Describing and optimising travel distances in amateur football and handball in Germany

Explorative nature of the study

Numerous adverse environmental effects of playing football and handball could be examined, such as the energy expended on conditioning gymnasiums (handball), the water required to maintain football fields (football), transportation to team practices, and equipment and attire utilised. This study is an exploratory investigation whose scope is limited to travelling to regular league season games.

The study investigated the following questions:

Data collection and analysis

Vehicle travel distances were calculated for each league level (see Figure 1). If a league comprised multiple groups, two groups were randomly selected for analysis. Information on the leagues, groups, and regular season schedules for the 2022/2023 season was manually extracted from the official websites of the sports federations (bfv.de [Bayerischer Fußball-Verband] and bhv-online.de [Bayerischer Handballverband]). The locations of the sport facilities where the teams hold their home games were converted into geographical coordinates (latitude and longitude) using the Google Maps API https://mapsplatform.google.com.²⁶ Car distances and duration were computed using the Geoapify API https://www.geoapify.com.²⁷ The coordinates were subsequently displayed on a map using Mapbox GL Javascript https://docs.mapbox.com/mapbox-gl-js/api/²⁸ and manually checked to spot any major errors. The code used for the calculation is available at https://github.com/Tobs40/league-optimization.

Four men’s football groups from the Erlangen A League and two women’s handball groups from the Landesliga were selected to carry out a study on optimized travel distances. Initially, the travel distances were calculated in accordance with the current team distribution within each group. Subsequently, an optimal solution aimed at minimizing travel distances was calculated by interpreting the problem as the max-k-cut problem (with size constraints), modelling it as a mixed-integer linear program.²⁹ The model was then solved using a commercial MILP solver Gurobi Optimizer³⁰ (the same calculations also can be done in open source solver Scip³¹) Gurobi Optimizer is a prescriptive analytics platform and decision technology that uses mathematical optimisation to find solutions to complex problems. It allowed the rapid calculation of travel distances for all possible combinations of teams and, using a variety of optimisation algorithms, suggested the most efficient solution for reducing travel distances. Original and optimized distances were compared to identify differences.

Description of team travel in the 2022/23 season

The following section describes the travel distances for the different amateur football and handball leagues. In total, 8 women’s football and 10 handball groups and 12 men’s football and 10 handball groups were included in the analysis, with a total of 159 women’s teams and 180 men’s teams (Table 1). In total, travel distances for a full season of league play for 339 teams was calculated.

The average distance travelled by each team during the 2022/23 season is shown in Figure 2 for the different leagues. Participation in upper-level amateur leagues resulted in higher travel distances in both sports and in both women’s and men’s leagues. The distance from home to away games covered by all teams combined was 474.231 km during the 2022/23 season.

Figure 2. Average total travel distances for the 2022/23 season of play in different amateur football and handball leagues (in kilometres).

Original and optimized spatial distribution of groups

This section maps the spatial distribution of one football (men) and one handball (women) league for the 2022/23 season. In addition, an optimisation of the spatial distribution is presented.

Figure 3 displays the geographical distribution of the four groups of men’s football teams in the A-Klasse league, located in the Erlangen region. The total distance travelled by all the teams for games in the 2022/23 season schedule was 18,957 km. However, if teams were grouped in a way that minimised travel distances, the combined travel would have been 15,227 km, thereby reducing travel by 3,730 km (19.7%).

Figure 3. Original (left) and optimized (right) spatial distribution of four Men’s football groups of A-Klasse Erlangen (Groups 1-4).

Figure 4 displays the spatial distribution of two groups of women’s handball teams playing in the Landesliga. During the 2022/23 season, all teams collectively travelled 46,719 kilometres to attend regular season games. An optimized solution, aimed at minimising travel distances, includes the addition of one team to the southern group. Consequently, this group plays additional games (156 instead of 132). Nevertheless, the total distance travelled would be 42,762 km, a reduction of 3,957 km (8.5%).

Figure 4. Original (left) and optimized (right) spatial distribution of two women’s handball groups of Landesliga (Groups 1 and 2).

Summary of main findings

The study on travel distances to games in amateur women’s and men’s football and handball in Germany produced significant results. It revealed that in football, as the league’s level of play increases, teams travel farther to away matches. In handball, although the pattern was not as apparent, teams in the top two leagues still travelled the furthest distances to away games when compared to lower leagues. Throughout a single season, football teams at various levels of play for both women and men had an average travelling distance of 277 to 3,131 km for away games. Handball teams had to travel 111 to 2,314 km on average. The spatial mapping exploration revealed that not all teams were grouped optimally, which may lead to unnecessarily long travel distances for teams.

To our knowledge, this study represents one of the initial assessments of travel distances to amateur sport away games. The overall travel distance observed for certain teams, for instance, 2,958 km per season for Men’s Football Landesliga Bayern Südwest, brings into question the sustainability of amateur league football. The carbon footprint for travelling 2,958 km with a single combustion engine car is 502.86 kg/CO₂ (the equivalent for an electric vehicle is 139 kg/CO₂) using a standard emission conversion factor.³² Considering that many amateur teams may not possess mini-buses and will resort to using three or more private vehicles. For all 339 teams in the analyse, the 474.231 km travelled in the 2022/23 season would result in 80.62 t/CO₂ per petrol car used. These figures exclude all other travel that players might do to join practices, home games, or meeting other players to carpool. Transport emissions are important to consider, as they contribute globally to 16.2% of total greenhouse gas emissions (2016). This ratio rises to 25.9% in Europe,³³ and greenhouse gas emissions caused by transport have increased in Europe by 33% between 1990 and 2019.³⁴

Ethical issues of reducing the carbon footprint of amateur league play

These findings raise important ethical concerns regarding the permissible amount of carbon emissions from amateur league play in the future. Although amateur sport and exercise are strongly endorsed by public health experts for their positive impact on individual health,⁶ the potential negative effects of such activities on the environment necessitate a more nuanced assessment. Some forms of recreational sports and exercises result in significant greenhouse gas emissions, aligning individual health benefits with environmental harm. The question on how to equate the two is largely unanswered. As the sports industry also requires a solution in the future, addressing this question is of great importance. Shue³⁵ outlines another important ethical issue: should amateur football and handball players be limited to reduce carbon emissions, when professional players produce far greater emissions? For example, a male premier league football team emits 56,700 kg of CO₂ per season solely from travel,³⁶ this is about 100 to 150 times more than the average team we investigated. Should the Premier League teams not be the first to cut their emissions? However, this argument of climate justice could be countered by the argument that there are numerous recreational football teams and that their total emissions might ultimately be comparable to the ones of professional teams.

Implications of findings – optimising the spatial distribution of groups or changing game formats

In addition to raising challenging ethical questions, the findings suggest straightforward ways to reduce emissions. It appears from Figures 3 and 4 that certain teams are placed in groups that appear to result in unnecessarily long journeys. Adjusting the team placement could significantly reduce their travel and subsequently minimize the carbon footprint produced by amateur games. For example, restructuring the Men’s A-Klasse Football divisions (see Figure 3) could reduce travel by 3,730 km - almost 20% - resulting in shorter travel times and savings on petrol.

Obviously, in addition to such group re-organisation, there are alternative ways to mitigate travel, such as rethinking the game format. All teams assessed had employed the double round-robin structure, whereby each match day consists of a single game. A reduction in travel distances could be achieved by implementing a tournament-style season, modifying the round-robin arrangement, or reducing the number of games per season. Future projects should collaborate with sports federations to evaluate their willingness to adjust game formats for the purpose of minimising travel distances in amateur and possibly professional sports.

Limitations of our study

The study has certain limitation that should be considered when interpreting the results of the study. The findings of the study may not be representative of the entire league since only some groups per league were analysed. The calculation methods employed in the study were based on the available data of league play from the German Football and Handball Federations. However, potential changes in game schedules and cancellations must also be taken into account. Additionally, teams’ mode of transportation to away games was not surveyed. There is a possibility that some teams prefer using public transport that would result in lower greenhouse gas emissions during travel. Additional travel to team practices and of players travelling to home games or meeting points for away games was not considered Data from Google Maps on travel distances are time-sensitive, therefore the displayed results only present general estimates. Actual travel distances may vary depending on the day and time of travel.

Ethics and consent

All data about leagues, groups, team names, addresses, and tournament schedules sourced from publicly available open sources. As these types of data are publicly accessible and do not involve personal or sensitive information, ethical approval and consent procedures were not required for their collection.