By putting forward the Attacker-Defender Game (AD-G), the target articles makes an important contribution to behavioral game theory, which so far has neglected asymmetric games. While we applaud this contribution, the purpose of our commentary is to point to one type of asymmetric conflict that is central to modern societies yet is not captured by the model.

The AD-G models the distinct roles of attackers, who are motivated to increase their gains by changing the status quo, and defenders, who are motivated to defend against losses by protecting the status quo. The different psychological arrays that characterize attackers and defenders are said to be linked to evolved psychological mechanisms that are also evident in nonhuman animals (i.e., predator and prey). However, a major characteristic of modern societies, which is absent from societies of nonhuman animals or hunters-gatherers, is that they produce stable economic surplus, leading to the creation of group-based hierarchies (based on ethnicity, nationality, class, etc.) (Sidanius & Pratto Reference Sidanius and Pratto1999).

In these societies, disadvantaged-group members are typically more strongly motivated to change the status quo toward greater equality, whereas advantaged-group members are typically more strongly motivated to protect the status quo that privileges them (Saguy et al. Reference Saguy, Dovidio and Pratto2008). Thus, the motivations of disadvantaged-group members and attackers (i.e., change the status quo) and advantaged-group members and defenders (i.e., protect the status quo) are basically similar. Whereas the authors acknowledge that “asymmetry in power … can dramatically change the motivation to attack or to defend” (De Dreu & Gross, sect. 2.3, para. 2), we argue that power asymmetry creates substantially different psychological arrays. Specifically, we suggest that the psychological arrays that are typical of members of disadvantaged and advantaged groups are quite opposite to those of attackers and defenders (respectively). We now turn to illustrate these opposing patterns in six examples.

First, whereas in-group identification is stronger among defender than attacker groups, disadvantaged-group members, who often confront prejudice and discrimination due to their group affiliation, are typically more strongly identified with their groups than advantaged-group members (Simon & Brown Reference Simon and Brown1987). In fact, because they rarely confront prejudice and discrimination, advantaged-group members are sometimes oblivious to their group identity (e.g., whites who profess to be color-blind; Helms Reference Helms1990).

Second, whereas feeling superior is functional for attackers more than for defenders, superiority beliefs (e.g., in the form of racial or religious supremacy ideologies) are more common among members of advantaged versus disadvantaged groups (a pattern termed ideological asymmetry; Sidanius & Pratto Reference Sidanius and Pratto1999). In fact, disadvantaged-group members sometimes internalize the stigma on their group (e.g., sexual minorities who accept homophobic attitudes as legitimate; Herek & McLemore Reference Herek and McLemore2013).

Third, attackers, but not defenders, are likely to be overconfident (i.e., overestimate their relative strength). Yet, group efficacy (the belief that one's group has control over and can change society; Mummendey et al. Reference Mummendey, Kessler, Klink and Mielke1999) is typically stronger among advantaged groups compared with disadvantaged groups. To illustrate, individuals with lower (vs. higher) subjective social class showed cultural practices (including interests in education, arts, newspapers, TV, and shopping) that relate to recognizing the in-group's low group efficacy and, in turn, remained politically inactive when faced with an in-group–related social disadvantage (Becker et al. Reference Becker, Kraus and Rheinschmidt-Same2017).

Fourth, members of defender groups are more likely to cooperate with each other than members of attacker groups, who are more likely to show self-serving behaviors (e.g., free-riding). Yet, research on social class shows that members of disadvantaged groups (i.e., lower class individuals), who must rely on mutual aid, show more communal, prosocial behavior toward others than members of advantaged groups (higher class individuals), who enjoy greater independence and show more agentic, self-serving behavior (Kraus et al. Reference Kraus, Piff, Mendoza-Denton, Rheinschmidt and Keltner2012).

Fifth, attack is associated with activation of neural circuitries involved in the processing of rewards, whereas defense is associated with neural circuitries involved in threat detection and risk avoidance. By contrast, because elevated power activates approach-related tendencies, whereas reduced power activates inhibition-related tendencies (Keltner et al. Reference Keltner, Gruenfeld and Anderson2003), advantaged-group members are more attentive to rewards, whereas disadvantaged-group members are more attentive to punishments and threats in their environment (Rucker et al. Reference Rucker, Galinsky, Magee, Olson and Zanna2018).

Sixth, because attack means that targets may be harmed, subordinated, and exploited, attackers are more likely to experience guilt than defenders. However, especially when reminded of social inequality, advantaged-group members are more likely to experience group-based guilt than disadvantaged-group members. For example, when exposed to information about gender inequality, men expressed greater moral shame and wished that their in-group would behave more morally than women (Hässler et al. Reference Hässler, Shnabel, Ullrich, Arditti-Vogel and SimanTov-Nachlieli2018).

We conclude that although members of attacker and disadvantaged groups, on the one hand, and members of defender and advantaged groups, on the other hand, share the same basic motivation pertaining to the status quo, the AD-G should not be viewed as a representation of the dynamics between advantaged and disadvantaged groups. Why is it important to explicitly note this? The representations, framings, and language used by social scientists can inform and influence the public discourse about the topics that they study. Hence, our purpose in this commentary is to highlight that disadvantaged and advantaged groups should not be equated with attackers and defenders, and the relations between disadvantaged and advantaged groups should not be described as “a clash between one side seeking change and … victory, and the other side … protecting against loss and defeat” (De Dreu & Gross, abstract). Rather, as opposed to the AD-G's utility matrix, a disadvantaged group's success in changing the status quo does not necessarily involve a loss to the advantaged group. For example, from a broad historical perspective, white Americans did not “lose” because of the abolishment of slavery and racial segregation, and men did not “lose” because of the suffrage movement's success in achieving voting rights for women. Concurring with Paulo Freire's (Reference Freire1970) notion that “the great humanistic task of the oppressed is to liberate themselves and their oppressors,” we argue that changing the status quo toward greater equality may benefit both the advantaged and the disadvantaged – by allowing individuals to enjoy secure and positive social identities (Nadler & Shnabel Reference Nadler and Shnabel2015) and live to their full potential to the benefit of society as a whole.