In sports, the audience attending the stadium is key. On the one hand, it is an indicator of the club’s reputation and it is key to getting certain economic benefits. On the other hand, the audience, their chants, and shouts, improve the stadium atmosphere and contribute to the performance of the team playing at home. In a study carried out by Fabian Wunderlich from the German Sport University Cologne about the games played without an audience due to the pandemic, it was found that without fans, the teams that played at home won 43% of the time compared to the 45% they usually did before. The same happened in the case of victories when playing away. With empty stands, the away wins were 32% in comparison to the previous 28%. 

When summoning thousands of people to each event, it is vital to enhance the experience of the fan who goes to the stadium. In order for it to be fully satisfactory, one of the pillars is that it should be comfortable, as shown in the 2018 Fan Engagement survey of Deloitte in which most of the fans stated that they expected the stadium to be safe, comfortable, and clean. There are key actions to improve the spectator’s experience such as making the access easier, eliminating queues and waiting time, removing bottlenecks and agglomerations, offering services and products that suit their needs, and guaranteeing their security. There might even be exceptional cases such as a pandemic or construction work in the stadium that make it more difficult. That is why the organization of big events today has to do with the use of predictive support tools to be able to optimize the performance of the premises in any context that might be given on the day of the event. 

The perceived quality of play is also important. The stadium occupation rate influences this sensation directly. It is a key aspect of broadcasts and sports marketing. 

Traditionally, two variables have been considered when predicting attendance to stadiums: the opponent’s reputation and the competition level. Currently, this prediction is much more sophisticated. Data collection starts with the characteristics of the game. There is a lot of scientific literature about what makes fans go to the stadium. Whether it is the value of the opponent’s marking, the unpredictable result of the mathematical possibility of winning a title or being qualified for other competitions. However, all these data can be evaluated depending on the idiosyncrasies of each institution. 

In the same way, the characteristics of the demand must be analyzed based on diverse aspects such as the morphology of each city. For example, it is all different if the stadium is located in the outskirts of the downtown the city. One of the biggest problems in American cities where there are some baseball or American football franchises is the indirect expenses the stadium activity generates to the city not only when thinking about the infrastructure maintenance but also about the traffic jams generated to reach places that are normally located in the outskirts of the parking problems that might arise. This problem occurs not only in sports events but also in other types of events such as concerts held in the stadiums. 

The Camp Nou, for instance, is located in one of the downtown neighborhoods of Barcelona. This is not the standard in Europe and the rest of the world for stadiums of such magnitude. In order to develop predictive tools for the management of the stadium, the club is embarking on the IoTwins project together with BSC (Barcelona Supercomputing Centre) to carry out a pilot of a Camp Nou digital twin that analyses the movement of people in and around the stadium premises. In this context, one of the collaboration results has been the investigation When are they coming? Understanding and forecasting the timeline of arrivals at FC Barcelona stadium on match days by Feliu Serra Burriel, Fernando Cucchietti, Pedro Delicado, Eduardo Graells Garrido, Alex Gil, and Imanol Eguskiza studies the stadium attendees’ behavior based on variables such as their fans’ profiles (member, tourist, occasional buyer), the weather, and other determinants. This work was published in the Journal of Sports Economics and it was introduced at the 2022 Sloan Sports Conference. 

In this investigation, they tried to identify movement patterns through a mapping of the fan’s behavior.  In order to improve the mobility and security inside the stadium, the most crowded areas or the possibly inefficient or underused accesses can be identified. Its immediate use would be, for instance, being able to identify those accesses in which there are shorter queues in order to reduce the time of entry and the agglomeration of people. 

At FC Barcelona, a great number of seats are occupied by members who pay an annual fee. However, the club never knows how many of them will attend the match until the same game day. There is a system that refunds the ticket price to the member who tells in advance they are not attending the game so the ticket can be sold to someone else. For the stadium management, the club needs to have an idea about the number of attendees as well as the number of member absences on the day of the game. 

To carry out the study, data from the last four seasons of FC Barcelona have been taken into account. Even though there has been more data available in recent years, the characteristics of the opponents and other situations and scenarios affecting stadium attendance have changed. Data come from a total of 108 games played between the 2016-17 and 2019-20 seasons. However, some games were not taken into account since they had no audience due to the pandemic. The study identified, preserving people’s privacy, every person who entered the stadium and registered the times since they got their tickets in a place with a holding capacity of 99,354 attendees, more than 100 doors, 6 flood, and 292 entrances.
Considering the data provided by this investigation, it is possible to predict a stadium arrival schedule 72 hours in advance. 

The results showed that in the games not included in the season tickets, such as the Gamper trophy or Super Cup games, people tend to arrive much earlier. Most of the fans who bought tickets arrived much earlier and there wasn’t the typical agglomeration of people some minutes before the start of the game.  There were also differences found between club members and spectators who buy the ticket. The latter arrives at the stadium between 30 to 44 minutes in advance. However, the former arrive, in average, about 20 minutes before the start. 

In order to process the data provided by the investigation, two covariates were used:  factors that are timeless, known beforehand, such as the characteristics of the game, the day of the week and the time of the game, whether it is a direct rival or a local derby; and the ones that change in time such as the number of seats available, or the tickets sold the day before the game. Besides, it was also observed that meteorology had an influence too. If it rains, attendance gets affected negatively and there are about 8,500 fewer spectators. 

These results open doors to the creation of a tool to assign the arrival time to spectators. It would be the best way to avoid agglomerations or unnecessary proximity between people and to be able to set the conditions to optimize the economic performance of a stadium.  By knowing the fans’ preferences, it is possible to segment the audience based on their entry time. 

In this way, clubs would also be able to maximize their revenues. Those spectators who buy tickets and arrive earlier at the stadium are more likely to spend more money on merchandising and have the chance of getting the products that suit their needs. Club members, on the other hand, could enter the stadium in a more direct way. 

The tool would also be useful to manage situations in which the capacity of the venue is limited such as the ones experienced during the COVID pandemic or the construction works of the stadium.

All those findings can be used in other places in the city such as airports or train stations where different commercial services are offered and in which there are agglomerations during the rush hours of arrivals and departures. In conclusion, it would be useful to streamline and facilitate urban mobility at their busiest and more uncomfortable spots as well as to maximize revenues. 

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 857191