 (2021-22) different countries in the respective seasons. This strategy could be a measure against corruption but also provides a good experience for the Greek teams in case they participate in international tournaments. France and Portugal also had an exchange agreement for the 2021-22 season where a few referees only handled one game for the other league.
Number of games officiated
The utilisation of the eligible referees in the considered leagues differs. It may be desirable to use the referees uniformly by giving them about the same number of games to officiate. However, certain referees may need to be rested based on their weekly evaluations and appointments in European championships. The number of games assigned to each referee in a league highly depends also on the number of referees used by the league. If that number is high, it is natural that each referee will receive fewer games. Table 2 shows the median number of games officiated together with the standard deviations of each league in each season. La Liga stands out with its low standard deviation in line with a uniform appointment policy. For example, all referees officiated between 16 and 22 games (moreover, only one referee had 16 games, and only one referee had 22 games) in the 2021-22 season. The Bundesliga of Germany and the Ligue 1 also prefer a relatively more uniform distribution. However, Ligue 1 has exchange agreements with different countries, and their standard deviations in some seasons would be higher if these appointments were included. In the 2021-22 season, four Portuguese referees called only one game each in Ligue 1. If one includes those referees, the standard deviation for 2021-22 is increased to 6.44 from 1.90. Similarly, the standard deviation goes up to 6.45 from 4.08 when the two Swiss referees with one game each are included in the 2017-18 season. Moreover, Ligue 1 is also integrating female referees into its system. Stéphanie Frappart, who also was chosen by FIFA to participate in the 2022 World Cup as one of the three woman central referees (FIFA, 2022), started with only two assigned games in the 2018-19 season and had a gradually increased number of games in the later seasons. On the other hand, officials in the EPL do seem to prefer to have a non-uniform number of appointments possibly distinguishing between experienced and rookie referees. Even though there are about the same number of referees available in the EPL (between 18 and 22 different referees in the five seasons) as in La Liga, the standard deviations of the appointments in the EPL are high; in fact, the EPL has the highest average standard deviation among all the reported leagues. Among the Big Five, Serie A also has high standard deviations but they are utilizing comparatively a very high number of referees; in the 2020-21 and 2021-22 seasons, most referees in Serie A ended up calling for a few numbers of games during the season reflected in the comparatively smaller medians.
Appointments to teams
Each game appointment exposes referees to a couple of teams. Long exposures to teams may lead to biases towards certain teams (Hlasny and Kolaric, 2015). Thus, it may not be desirable to assign a referee too often to the games of the same team especially when they are home games. From a fairness perspective, it may also be preferable to assign a referee more or less uniformly to teams.
Home games of teams
One of the well-studied phenomena in football is the home advantage (Goumas, 2014). That is why a referee would most likely not be assigned to a game of his hometown team (Watson, 2020). The crowd also has some impact on referees’ decisions leading to biases towards home teams (Dawson et al., 2007; Goumas, 2014). Dawson et al. (2007) find that the away teams receive more yellow and red cards due to a refereeing bias favouring the home team. The Covid-19 pandemic provided a natural experiment to verify the impact of home crowds on home advantage and referee bias (see for example Sors et al., 2021). Thus, it would not be advisable to let a referee officiate in front of the same crowd too often. In general, no league allowed this to happen more than 6 times in all of the seasons investigated. Assigning a referee to 6 home games of a team has only happened once in the Liga Portugal during the 2020-21 season when Luís Godinho officiated six home games of Famalicao. In a league with many referees such as Serie A, this can be avoided easily. But La Liga with 20 referees seems to have paid special attention to this issue in the 2017-18 season; no more than two home games of the same team have been assigned to the same referee in the whole season (Table 3). The same is true for the Greek Super League in the 2019-20 season (Table 3). A low number on this statistic can not easily be achieved without planning. Interestingly, the EPL lets its referees call for more home games of the same team on average. For example, all of the five Arsenal appointments of Michael Oliver in the 2020-21 season happened to be Arsenal home games. To see which ones of these differences among the Big Five leagues were statistically significant, we ran a one-factor ANOVA and then applied the Tukey HSD test. ANOVA was applied to the average percentage of the possible home games (which is equal to the number of teams in the league minus 1) referees officiated. Statistically significant differences among the Big Five are shown in Table 4. The EPL is showing statistically significant differences with all the other Big Five leagues except with the Bundesliga in this regard.
Consecutive appointments to teams
Table 5 shows how many times a referee has been assigned to a game of the same team in his consecutive appointments. Especially, when a referee has drawn heavy criticism due to bad performance, consecutive appointments to the same team may not be preferred. Note that, these consecutive appointments of the same referee can be several weeks apart. Relatively low numbers of such appointments in La Liga indicate that the officials are paying attention to this. Doing this only once or twice in the whole season is a strong sign of deliberate action. The Serbian Super Liga, on the other hand, made quite a significant proportion of its appointments in the 2020-21 season this way. They had 20 teams and 370 games after the first week of that season. Hence, 54/370*100 = 15% of the time the same referee was assigned to a game of the same team consecutively. A one-factor ANOVA together with a Tukey HSD applied to these percentages for the Big Five confirms that La Liga is shying away from having consecutive appointments indeed (Table 6). The German Bundesliga has relatively higher percentages in this aspect and thus distinguishes itself not only from La Liga but also from Ligue 1 (Table 6).
As a note, while officials may assign the same referee to games of the same team in the referee’s consecutive appointments, this practice is very rare in consecutive rounds – almost non-existent in all the leagues.
Same opponents
In the double-round-robin format, the same opponents face each other twice. Different league officials seem to have different opinions about assigning the same referee to both of these games. In general, though, considering the ratios to the overall number of games, this practice is more or less shied away from by all of the leagues (Table 7). Looking at their consistently low number of such appointments the French do not want this to happen. On the other hand, the EPL lets this happen most frequently among the Big Five. The Spanish also do not mind making such appointments more often although they seem to have a change of heart in the 2021-22 season with significantly lower appointments compared to previous seasons. It can be argued that assigning the same referee to both of these games can be beneficial in terms of fairness since both games would be officiated by the same person. This might be what the Turkish had in their mind when they let slightly more than 20% of the rivalries be officiated by the same referee in the 2018-19 season (Table 7). However, if a league turns this into a regular policy, all parties would automatically know the referee for the second game well in advance. This type of certainty is probably not something the leagues want to have since it is a common practice to announce referee appointments just shortly before the games primarily as a measure against corruption.
Public opinion
Refereeing decisions in games are scrutinized in the media, and this scrutiny may be a source of pressure for referees. This is expected but the level of this pressure is debated; Webb (2018) finds that this pressure is not as great as previously thought. It is also not uncommon that the referee appointment decisions are publicly criticised by sports actors; people may even start looking for bad intentions especially when the referee makes some bad calls against one or the other team. One of the first things that may come to mind is corruption since selecting specific referees for certain games has been one common way of fixing matches in some known cases (Manoli et al., 2019; Manoli et al., 2021). In Portugal, there were public statements from the president of the Portuguese Football League criticising that the Council of Refereeing is not transparent supporting suspicions about institutional involvement in appointing certain referees to certain games as reported in Moriconi (2018). Moriconi and De Cima (2021) report the details of the so-called ‘Golden Whistle Affair’ in the Liga Portugal discovered in 2004. In general, though, it is more plausible that match-fixing or spot-fixing with referee involvement in a particular game occurs, if at all, after appointments are made and not when the responsible committee takes these appointment decisions.
How effective are these critical public opinions on the appointment process? An extreme example occurred in the 2021-22 season of the Süper Lig, the highest football division in Turkey. Cüneyt Çakır was the world’s second-best man referee of the decade between 2011-20 according to the International Federation of Football History and Statistics (IFFHS, 2021). On March 8, 2022 (just before the 29th round of games with 10 more rounds to go), the central committee of ex-referees (MHK) under the Turkish Football Federation which takes all referee-related decisions in Turkish football suddenly decided to not assign any games to 13 central referees anymore. In the following weeks, there were only 14 eligible referees available for calling the games in the Süper Lig; hardly enough when referees have to be rested for some reasons. Around that time, the EPL was making use of 22 referees (Bundesliga 23, La Liga 20, Ligue 1 25, and Serie A 44). One of the forced-out referees was Cüneyt Çakır who also had a good chance of going to the World Cup in 2022 (not possible, though, if he did not referee any more games). Interestingly, the decision was overturned only after a few weeks on March 26 by a higher committee to which the referees had objected. Soon afterwards, on April 4, the president of the Turkish Football Federation resigned without waiting until the end of the season; he had mentioned that the original decision was backed by the whole federation. On April 17, the president of the MHK also resigned having lasted only 173 days. He was appointed there on October 19, 2021, after the previous head had resigned at the beginning of the season. By the end of the season, Çakır only called two more games after his dismissal (he had called 17 games before his dismissal) and was not chosen to officiate in the World Cup 2022. Before the 2022-23 season started, Çakır quit officiating in the Turkish Süper Lig and mentioned that he will be educating new referees. The public was never provided with an explanation of these developments.
After the Chelsea-Tottenham game that ended 2-2 in the 2022-23 EPL season, Chelsea fans’ petition against Anthony Taylor for him to never officiate a Chelsea game in the future quickly reached 140,000 signatures making it eligible to be discussed in the Parliament (Wells, 2022). A similar incident was encountered in the 2015-16 season (Downes, 2015) against Mike Dean who, by coincidence, was the VAR official in the 2022-23 Chelsea-Tottenham game. This time, Arsenal fans were not happy with the decisions of Mike Dean when they met Chelsea at the beginning of the season in an away game and lost 2-0. That petition had also surpassed 100,000 signatures. It may not have been due to the petition but we make the following observation: Mike Dean waited for an average of 8.5 days between games, however, his next Arsenal game was assigned to him after 218 days and he -who officiated the most number of games among all referees in the EPL in that season- called only two games of Arsenal (both were away games) in the 2015-16 season.
It is plausible that the public and important stakeholders’ such as club owners’, managers’, and politicians’ opinions influence the decision-makers when assigning referees to the games. We looked at the appointment counts of the common referees to the common teams in the five seasons from 2017 to 2022 to detect any anomalies in the appointments and tagged outstanding numbers. Then, we searched for the explanations of these occurrences we tagged. Tables 8 through 13 give the total number of times each referee was assigned to each team in five seasons. The listed referees and teams were present in each one of those seasons; teams and referees absent in any of those seasons were not included. The tables also show the row and column totals but note that the row totals are not necessarily the total number of games a referee has officiated involving the listed teams as a game among a pair of the listed teams is counted twice for the referee of that game. Nevertheless, these totals help us in comparing relatively how often a referee has been chosen (row totals), or the number of games a team has been assigned to this set of (potentially the better-performing referees since they were involved in all of the seasons) referees (column totals). We did this exercise for all of the leagues for which we obtained data. Whereas we include these tables for all leagues in the Big Five, for sake of brevity, we only report cases where we thought the numbers were disproportionate, i.e. either relatively low or high, for the rest of the leagues. One common policy in all of the leagues seems to not assign a referee to the games of his hometown (birthplace) or region (home state) he is affiliated with as a referee. Exceptions were made in some cases, but the officials try to avoid these situations if they can. In the EPL, for example, referees have to fill in a form ahead of the season including the team they support and where they live to avoid situations that can lead to bias or possible speculations (Burton, 2022; Okey, n.d.). Association football has many rivalries specific to each country such as El Clásico between Barcelona and Real Madrid in Spain (Wikipedia, n.d.). While some of these are long-standing rivalries, the list can change, for example, when a new successful team appears in the respective league. We also checked how many games each referee has officiated for each one of these teams involved in the important rivalries of the countries and tagged cases when we thought there was some imbalance (Tables 8-13). In the tables, we use a double strikethrough when the small number of appointments was due to where the referee is from. In most of these cases, no appointment was made but sometimes the referee has been assigned a few times. The important team rivalries are indicated by bolding or italicizing the team names in the tables. For example, Liverpool, Manchester City, and Manchester United are rivals for the EPL championship. There is also a rivalry between Arsenal, Chelsea, and Tottenham. The list of derbies in England is quite long such as the Merseyside derby involving Everton and Liverpool. We chose some of these rivalries when looking for numbers that are not in balance among the teams. Such numbers are underlined in the tables. Other disproportionate numbers have been highlighted in grey. Where possible, we provide (potential) explanations for the tagged table entries. These explained entries are bolded in the tables.
Table 8 provides the appointment counts for the EPL. Michael Oliver called for relatively fewer games of Manchester United compared to its rivals Manchester City and Liverpool. In this case, the fans’ opinions may have mattered (Coleman, 2019; Jones, 2018). Jonathan Moss who retired at the end of the 2021-22 season (Feringa, 2022) officiated a disproportionate number (25, his next highest number for officiating for a team is 18) of Manchester United games during the studied five seasons (Table 8). However, this seems to have occurred despite fans’ dislike of him (Ward, 2020). Mike Dean -also retired at the end of the 2021-22 season (Feringa, 2022)- had the second-highest number of Manchester United games officiated (22) despite his decisions having been criticised by the legendary Alex Ferguson (Smith, 2022). Of course, the more a referee officiates for a team, the more likely it is for the referee to make bad decisions and there will be more criticism yet this does not seem to have affected the committee decisions in these cases. Since Simon Hooper served as the main referee in much fewer games compared to other referees in the EPL, we did not tag his zero appointments to several teams (none of which are his hometown teams).
In La Liga’s case, all zeroes (Table 9) are due to not letting a referee call for a team in his hometown (country). For example, since Antonio Mateu Lahoz is from Valencia he was not assigned to any games of Levante, Valencia, and Villarreal.
While IFFHS ranks referees from different countries, Spain has a unique event worth mentioning that is related to referee performance nationally: Spanish sports newspaper Marca awards the so-called Guruceta Trophy to the best referee each season since 1987. The trophy is named after the Spanish referee Emilio Guruceta (1941-1987) who died in a traffic accident (Back Page Football, 2017). Interestingly, a former president of Anderlecht (Constant van den Stock) admitted that he has given a bribe of 18,000 pounds to Guruceta during their 1984 UEFA cup game against Nottingham Forest in the second leg of the semi-finals during a Belgian civil court hearing in 1997 after Guruceta’s death (Taylor, 2013). The match was won by Anderlecht 3-0. Anderlecht was later banned from competing in UEFA cups for a year although the decision was overturned on appeal. Guruceta was also not liked by Barcelona fans who accused him of fixing a match against Real Madrid in 1970 (Back Page Football, 2017). Events such as Guruceta Trophy can help with increasing the quality of decisions by providing additional motivation for the referees to perform better.
The number of appointments of referees to teams in Serie A (Table 10) and the German Bundesliga (Table 11) has been coded as described above. We could not find any explanation for the greyed numbers in both leagues, they may have just happened as a result of the nature of things. Yet, some of those numbers look quite disproportionate; for example, M. Di Bello was assigned to 20 games of Roma when the next highest number of appointments to Roma games by any other referee is 13. In the German Bundesliga, T. Stieler was assigned to only one E. Frankfurt game since he was born near Frankfurt.
In France, Ligue 1 officials are also paying attention to not assigning a referee to his hometown (Table 12). However, they are not as strict as in La Liga. For example, Jérôme Brisard who is associated with Maine has been assigned to four Angers games. As another seemingly cautionary decision, Benoît Bastien, a referee associated with the Lorraine region, has only been assigned to two games of Strasbourg which is nearby. Mikael Lesage’s and Ruddy Buquet’s number of appointments to Angers games stand out as they have been assigned to only a handful of Angers games when compared to their appointments to other teams (Table 12). However, many Angers games were not called by the listed referees, and Angers games may not have been deemed critical enough to use arguably these better referees common to all of the seasons. Thus, it appears that the imbalance in the case of Lesange and Buquet is just coincidental.
Table 13 is a unified table that includes team appointments of selected referees from leagues other than the Big Five between 2017-22. In Portugal, many referees have a disproportionate number of appointments to certain teams indicated with grey entries. In some referees’ cases in Portugal, the distribution among rival teams (underlined entries) is also uneven. The same can be said for some referees in Greece and the only Serbian referee on the table. As a note, there were only four Serbian referees who were common in all of the investigated five seasons which was the lowest number of common referees in the compared leagues. No Dutch entry has been tagged but we listed some referees in Table 13 because it turns out that only those few referees among the total of 15 referees common to all of the seasons officiated most of the games for the listed teams. In the case of Turkish Süper Lig the most successful Turkish referee C. Çakır officiated only 10 games for Trabzonspor, a team in the so-called Big Four, but 18, 16, and 19 games for the other three teams between 2017-22 (Table 13). Interestingly, he does not seem to be liked much by the people of the city of Trabzon (Yeni Şafak, 2021). H. Göçek could only get back to officiating games of Galatasaray in 2021 three years after being under fire in a game of Galatasaray against Konyaspor in 2018 (Akşam, 2021). Between 2017 and 2022 he was assigned to only three games of Galatasaray (Table 13). In March 2022, the president of Fenerbahçe, a very prominent businessman, heavily criticised Z. Küçük’s appointment to their game against Trabzonspor: Citing the irregularities in the appointment process, he claimed that his club thought that the process was being ‘interfered with’ and that the original referee decision had been overturned (TRT Haber, 2022). In fact, Z. Küçük has been assigned to a disproportionately high number of Trabzonspor games in the past (Table 13). Numerato (2016) classifies the appointment of a preferable referee to a game as ex-ante meta corruption. A. Kardeşler’s case of only one appointment to Trabzonspor games can be possibly explained by his brother who played as a substitute goalkeeper for that team.