Keywords

1 Introduction

Games are often framed as products bought purely for their entertainment value but emotions in games have attracted the interest of researchers in recent years. With the development of computer graphics and advances in game artificial intelligence (AI), the linear game experiences have evolved into games with stories or games with choices. As the player is provided a chance to select between a few alternatives that may affect the story or the virtual world, emotions and empathy in video games have become more important. It is critical to note here that games create a different kind of empathy when compared with the traditional media and the distinguishing feature of this medium is the ability of the player to change or affect the things that are happening in the game. So new questions arise for both the designers and the researchers: What happens when the player feels empathy towards the game characters? How can the player know what an AI controlled character is thinking and feeling? What leads the player to respond with sensitivity and care to the suffering of another character? When the player starts thinking and feeling from the protagonist’s perspective (also called role-playing), does this affect the choices they make? Is it possible for an empathetic person in real life to transform into an egoist person as a player? In this regard, this paper attempts to investigate the relationship between the empathy of players and the moral choices they made in story driven games.

When we look at the etymology of “empathy,” we can see that it is acquired from the Ancient Greek word “έμπάθεια” (empatheia). The meaning of “empatheia” is physical affection or passion. However, Edward Bradford Titchener first uses the word “empathy” in English in 1909 (Stueber 2008). He translated empathy from the German word “Einfühlung,” which means “in-feeling” or “feeling into.” The Austrian-American psychoanalyst Heinz Kohut approached empathy from the perspective of clinical theory and defined empathy as “vicarious introspection” or “the capacity to think and feel oneself into the inner life of another person” (Kohut 1959). Since the late 1970s empathy referred to the special type of proper vicarious emotion, which was more other-oriented than self-oriented (Batson and Coke 1981). A broader definition of empathy is provided by Batson (2009) who suggested eight different conceptualizations of empathy: (a) knowing another person’s internal state, including his or her thoughts and feelings, (b) adopting the posture or matching the neural responses of an observed other, (c) coming to feel as another person feels, (d) intuiting or projecting oneself into another’s situation, (e) imagining how another is thinking and feeling, (f) imagining how one would think and feel in the other’s place, (g) feeling distress at witnessing another person’s suffering, and (h) feeling for another person who is suffering. The conceptualizations of Batson are more suitable for interactive stories or story driven games, which aim to convey an experience rather than the interactive systems, which focus on exploring the game’s mechanical depth (Fischer 2017).

Empathy is a multidimensional phenomenon and there are many different definitions from different disciplines as philosophy, psychology, and sociology. Nowak (2011) presented 57 diverse scientific definitions of empathy and Cuff et al. (2016) provided 43 different interpretations. This study accepts the basic definition well accepted by many researchers: Empathy is understanding and sharing the emotions of others (Davis 2018; Decety and Jackson 2006; Baron-Cohen and Wheelwright 2004). From a gaming perspective, others can be both the non-player characters (NPCs) and other players that the player can interact with. Another relevant debate is whether empathy is an affective or a cognitive concept. Generally, affective empathy is an emotional or reactive response to the emotions of other individuals and it is an instinctual reaction. Affective empathy is also described as emotional empathy (Davis 2018; Dziobek et al. 2008). As stated in Hoffman’s (1987) definition, empathy is an affective response and more appropriate to another’s situation than one’s own. On the other hand, according to the cognitive approach, empathy is an understanding of the emotions of other individuals (Baron-Cohen and Wheelwright 2004) and it is a conscious attempt to put oneself in another’s shoes. Ickes used the term “empathic accuracy” and defined as a one’s capability to accurately infer the specific content of another individual’s feelings and thoughts (Ickes 1993). In terms of the chosen games of this study, players are expected to react to the emotions of others or to understand them, whether the emotional response originated from an affective or a cognitive nature. The core of the non-critical interactions in the story driven games is caring about others and understanding how they will be affected by player decisions, which are important in fostering this sense of empathy (Ryan et al. 2016).

2 Theoretical Framework

In order to analyze player decisions in a story driven game, it is expected that players with high empathy will show prosocial behavior and players with low empathy will show antisocial behavior. In this regard, first the relationship between empathy and prosocial behavior will be summarized. Aiming to categorize player decisions within a story driven game, several different frameworks were analyzed and the hot cognition/cold cognition dual system is chosen for the purposes of this study. This dual system, which is supposed to facilitate the analysis of player choices in pre-scripted story driven games, is also explained in this section.

2.1 Prosocial Behavior and Empathy

We humans are motivated to helping others. Not only to humans, we also work hard to help and rescue animals. Why helping others is so important for us? What makes us help earthquake, hurricane, or war victims from other parts of the world? Helping is the completely voluntary act and it is a part of the prosocial behavior which consists of acts such as helping, sharing, cooperating, comforting, donating, caring, volunteering, and other positive attitudes which intended to benefit another. Many human behaviors can be explained by the empathy phenomena, and the lack of empathy is usually translated to antisocial behavior. Similarly, high empathic ability is assumed to lead to prosocial behavior. One of the first researches that focus on the connection between empathy and altruistic behavior was Krebs (1975) but it was difficult to diagnose whether egoistic or altruistic motivation caused to help others. Similar experiments on the empathy-altruism hypothesis (Batson et al. 1981, 1988, 1991) claimed that the prosocial motivation caused by empathy is aimed towards the final goal of increasing welfare of a person in need. But in a virtual world defined by the rules of a video game, are the players more interested in increasing their own welfare or the other’s? It is important to note here that the morality systems of video games are inherently restricted and limited by ludic and business considerations where moral choices are usually flattened down into mere narrative flavoring rather than a reflection of an individual’s ethical makeup (Heron and Belford 2014).

As stated by Batson et al. (1981), if you try to help another individual from a desire for personal gain or to avoid personal distress, then you are “egoistically motivated” and “directed toward the end-state goal of increasing your own welfare.” On the contrary, if you help from a desire to decrease the distress or increase the benefit of another individual’s, then you are “altruistically motivated” and “directed toward the end-state goal of increasing the other’s welfare.” You may be feeling empathy for another person and you may try to reduce his or her distress but if the end-state goal of your action is to reduce your own distress, this is not an altruistic behavior but an egoistic response. Altruism is utterly an unselfish act. Batson et al. (1981, p. 291) have three observations for differentiating egoism and altruism: (a) helping, as a behavior, can be either egoistically or altruistically motivated; it is the end-state goal, not the behavior, which distinguishes an act as altruistic, (b) motivation for helping may be a mixture of altruism and egoism; it need not be solely or even primarily altruistic to have an altruistic component, and (c) increasing the other’s welfare is both necessary and sufficient to attain an altruistic end-state goal. Regarding these observations, games are also known for rewarding the act of helping others where increasing the welfare of others also increases the welfare of the player, which makes it more difficult to understand whether players are egoistically or altruistically motivated.

2.2 Hot Cognition and Cold Cognition

Hot versus cold cognition is one of the dual-system models used broadly in psychological science. Generally, this temperature metaphor is very well known in psychology and used to categorize emotion-driven and motivational processes against information-driven and cognitive processes. “Hot” cognition refers to affect-laden, emotion involved cognitive processes contrary to the “cold” cognition, which mainly related to logic-based, emotionless, rational, information-driven cognitive processes (Abelson 1963). “Hot” cognition is directed by feelings and desires, which linked with our motivations, and in conclusion, it affects our decision-making (Kunda 1999). Metcalf and Mischel classified “hot” emotional “go” system as “the basis of emotionality, fears as well as passions-impulsive and reflexive-initially controlled by innate releasing stimuli” and on the contrary, “cool” cognitive “know” system as “emotionally neutral, contemplative, flexible, integrated, coherent, spatiotemporal, slow, episodic, and strategic” (Metcalfe and Mischel 1999). However, according to the appraisal theory emotions are the result of evaluations, which makes “hot” cognition itself a little ambiguous (Scherer 1999; Roseman and Smith 2001). Players of a computer game also evaluate the situation they are in and try to guess the implications of their choices such as rewards or punishments in the form of experience points or gold, which may lead to a cold cognition oriented decision-making process.

There are also similar classifications in literature with different names. For example, another similar practice “system 1” refers to automatic, quick, and cognitively effortless operations in contrast to “system 2” which refers to more concentrated, controlled, cognitively effortful operations (Kahneman and Egan 2011). We used “hot” versus “cold” cognition system for classification of the major game choices in the selected games. The developers identified these major choices and they were also tracked for each player for demonstrating a summary of worldwide statistics on each one. After an analysis of the major choices provided in each game, they are classified into two categories: choices related with emotions like empathy, fear, anger, disgust, admiration, sadness, panic (hot cognition) and other choices related to cognitive processes as logical thinking, coherence, rationality, decision-making (cold cognition). For example, two different but similar events analyzed in this context which the player can choose is: “killing” or “not killing” an NPC. In this instance, the action of “killing” can be defined as “hot” or “cold” cognition depending on the particular game event and the emotional arousal state. It means that the action—“killing”—can be logical or emotional, based on game context. The games analyzed in this study were pre-scripted and story based, which simplified the categorization process. Besides that, the choices were analyzed from the players’ point of view during classification.

2.3 Analyzing Games in the Context of Empathy

Research on empathy and games have usually been narrowed down to violent video games and aggression (Bartholow and Anderson 2002; Coeckelbergh 2007; Anderson et al. 2010; Happ et al. 2013) but other recent studies found that exposure to prosocial video games increases empathy and accessibility of prosocial thoughts, as well as decreasing reported pleasure at another’s misfortune (Greitemeyer et al. 2010; Greitemeyer and Osswald 2011). But how do game designers create game choices that may be affected by the empathy level of the players and do the empathy level of the players really affect their choices? Games commonly use tragic events in order to create an empathic situation, such as an unfortunate accident or the loss of a character. This design perspective is quite in line with the famous proverb: “Before you criticize a man, walk a mile in his shoes.” According to Bogost (2011) video games frequently ask the players to fill the “big shoes” by assuming powerful roles to achieve and experience great things in a game but Bogost (2011, p.19) also suggested that players should also try “smaller shoes”:

If a game about the Sudanese genocide is meant to foster empathy for terrible real-world situations in which the players fortunate enough to play video games might intervene, then those games would do well to invite us to step into the smaller, more uncomfortable shoes of the downtrodden rather than the larger, more well-heeled shoes of the powerful.

The games that allow players to empathize with the story or the game characters are still few since it requires intricate narrative design and believable non-player characters (NPCs). Recently, games started to use more emotion-driven narrative stories. Choice-based games give the player the chance to be more interactive and emotionally responsive during the gameplay. Meaningful choices help players to contact the game characters personally and empathize with them. Given below are some significant examples from the games industry.

This War of Mine (11 bit studios 2014) is a point and click survival-management war game created by 11 bit studios. The game is based on the siege of Sarajevo, Bosnia and inspired by the stories of real war victims. This War of Mine is one of the unique games, which gives the player a chance to look at war from a totally different perspective. Focusing on the civilian experience of war rather than common soldier’s point of view in order to show the horrors of war. The player controls a group of civilians trying to stay alive in a city surrounded by the enemy.

That Dragon, Cancer (Numinous Games 2016) is a documental point and click exploration game that demonstrates the experience of parents supporting their five-year-old son Joel, who fights against cancer. The designer of the game Ryan Green is the father of Joel. The game consists of several short vignettes that show the Greens’ tragedy. This game gives player a chance to look at all these events from the perspectives of parents and even doctors.

2.4 Selected Games: The Walking Dead and the Wolf Among Us

The two games selected for this study are The Walking Dead (Telltale Games 2012) and The Wolf Among Us (Telltale Games 2013), both developed by Telltale Games. Although they do not emphasize puzzle solving but focus on story and character development, both games are classified as episodic adventure video games. The Walking Dead is based on The Walking Dead comic book series and tells the events occurring shortly after the onset of a zombie apocalypse in Georgia. Lee Everett is the protagonist of the story, convicted for murder and being transferred to prison. The vehicle transporting Lee hits a zombie and crashes, he manages to escape and later finds a little girl named Clementine, and promises to help her find her parents. The Wolf Among Us is based on Fables comic book series, which features various characters from fairy tales and folklore that are living in a close community within New York City known as Fabletown. The protagonist of the story is the Sheriff Bigby Wolf, known as Big Bad Wolf in fairy tales, investigating the murder of a woman. Assuming that the setting of a game may affect the nature of player choices in a game, both games are used in the study where one features a post-apocalyptic Georgia and the other takes place in a hidden, magical community in New York.

These games have a special cartoon-like art style, which separates it from a realistic appearance. Both are story driven games and they are also classified as interactive movies with special emphasis on ethical decisions and interpersonal relationships. They can also be defined as story driven emotional experiences that use the Choose Your Own Adventure (CYOA) concept very successfully. Telltale tagline at the beginning of each game is: “This game series adapts to the choices you make. The story is tailored by how you play.” Steam also announces one of the key features of Telltale games as: “Now, it’s not only WHAT you choose to do that will affect your story, but WHEN you choose to do it.” Conversation with characters is established via dialogue choices where player decisions play an important role. When the players focus on the action on-screen, it is questionable if they are going to show reactive responses to the emotions of other individuals (Newman 2002). So, the major choices in the storyline of Telltale games are usually given in off-line moments to make the player feel both emotional and cognitive empathy towards the characters (Smethurst and Craps 2015).

The selected games offer meaningful choices for the player but in terms of story structures they are more like an illusion of a choice. Player’s relationship with the characters is affected by the choices but the game remains the same without a direct impact to the game’s narrative. In order to motivate the player to make moral decisions for moral reasons, both games try to give the player a strong moral identity to role-play, which is defined as an effective way of achieving that end (Ryan et al. 2016). The role-playing value does not create branches in the storyline but aims to create persistent responses in the NPC dialogue options. For example, at the beginning of the Wolf Among Us game, the protagonist Sheriff Bigby responds to a public disturbance. After a small conversation with a character in front of the building, the player heads upstairs and faces another character, Woodsman, who hits a woman. It does not matter how the player handles the conversation here, the game forces the player to fight with Woodsman. After the situation is under control, the player will learn that the woman, Faith, is a prostitute who wants his money back from Woodsman. At this point, the player must choose between “Give her some money” or just saying “Wish I could help.” If the player chooses to give his money to Faith, he/she will not have a chance to use money in another point of the game. And after this major choice, the story unfolds the same way it was designed to be (see Fig. 16.1 above). The story do not branch but the game aims to create a bond between the player and the women in distress, who will be murdered later and the player will be responsible for the investigation of her death.

Fig. 16.1
figure 1

Choice tree of the wolf among us episode one opening

Full size image

3 Participants and Methodology

Participants were volunteers from a Turkish university and they were interviewed on which Telltale games they have played before, if they played any. The selection criterion of students is that they should not have played at least one of the selected games (The Wolf Among Us or The Walking Dead) before. Those that played both games did not participate in the study and if they played one of the games before, the game that they have not experienced is selected for them. Those that have not played both games were randomly assigned to one of the two video games. At the end, 51 students (11 women, 40 men) participated in the study (mean age 21 years). Before they know the specifics of the experiment, informed consents were obtained from the participants and they were instructed to complete The Empathy Quotient (EQ) questionnaire. The measurement of empathy is a serious challenge for researchers from different disciplines and the selected EQ scale defines empathy as “the drive to identify another person’s emotions and thoughts, and to respond to these with an appropriate emotion” (Baron-Cohen and Wheelwright 2004, p. 361). A team of research assistants recorded the scale results and then the participants were informed that they would take part in a study where they have to play two episodes from a selected game. During the gaming sessions all participants were tested individually and the gameplays (4–5 h for each participant) were recorded for future analysis.

There were 11 major choices for The Wolf Among Us and 10 for The Walking Dead, which were identified and tracked by the developers. The players can review their choices at the end of each chapter and see what percentage of other players made the same choice. These major choices were analyzed by three researchers separately and coded as either a cold cognition choice or a hot cognition choice. The codes were compared for inter-coder reliability and choices that were coded differently were removed from the study. At the end, 7 choices from The Wolf Among Us and 8 choices from The Walking Dead were retained in the study. The recorded gameplay videos were analyzed by both researchers: (1) to detect the dialogue options they chose, (2) to identify the major choices they made, and (3) to determine their response time to important choices provided by the game. During the gameplay analysis, interview questions were identified and notes were taken for each participant. Then the participants were called again for semi-structured interviews where the researchers discussed the choices they made, the emotions they felt while playing the game and the empathy they felt towards the game characters.

4 Findings and Discussion

4.1 Statistical Tests

Our hypothesis was that the number of hot or cold choices made by the participants would have a relationship with their empathy score. In other words, we wanted to analyze if the total number of hot or cold choices differed due to the empathy score of the participant. More specifically, we wanted to see if there is a significant difference between participants of high/low empathy in regard to the total number of hot or cold choices they made. Empathy score of the participants is a continuous (scale) variable, which ranges between 1 and 80 in the questionnaire. “Hot-Total” is conceptualized as “the total number of hot choices a player makes in the game.” These are the choices coded as choices of hot cognition in the qualitative coding process. “Cold-Total” is conceptualized as “the total number of cold choices a player makes in the game.” These are the choices coded as choices of cold cognition in the qualitative coding process. After such conceptualization and operationalization, we firstly started to check if any meaningful relationships existed between Empathy and “Hot-Total” or Empathy and “Cold-Total” as we have predicted in our research hypothesis.

Accordingly, we have searched for any meaningful bivariate correlations among Empathy, Hot-Total, and Cold-Total. As shown in Table 16.1, there were no meaningful bivariate correlations either between Empathy and “Hot-Total” or Empathy and “Cold-Total” in spite of being predicted to exist (p > 0.05). Besides, a meaningful, strong (Pallant 2013) bivariate negative correlation between “Hot-Total” and “Cold-Total” (p < 0.001) has been observed as predicted as the normal consequence of the conceptualizations of the two concepts.

Table 16.1 Bivariate correlations among Empathy, Hot-Total, and Cold-Total
Full size table

Secondly, we have categorized Empathy, Hot-Total, and Cold-Total in two sub-categories to check whether we could identify any meaningful relationships among these variables. Accordingly, Empathy has been divided as Low vs. High, based on the average empathy of participants. Hot-Total has been divided as Low Number of Hot Choices vs. High Number of Hot Choices, based on the average number of hot choices made by the participants; and Cold-Total has been divided as Low Number of Cold Choices vs. High Number of Cold Choices, based on the average number of cold choices made by the participants. Then crosstabs and Pearson Chi-Square tests have been applied with these new categorical variables. However, again no meaningful relationship has been found between Empathy and Hot-Total (X 2 = 0.126; df=1, p > 0.05] and Empathy and Cold-Total (X 2 = 0.345; df=1, p > 0.05] as shown in Tables 16.2 and 16.3.

Table 16.2 Relationship between two-levels of Empathy and two-levels of Hot-Total
Full size table
Table 16.3 Relationship between two-levels of Empathy and two-levels of Cold-Total
Full size table

Thirdly, we have categorized Empathy in four sub-categories (namely: Low, Average, Above Average, and High) just like it was done in the original questionnaire. Then, again, we checked if we could identify any meaningful relationships between those sub-categories of Empathy and total number of hot choices in the game (Hot-Total) or total number of cold choices in the game (Cold-Total). In other words, we wanted to see if the total number of hot or cold choices in the game differed due to different levels of Empathy. As shown in Table 16.4, the one-way unrepeated MANOVA test between these Empathy categories and the hot/cold choice totals also yielded no significant relationship [F (4, 94) = 1.85, p > 0.05; Wilks’ Λ = 0.859, partial η 2 = 0.07].

Table 16.4 Relationship between four-levels of Empathy and Hot-Total or Cold-Total
Full size table

According to the findings all above, we concluded that the hot/cold dual system and the analysis based on the total number of hot/cold choices players made might not be working. So, at this final stage, instead of looking at the total number of hot/cold choices, we decided to take a look at each individual choice for a relationship between the choice and the empathy scores of the individuals with four categories (low/average/above average/high). Among the 21 choices in the selected games, only two of them show a significant relationship between the player choice and empathy, these choices are chronologically the 13th and 14th choices in the study and we named them as C13 and C14, respectively. Both choices are from The Walking Dead game and the findings are shown in Tables 16.5 and 16.6.

Table 16.5 Relationship between four-levels of Empathy and the player choice for choice #13
Full size table
Table 16.6 Relationship between four-levels of Empathy and the player choice for choice #14
Full size table

In Table 16.5, the relationship between four-category-empathy and the player choice (being revive or kill Larry) in choice #13 is analyzed. The relationship between these variables has been found to be significant (X 2 = 7.953; df = 1, p < 0.05, two-tailed Fisher's exact test), strong (|Phi| > 0.50), and in the diagonal cells. Accordingly, having low level of empathy making it more likely that the player’s #13 choice is to revive and having average level of empathy making it more likely that the player’s #13 choice is to kill.

This is a choice in the Walking Dead game where the NPC character Larry suffers a massive heart attack, dropping to the floor unconscious and appears to have stopped breathing. The Walking Dead WikiFootnote 1 defines Larry as: “Though he mainly cares about the safety of his daughter and generally has good intentions, his loud, cantankerous, obstreperous, and judgmental attitude causes him to be a thorn in the side of most of the group.” The player now has two choices: to help and revive Larry or assuming that it is too late to help him, killing him before he reanimates as a zombie. In this fictional setting, one becomes a zombie upon death irrespective of the manner in which one dies. So, killing Larry is coded as a cold choice since it is the rational and logical way to deal with the situation at hand and reviving Larry is coded as a hot choice, an action related with either the panic of losing a major character or the sympathy felt towards him. Although it can be concluded that players with low empathy choose to revive Larry and players with average empathy choose to kill the character, this correlation fails to explain a meaningful relationship between the empathy of players and the choice provided by the game. It can also be argued that players with low empathy should choose to make the rational choice and kill him but the relationship shows the opposite.

In Table 16.6, the relationship between four-category-empathy and the player choice (whether the player killed both brothers or not) in choice #14 is analyzed. The relationship between these variables has been found to be significant (X 2=6.390; df = 1, p < 0.05, two-tailed Fisher’s exact test), strong (|Phi| > 0.50), and in the off-diagonal cells. Accordingly, having average level of empathy making it more likely that the player’s #14 choice is killing both brothers and having low level of empathy making it more likely that the player’s #14 choice is not killing both brothers.

This choice in the Walking Dead game is related with two major characters, Danny St. John and Andrew St. John. In the game, the St. John family lures unwary survivors into their farm to murder them, dismembering and trading their meat to bandits in return for protection and supplies. The Walking Dead wikiFootnote 2 defines Andrews as: “… a cannibal who hides his true nature behind a kind and caring personality,” whereas defines Danny as: “Unlike his mother and brother, who appear to commit their horrendous actions out of necessity and to survive, Danny is more sadistic and appears to relish in violence and killing and takes sick pleasure in toying with his victims before dismembering them into human meat.” Killing both brothers is coded as a cold choice since the player assesses the situation and makes the rational choice since both brothers pose a danger to everyone. Not killing both brothers is coded as a hot choice since the player empathizes with their situation and desperation, deciding to spare one or both of them. Although it can be concluded that players with low empathy choose not to kill both brothers and players with average empathy choose to kill both brothers, this correlation again fails to explain a meaningful relationship between the empathy of players and the choice provided by the game. It can also be argued that players with low empathy should choose to make the rational choice and kill both brothers but the relationship shows the opposite.

4.2 Qualitative Analysis

The semi-structured interviews conducted for this study explored participants’ experiences through discussions about ethics, choices, and empathy in the game they played. All interviews were recorded and transcribed for further analysis. Although the players were expected to form an emotional bond with the story of the game they play, four recurring concepts that contradict this assumption were identified during the qualitative coding process:

  • The curiosity of the player

  • The “not real” phenomenon

  • The protagonist effect

  • Role-playing value

The first repeating idea is the curiosity of the player. Some players wish to see what will happen if they make a good or bad choice and they were just experimenting with the game.

If I made a very bad choice in the game, I made it because I wanted to see what would happen next. Sometimes, I just play the bad guy to play the bad guy. (P3, male)

I tried to see the consequences of my choices, both for good and bad choices. I wondered how the game would progress. (P4, male)

I do not do the right things in the game. I made different choices to experience different things. (P7, male)

Sometimes I do the things I can't do in real life. Normally, I am not an aggressive person but I tried to experience being aggressive in this game. (P8, female)

I intentionally made the wrong choices to see what will happen. (P36, male)

The second repeating idea is named as the “not real” phenomenon. Some players classified their experience as “just a game,” so they do not feel themselves responsible for their actions and do not feel empathy towards the game characters.

I was not trying to the right thing in the game because I know that it does not effect my real life in any way. (P3, male)

I know that it was just a game and I am usually not carried away by the events of the game. (P4, male)

This is not real life. I can lie, I can do whatever I want at that moment. (P11, male)

I was not trying to the right thing because it is just a game. (P13, male)

I killed some characters because I wanted to kill them, it is just a game. (P23, male)

I am not responsible for my actions because this is a game and I tried to make the choices that I can't make in real life. (P30, female)

I never regret my choices in a game because I know that it is just fiction, just like a movie. (P37, male)

The third repeating idea is named as the protagonist effect because some players only focus on the character they play, the protagonist of the story, and their choices depend on the empathy they feel for him/her. Special emphasis was given on the control of a character. Kway and Mitchell (2018) also reported that players feel the need to ensure that their actions led to a consistent characterization of their playable character.

I was trying to feel what the protagonist should feel at the moment but I do not care for the other characters since I am not controlling them. (P5, male)

I feel more empathy towards the protagonist of the story since I play him. (P15, female)

I feel more empathy towards the protagonist because I was making choices for him and I was playing him. (P17, male)

There was only one playable character so I just feel empathy towards him. (P18, male)

You see the NPCs occasionally and you either forget them or ignore them. I feel more empathy towards the protagonist. (P40, male)

You see the world through the protagonist’s eyes so you feel more empathy towards him, as for the other characters I do not feel much. (P45, male)

The fourth repeating idea was the role-play value. They imagine a role for themselves (good or bad) and try to role-play it throughout the game. In the case of The Wolf Among Us, the game gives the player a tough guy with a dark past, the Big Bad Wolf and The Walking Dead give the player a criminal convicted for murdering a state senator who slept with his wife, Lee Everett.

I do not remember most of my choices but I know that I focused on role-playing myself. (P8, female)

I tried to make the decisions my character should make. Even if my decision looks like a bad one, I chose it if it suits my character. (P12, male)

I focused on what I aim and try to make the choices the game expects of my character. (P13, male)

I was trying to role-play a bad guy, a villain. So, I do not feel empathy in any of my choices. (P19, male)

I made the choices that are ‘right’ for my character. I feel connected to him. (P20, female)

I tried to put my empathy aside because there is a goal to achieve and a character to role-play (P32, male)

5 Conclusion

Games are complex goal oriented systems where players engage in an artificial conflict that results in a quantifiable outcome. Player’s motivation to reach a goal is influenced by both personal and situational factors. The needs, motives, and goals of the player are effected by the opportunities and incentives provided by the virtual world (Bostan 2009). People play games to win and to have fun in the process and this does not mean that they have to do the “right” or “moral” things that they would do in their real lives. Although story driven games are designed to strengthen the emotional bond between the player and the game, players are still constrained by the rules and the mechanics of the game. Formal constraints of a game determine the invisible borders of what “makes sense” to do within the limits of the narrative and the overall genre of the digital text (Mateas 2004). Players have their own agendas; their own goals and they expect to be rewarded at the end regardless of the nature of their choices. The classical psychological constructs such as the hot/cold cognition dual system might not work in the context of games because player decision-making process is different than how they make their decisions in real life. They usually evaluate what they will lose or what they will win at the end of a choice and they know that it is just a game, it is “not real,” which is one of the recurring concepts identified by the semi-structured interviews of this study. This may also be the result of extremely shallow choice systems of games, offering few genuine opportunities for real moral reflection (Heron and Belford 2014).

When the player knows that the events that unfold in front of his/her eyes are “not real,” they are also inclined to experiment with the game, such as making choices that they would never make in the real life or intentionally making a bad decision just to see what will happen next. This curiosity factor determined in the qualitative analysis of this study is also related with the fun value of a game. Some people like to experiment with the game, testing the limits of the narrative and the game mechanics. Belman and Flanagan said that people are likely to empathize only when they make an intentional effort to do so at the beginning of the game but most people play unempathetically (Belman and Flanagan 2010). With this inclination to play a game without empathy, to truly challenge a player’s ethical reasoning a game should give the player opportunities to reflect, learn and improve, just as they would learn to improve other physical or intellectual skills (Ryan et al. 2016). In this regard, the protagonist of the story also effects how people make their decisions in a game. Some people like to step into the shoes of the protagonist so that they make choices not as themselves but as the character they play in a fictional world. This protagonist effect is closely related with the role-playing value identified during the interviews of this study. Some players imagine a fictional role for themselves and like to make their decisions within the boundaries of this role, which they found enjoyable. The findings of this study are also in accordance with the fact that getting the player to break the habits cultivated by playing dozens of morally inert, amoral games is not an easy task because most of the players already learned to ignore the moral dimension of their in-game behavior in favor of maximizing ludic outcomes (Ryan et al. 2016).

It should also be noted that the empathy scores of players in this study reflect the level of empathy they have in real life but players do not have to replicate their personalities in a fictional world. People of high empathy may simply experiment with the game or may desire to role-play a very different type of person since they know that this is just a game. Similarly, people of low empathy may feel connected to the protagonist of the story and make empathetic decisions that will benefit the character they play. Turkle explained this phenomenon with the differences between the “real self” and the “second self,” pointing out to the unparalleled opportunities to play with one’s identity and try out new ones in virtual worlds (Turkle 1994). Lastly, it should always be kept in mind that story and the characters in it usually provide a context for gameplay, a justification for the activities the player is expected to engage in the game. The distinctive feature of the medium of games is the ability to interact with it, which creates a more active experience than the passive experience of watching a movie and leads to new opportunities like experimenting with the game or role-playing a character.