Below are three articles (published in The Conversation, Scientific American, and ZME Science)
Published in The Conversation, 8.11.2023, "We blurred the gender of soccer players and had people rate their performances − with surprising results"
During the 2023 FIFA Women’s World Cup, the telecommunications company Orange ran a powerful TV ad. It depicts the graceful agility and dramatic goal-scoring shots from French national players such as Antoine Griezmann, Kylian Mbappé and Olivier Giroud.
Then comes the catch. After about a minute, the viewer sees that the highlights had been artifically modified: All of the players were actually from the French women’s national team. The ad seeks to push back on a common criticism of women’s sports: that female players aren’t as entertaining as their male counterparts, and the action is less exciting than it is in men’s sports. In our recently published study, we decided to put this notion to the test. We obscured the gender of soccer players and asked participants to rate the performances of the athletes they viewed.
Sports is one of the world’s largest markets – estimated to be around US$83 billion in the United States alone. But any fan, casual or serious, can readily observe striking gender differences in media coverage and player salaries. Outside of the Olympics, only about 4% of all sports media coverage around the world is devoted to women’s sports. Live events are much less likely to be broadcast, and only one woman, retired tennis star Serena Williams, is among the Forbes’ 50 highest-paid athletes in the world.
In 2019, the U.S. women’s national soccer team filed a gender discrimination lawsuit against U.S. Soccer. Even though the squad had won the past two women’s World Cups – and the men’s team hadn’t reached a semifinal since 1930 – the women weren’t paid as much as their male counterparts. In 2022, the two sides came to an agreement guaranteeing equal pay. But the dispute offered a window into the thinking of many sports executives and fans. U.S. Soccer’s legal counsel stated that the women’s team plays “a different game” from the men’s “in the sense that men are bigger, stronger, faster.” Research has proven that women’s and men’s soccer does indeed differ in several physical aspects – for example, male soccer players cover more ground and run faster during games.
But the question is whether the physical differences of women necessarily make the games less entertaining. The existence of stereotypes points to an alternative possibility: Gender biases might influence perceptions of the quality of the games.
Physical differences are often used to sustain sexist assertions that most women’s sports are boring and slow. This narrative – especially prominent among detractors of women’s sports and internet trolls – likely influences mainstream attitudes toward sports such as women’s basketball and women’s soccer.
When gender is obscured, differences disappearPrior research shows that biases likely play a role when people evaluate the performance of women on the field and in the workplace. It includes the work of Claudia Goldin, who was recently awarded the Nobel Memorial Prize in Economics. Her fascinating 2000 study with economist Cecilia Rouse showed how blind auditions for symphony orchestras resulted in more women being hired.
Economist Claudia Goldin was able to show that when female musicians participated in blind auditions, it improved their chances of being hired. Goldin’s work inspired our study. What if we could prevent soccer fans from identifying players and test for whether gender bias influenced evaluations of the players’ athletic performance?
We set up an experiment and showed more than 600 participants highlights from men’s and women’s professional soccer videos. We asked the group, which was made up of 55% men and 45% women, to watch 10 videos and then rate the performance of the players during the goal-scoring plays on a scale of 1 to 5. In the control group, participants watched and evaluated regular videos. For the experimental group, we blurred the gender of the players, making it impossible for participants to distinguish the men from the women. To do this, we painstakingly went through each frame of every video and used a program to blur the bodies. Participants who watched regular highlights evaluated the goals scored by men significantly higher. However, the difference evaporated under the experimental condition, where participants did not know if they were watching men or women playing. The results held even after controlling for demographics, whether they preferred men´s or women´s soccer and how often they watched soccer.
The findings reveal that gender biases influence fans’ perceptions of women’s soccer – and possibly other women’s sports. Precisely quantifying the influence of these biases is difficult, and these types of experiments have several limitations, such as the focus on highlights and the practice of blurring, that prevent us from drawing definitive conclusions. The findings, however, challenge conventional wisdom about the potential of the women’s sports market. Any evaluation of quality in women’s sports should stop and think about whether biases are playing a role.
People often apply a double standard in evaluating the quality of women’s sports, particularly soccer. You’ll hear the argument that “a women’s team can never beat a men’s team,” which is used to justify why people shouldn’t care as much about women’s sports. But in our view, that argument is beside the point. You don’t hear the same argument when it comes to the performance of boxers from different weight classes, or whether college basketball is an inferior product to pro basketball. Despite stereotypes, biases and a late start in the business, women’s soccer keeps growing; the most recent World Cup shattered viewership records in multiple countries. Clearly, there’s a market. And clearly, it has plenty of room to grow.
Then comes the catch. After about a minute, the viewer sees that the highlights had been artifically modified: All of the players were actually from the French women’s national team. The ad seeks to push back on a common criticism of women’s sports: that female players aren’t as entertaining as their male counterparts, and the action is less exciting than it is in men’s sports. In our recently published study, we decided to put this notion to the test. We obscured the gender of soccer players and asked participants to rate the performances of the athletes they viewed.
Sports is one of the world’s largest markets – estimated to be around US$83 billion in the United States alone. But any fan, casual or serious, can readily observe striking gender differences in media coverage and player salaries. Outside of the Olympics, only about 4% of all sports media coverage around the world is devoted to women’s sports. Live events are much less likely to be broadcast, and only one woman, retired tennis star Serena Williams, is among the Forbes’ 50 highest-paid athletes in the world.
In 2019, the U.S. women’s national soccer team filed a gender discrimination lawsuit against U.S. Soccer. Even though the squad had won the past two women’s World Cups – and the men’s team hadn’t reached a semifinal since 1930 – the women weren’t paid as much as their male counterparts. In 2022, the two sides came to an agreement guaranteeing equal pay. But the dispute offered a window into the thinking of many sports executives and fans. U.S. Soccer’s legal counsel stated that the women’s team plays “a different game” from the men’s “in the sense that men are bigger, stronger, faster.” Research has proven that women’s and men’s soccer does indeed differ in several physical aspects – for example, male soccer players cover more ground and run faster during games.
But the question is whether the physical differences of women necessarily make the games less entertaining. The existence of stereotypes points to an alternative possibility: Gender biases might influence perceptions of the quality of the games.
Physical differences are often used to sustain sexist assertions that most women’s sports are boring and slow. This narrative – especially prominent among detractors of women’s sports and internet trolls – likely influences mainstream attitudes toward sports such as women’s basketball and women’s soccer.
When gender is obscured, differences disappearPrior research shows that biases likely play a role when people evaluate the performance of women on the field and in the workplace. It includes the work of Claudia Goldin, who was recently awarded the Nobel Memorial Prize in Economics. Her fascinating 2000 study with economist Cecilia Rouse showed how blind auditions for symphony orchestras resulted in more women being hired.
Economist Claudia Goldin was able to show that when female musicians participated in blind auditions, it improved their chances of being hired. Goldin’s work inspired our study. What if we could prevent soccer fans from identifying players and test for whether gender bias influenced evaluations of the players’ athletic performance?
We set up an experiment and showed more than 600 participants highlights from men’s and women’s professional soccer videos. We asked the group, which was made up of 55% men and 45% women, to watch 10 videos and then rate the performance of the players during the goal-scoring plays on a scale of 1 to 5. In the control group, participants watched and evaluated regular videos. For the experimental group, we blurred the gender of the players, making it impossible for participants to distinguish the men from the women. To do this, we painstakingly went through each frame of every video and used a program to blur the bodies. Participants who watched regular highlights evaluated the goals scored by men significantly higher. However, the difference evaporated under the experimental condition, where participants did not know if they were watching men or women playing. The results held even after controlling for demographics, whether they preferred men´s or women´s soccer and how often they watched soccer.
The findings reveal that gender biases influence fans’ perceptions of women’s soccer – and possibly other women’s sports. Precisely quantifying the influence of these biases is difficult, and these types of experiments have several limitations, such as the focus on highlights and the practice of blurring, that prevent us from drawing definitive conclusions. The findings, however, challenge conventional wisdom about the potential of the women’s sports market. Any evaluation of quality in women’s sports should stop and think about whether biases are playing a role.
People often apply a double standard in evaluating the quality of women’s sports, particularly soccer. You’ll hear the argument that “a women’s team can never beat a men’s team,” which is used to justify why people shouldn’t care as much about women’s sports. But in our view, that argument is beside the point. You don’t hear the same argument when it comes to the performance of boxers from different weight classes, or whether college basketball is an inferior product to pro basketball. Despite stereotypes, biases and a late start in the business, women’s soccer keeps growing; the most recent World Cup shattered viewership records in multiple countries. Clearly, there’s a market. And clearly, it has plenty of room to grow.
Published in Scientific American, 05.10.2022, “Video Game Players Avoid Gay Character”
I like video games because they give me choices that I don’t have in real life. When I want to feel like a princess, I choose Peach in Mario Kart. If I want to fight like Bruce Lee, I can choose Marshall Law in Tekken.
When picking out which character to play in a game, players factor in an avatar’s strengths and weaknesses, and even their appearance. But players also make their choices based on attributes that aren’t apparent during gameplay and have no influence on the game, such as a character’s sexuality. And my research has found that other players in turn might treat those characters differently within the game. This is important for the gaming industry. Some companies actively support social movements through publicity and marketing campaigns and product labeling. However, backlash could deter these companies from supporting marginalized groups and discourage gay players from participating.
How do players factor in these kind of character traits, and what are the implications for online harassment and discrimination? The first-person shooter game Overwatch, which is widely popular and has generated more than $1 billion in revenue so far for its owner, Blizzard, has turned out to be an ideal laboratory for me and my colleagues to investigate these questions. In that sense, what happens within Overwatch has economic and societal implications.
In Overwatch, players can choose between 32 avatars that have different appearances and skill sets. As an additional marketing highlight, Blizzard regularly updates the background stories of these characters. Their stories have no effect whatsoever on the game; they don’t change the characters’ skills or appearance. Still, these stories are so important for the fan base that they influence Overwatch in another way. In May 2019, an Overwatch writer updated the background story for one male character, Soldier: 76. He wrote that Soldier: 76 had been in a romantic relationship with another man and identified as gay. It was a bit of information that is far from unusual in the real world, but in online gaming, it was somewhat revolutionary.
The announcement resulted in hundreds of responses on Twitter and other social media channels. Many players supported the announcement and liked the tweet. But some players were uncomfortable with the decision, accusing Blizzard of making money through political correctness, complaining that the announcement was unnecessary, and even threatening to stop playing Overwatch.
Our research team wanted to know if this announcement influenced the game itself. Would players change their attitude toward Soldier: 76? We examined the pick rate—the frequency with which players choose a specific character—before, during and after the announcement. Surprisingly, we found an extreme drop in the pick rate for Soldier: 76 after the news. Players shied away from playing with the newly outed character.
Interestingly, instead of choosing Soldier: 76, several players chose the only other LGBTQ character in the game: a lesbian called Tracer. Though Soldier: 76 had previously been assumed by most to be straight, Tracer had been openly gay since the game was published.
To understand what was going on, we used an online survey to ask players what they thought about the announcement and how it influenced their gaming experience. Hundreds of players responded. Most respondents were men (83.77 percent) and from North America (49.4 percent) and Europe (39.5 percent). The majority had heard about Soldier: 76 being gay but didn’t care. They reported that changing a character’s sexuality had no influence on the game. A large minority, however, felt uncomfortable and that other players discriminated against them when they played as Soldier: 76 after the announcement. They got tired of homophobic slurs and constant harassment and temporarily switched to other characters to avoid it.
Additionally, many players mentioned that the gaming community makes an important distinction between the only two LGBTQ characters in the game. Whereas Soldier: 76 is a man, Tracer is a woman. Some players reported in the survey that they felt other players see lesbians as appealing and interesting, but that the same players voice disgust toward gay men.
Detecting discrimination is important and sheds light on a problem. The next step, however, is to find a way to decrease discrimination. In another research project, my team worked with a soccer federation trying to decrease discrimination. In a randomized trial, we found that an e-mail to coaches about how the sport can promote inclusivity and reduce racism helped raise awareness. We hope to do the same in e-sports by working together with people from within the industry who want to improve the equity and inclusivity of the community.
Some people like to imagine that video games offer idealized versions of reality and teach values that seem to be lacking in the real world, such as fighting for a common cause or working together as a team. But our findings show that the world of video games has the same biases and discrimination we find in our everyday lives, and that games can be much more welcoming for some than others. Video games are always innovating, pushing art and technology forward to create new experiences for players. In the past, players and fans have adapted to this fast-changing environment. The online gaming community needs to embrace its diverse players and reckon with its biases. Players should get over themselves and accept all characters and the players who pick them.
When picking out which character to play in a game, players factor in an avatar’s strengths and weaknesses, and even their appearance. But players also make their choices based on attributes that aren’t apparent during gameplay and have no influence on the game, such as a character’s sexuality. And my research has found that other players in turn might treat those characters differently within the game. This is important for the gaming industry. Some companies actively support social movements through publicity and marketing campaigns and product labeling. However, backlash could deter these companies from supporting marginalized groups and discourage gay players from participating.
How do players factor in these kind of character traits, and what are the implications for online harassment and discrimination? The first-person shooter game Overwatch, which is widely popular and has generated more than $1 billion in revenue so far for its owner, Blizzard, has turned out to be an ideal laboratory for me and my colleagues to investigate these questions. In that sense, what happens within Overwatch has economic and societal implications.
In Overwatch, players can choose between 32 avatars that have different appearances and skill sets. As an additional marketing highlight, Blizzard regularly updates the background stories of these characters. Their stories have no effect whatsoever on the game; they don’t change the characters’ skills or appearance. Still, these stories are so important for the fan base that they influence Overwatch in another way. In May 2019, an Overwatch writer updated the background story for one male character, Soldier: 76. He wrote that Soldier: 76 had been in a romantic relationship with another man and identified as gay. It was a bit of information that is far from unusual in the real world, but in online gaming, it was somewhat revolutionary.
The announcement resulted in hundreds of responses on Twitter and other social media channels. Many players supported the announcement and liked the tweet. But some players were uncomfortable with the decision, accusing Blizzard of making money through political correctness, complaining that the announcement was unnecessary, and even threatening to stop playing Overwatch.
Our research team wanted to know if this announcement influenced the game itself. Would players change their attitude toward Soldier: 76? We examined the pick rate—the frequency with which players choose a specific character—before, during and after the announcement. Surprisingly, we found an extreme drop in the pick rate for Soldier: 76 after the news. Players shied away from playing with the newly outed character.
Interestingly, instead of choosing Soldier: 76, several players chose the only other LGBTQ character in the game: a lesbian called Tracer. Though Soldier: 76 had previously been assumed by most to be straight, Tracer had been openly gay since the game was published.
To understand what was going on, we used an online survey to ask players what they thought about the announcement and how it influenced their gaming experience. Hundreds of players responded. Most respondents were men (83.77 percent) and from North America (49.4 percent) and Europe (39.5 percent). The majority had heard about Soldier: 76 being gay but didn’t care. They reported that changing a character’s sexuality had no influence on the game. A large minority, however, felt uncomfortable and that other players discriminated against them when they played as Soldier: 76 after the announcement. They got tired of homophobic slurs and constant harassment and temporarily switched to other characters to avoid it.
Additionally, many players mentioned that the gaming community makes an important distinction between the only two LGBTQ characters in the game. Whereas Soldier: 76 is a man, Tracer is a woman. Some players reported in the survey that they felt other players see lesbians as appealing and interesting, but that the same players voice disgust toward gay men.
Detecting discrimination is important and sheds light on a problem. The next step, however, is to find a way to decrease discrimination. In another research project, my team worked with a soccer federation trying to decrease discrimination. In a randomized trial, we found that an e-mail to coaches about how the sport can promote inclusivity and reduce racism helped raise awareness. We hope to do the same in e-sports by working together with people from within the industry who want to improve the equity and inclusivity of the community.
Some people like to imagine that video games offer idealized versions of reality and teach values that seem to be lacking in the real world, such as fighting for a common cause or working together as a team. But our findings show that the world of video games has the same biases and discrimination we find in our everyday lives, and that games can be much more welcoming for some than others. Video games are always innovating, pushing art and technology forward to create new experiences for players. In the past, players and fans have adapted to this fast-changing environment. The online gaming community needs to embrace its diverse players and reckon with its biases. Players should get over themselves and accept all characters and the players who pick them.
Published in ZME Science, 04.05.2021, “Amateur sports show how people with foreign-sounding names are discriminated against”
The current pandemic has made it abundantly clear just how important social relations are. People all around the world are staying and working from home and many suffer from a lack of interaction with friends and family. Social relations are especially important when moving to another country or city. While economic research has shown that people with foreign-sounding names suffer from discrimination in various interactions (when they apply for a job, when renting an Airbnb apartment, or when trying to get an Uber) it has neglected social integration. This is very surprising as a large share of the population in many European countries is foreign-born (e.g., 29.7% in Switzerland, 19.3% in Austria, or 15.6% in Norway) — and people don’t just work and buy things, but also want to make friends and establish a social network. Without friends, life can be very unpleasant, despite a good job or a nice house. Having a social network is essential for people to feel comfortable in their (new) surroundings.
In many European countries, newcomers have a straightforward way to get access to a new social network: amateur sports. It can be hard to make social connections as an adult, and for many people, contacting one of the many local football, tennis, basketball, handball, or hockey clubs is a great way to start. Basically, that means that players meet once or twice a week, play the game, have the occasional beer after practice, and play a match against another team every other weekend or so. While being competitive is important, the focus is often on having a good time with friends while trying to stay fit. It’s a social activity as well as exercise. I am part of a team of researchers from the University of Zürich in Switzerland and the NTNU in Norway who used this setting for an interesting research idea. We created fake email accounts with typical foreign- and native-sounding names and contacted amateur football clubs asking to participate in a training session. We estimate that it is an accurate method to measure access to social integration.
Declining to invite someone for a training session is similar to not giving the person access to a social network — they’re excluded from the social circle they’re trying to access. We performed the experiment not in one but in several countries to have a comprehensive overview of access to social integration for the respective countries. We decided to perform the experiment using amateur football clubs. Football is by far the most popular amateur sport in Europe and tens of thousands of amateur football clubs exist throughout Europe. We identified 22 European countries that fit with the experimental setting. First, we gathered the emails of all amateur football clubs. Then, we translated the mail into the respective languages of the countries and created foreign-sounding names for natives and the three largest foreign groups in each country. Finally, we clarified in surveys in the respective countries that the names sounded indeed native or foreign. Before the pandemic started, we sent more than 23,000 emails to amateur football clubs all over Europe. The results showed that, on average, people with a foreign-sounding name are 10% less likely to receive a response (see the figure below). Many countries, e.g., Netherlands, Norway, Poland, and Romania, were fairly close to the average for all European countries.
But there were notable exceptions. In Hungary, Austria, and Croatia, people with foreign-sounding names were 20% less likely to receive a response. On the other hand, football clubs in Ireland, France, and Portugal discriminated only marginally against people with foreign-sounding names.Our results are important both for the general public and for policymakers. National and local governments invest considerable sums into integration-related programs, and rightfully so. Individuals who leave a city or country because they were never able to establish a social network generate real costs for themselves and for society. Thus, social integration is a field where policymakers and researchers should closely work together.
Amateur sports clubs remain a relatively easy way for foreigners to develop a social circle, but the present field experiment shows that discrimination against ethnic minority groups is still present. If we want to foster healthy social interactions for everyone, there’s still a lot of work to be done.
In many European countries, newcomers have a straightforward way to get access to a new social network: amateur sports. It can be hard to make social connections as an adult, and for many people, contacting one of the many local football, tennis, basketball, handball, or hockey clubs is a great way to start. Basically, that means that players meet once or twice a week, play the game, have the occasional beer after practice, and play a match against another team every other weekend or so. While being competitive is important, the focus is often on having a good time with friends while trying to stay fit. It’s a social activity as well as exercise. I am part of a team of researchers from the University of Zürich in Switzerland and the NTNU in Norway who used this setting for an interesting research idea. We created fake email accounts with typical foreign- and native-sounding names and contacted amateur football clubs asking to participate in a training session. We estimate that it is an accurate method to measure access to social integration.
Declining to invite someone for a training session is similar to not giving the person access to a social network — they’re excluded from the social circle they’re trying to access. We performed the experiment not in one but in several countries to have a comprehensive overview of access to social integration for the respective countries. We decided to perform the experiment using amateur football clubs. Football is by far the most popular amateur sport in Europe and tens of thousands of amateur football clubs exist throughout Europe. We identified 22 European countries that fit with the experimental setting. First, we gathered the emails of all amateur football clubs. Then, we translated the mail into the respective languages of the countries and created foreign-sounding names for natives and the three largest foreign groups in each country. Finally, we clarified in surveys in the respective countries that the names sounded indeed native or foreign. Before the pandemic started, we sent more than 23,000 emails to amateur football clubs all over Europe. The results showed that, on average, people with a foreign-sounding name are 10% less likely to receive a response (see the figure below). Many countries, e.g., Netherlands, Norway, Poland, and Romania, were fairly close to the average for all European countries.
But there were notable exceptions. In Hungary, Austria, and Croatia, people with foreign-sounding names were 20% less likely to receive a response. On the other hand, football clubs in Ireland, France, and Portugal discriminated only marginally against people with foreign-sounding names.Our results are important both for the general public and for policymakers. National and local governments invest considerable sums into integration-related programs, and rightfully so. Individuals who leave a city or country because they were never able to establish a social network generate real costs for themselves and for society. Thus, social integration is a field where policymakers and researchers should closely work together.
Amateur sports clubs remain a relatively easy way for foreigners to develop a social circle, but the present field experiment shows that discrimination against ethnic minority groups is still present. If we want to foster healthy social interactions for everyone, there’s still a lot of work to be done.