Original Article
Hormonal responses differ when playing violent video games against an ingroup and outgroup

https://doi.org/10.1016/j.evolhumbehav.2009.07.002Get rights and content

Abstract

For 14 teams of three young men, salivary testosterone and cortisol were assessed twice before and twice after competing in within-group and between-group video games that simulated violent male–male competition. Men who contributed the most to their teams' between-group victory showed testosterone increases immediately after the competition, but only if this competition was played before the within-group tournament. High-scoring men on losing teams did not show this immediate effect, but they did show a delayed increase in testosterone. In contrast, high-ranking men tended to have lower testosterone and higher cortisol during within-group tournaments. The results are consistent with the hypothesis that men's competitive testosterone response varies across ingroup and outgroup competitions and is muted during the former. The testosterone response during the between-group competition also suggests that violent multiplayer video games may be appealing to young men because they simulate male–male coalitional competition.

Introduction

According to the Entertainment Software Association (2008), annual video game sales in the United States have grown from $2.6 billion in 1994 to $9.5 billion in 2007. Online games that incorporate a group-level social component have an unprecedented level of player commitment. Yee (2006) surveyed 30,000 online gamers (85% were male) from 2000 to 2003 and found that they spend an average of 22 h/week playing their chosen game and 9% reported spending more than 40 h/week. It is clear that substantial amounts of time and money are spent on video games, and substantial research efforts are devoted to understanding how game play influences aggression (Anderson & Dill, 2000, Bartholow et al., 2006, Kirsch, 2003). At the same time, there is comparatively little research on the underlying appeal of video games. We focus on this question from an evolutionary perspective and on how these games can be used to test evolutionary predictions.

Among the many components of human social evolution are one-on-one and coalitional male–male competition, and the ability to mentally simulate associated behavioral strategies (Alexander, 1989, Flinn et al., 2005, Geary, 2005). Our first hypothesis is that aggressive multiplayer video games are appealing to men because the games engage men's motivational disposition to participate in coalitional male–male competition and engage the systems that enable the mental simulation of corresponding social strategies: these video games create competitive situations that mimic the mental strategizing that evolved, in part, to support male–male competition. If this hypothesis is correct, then during simulated coalitional competition men will show testosterone responses similar to those found with competitive male–male challenge in other species (Wingfield, Hegner, Dufty, & Ball, 1990). Our second hypothesis is that this competitive testosterone response will be muted when men compete against members of their ingroup (Geary & Flinn, 2002, Wagner et al., 2002).

Intrasexual competition in mammals is commonly manifested as one-on-one male–male contests in the form of direct threat and physical aggression for the establishment of social dominance and through this priority access to females (Andersson, 1994). For some species, including our closest living relative (chimpanzees, Pan troglodytes), competition also occurs between groups of males (Mitani et al., 2002, Wrangham, 1999). As with other species, within-group male–male competition occurs for priority access to estrous females (Goodall, 1986), but is intricately constrained and followed up by reconciliatory behaviors (de Waal, 2000). Between-group competition, however, does not evoke these same constraints and is often lethal (Goodall, 1986, Watts et al., 2006).

In other words, males face different cost–benefit trade-offs when they must establish within-group status and simultaneously maintain within-group support to compete against outgroups. Unlike species in which male–male competition is only one on one, males who kill or seriously injure ingroup competitors may be disadvantaged when threatened by an outgroup. In this situation, mechanisms that blunt the intensity of ingroup competitive responses could evolve (Boehm, 1993).

To lessen the costs of circulating testosterone (Folstad & Karter, 1992), mechanisms have evolved to elevate this hormone only at points in the life span or in situations when the corresponding physical and behavioral changes are beneficial. The social challenge hypothesis of Wingfield et al. (1990) captures core aspects of how testosterone levels adaptively vary. The gist is that testosterone is low during periods of social stability and rises during challenges over control of reproduction-related resources, such as breeding territory or mates. For species with a distinct breeding season, the increase in testosterone and its metabolites at season onset is associated with a spike in male-on-male aggression, territorial and courtship behavior, and changes in secondary-sexual characteristics (e.g., Neal & Wade, 2007). Once territories are established and a mate is found, testosterone levels drop. When challenged by another male, testosterone increases to allow the defender to confront the challenge. Successfully confronting the challenge often results in a postcontest surge in testosterone.

Although men are prepared to respond to challenge throughout the year, because there is no distinct mating season in our species, they vary considerably in how and whether they respond to any specific challenge (Archer, 2006). The variability in men's responses to social challenge is related to individual differences in competitive striving (Mazur & Booth, 1998, Schultheiss et al., 1999) and to the many ways in which men can achieve status in human societies (Irons, 1979). A challenge and a victory to one man may not be so to another, depending on how closely the activity matches the man's interpretation of how the event influences position in his status-relevant niche. Laboratory and field studies are generally consistent with this view; the relation between testosterone and competition-related challenges depends on the relation between the event and status in niches that are important to men, on how men interpret the outcomes of the event (Cohen et al., 1996, Gonzalez-Bono et al., 1999, Salvador, 2005) and their social confidence (Maner, Miller, Schmidt, & Eckel, 2008).

With respect to status-relevant challenges, Mazur (1985) predicted a precompetition increase in testosterone to enhance assertiveness in seeking or maintaining status. He also predicted that competition winners will experience an increase in testosterone, encouraging further challenges and status seeking, and losers a decrease, encouraging withdrawal from further competition. Archer's (2006) meta-analysis of related studies revealed that men do indeed show small to moderate increases in testosterone in anticipation of a competitive sporting contest and small increases during the event. For sporting events, there is a tendency for winners' testosterone to increase and losers' to decrease. A similar pattern is found for laboratory tasks, although the postcompetition differences between winners and losers are larger than for physical contests; the physical exertion during sporting events may mask some of these effects.

Even though these studies have typically involved status-related competition, only a few of them involved the manipulation of social context or the nature of the competition in ways explicitly related to evolutionary predictions (e.g., Wagner et al., 2002). Wagner et al. assessed men's testosterone and cortisol during within-village and between-village dominoes competitions in Dominica; these matches are culturally important and can influence social status. Due to small sample size, the results were inconclusive with respect to testosterone, but there was a trend toward higher testosterone levels after the between-village match than after the within-village match. All of the individuals who played in both matches won their between-village match and thus it is not clear whether their elevated testosterone was due to their win or because of the differences in social context (i.e., ingroup vs. outgroup competition).

Although not framed from an evolutionary perspective, Gonzalez-Bono et al. (1999) provided a direct test of the hypothesis that between-group competition will result in postoutcome testosterone change. Here, men's testosterone was assessed before and after a professional basketball game in Spain. Men on winning teams did not show the predicted postgame increase in testosterone, but men who contributed heavily (i.e., ratio of time played to points scored) to their team's win did. Gonzalez-Bono et al. did not provide a corresponding within-group contrast, i.e., an assessment of whether scoring points during a within-team competition also results in a testosterone surge.

Cortisol modulates physical and mental states in preparation for and response to male–male competition and other social-evaluative events (Dickerson & Kemeny, 2004); specifically, the diversion of energy and blood to exercising muscle, inhibition of energy storage, inhibition of reproductive physiology, decreased appetite and a sharpening of cognition and attention (Sapolsky, Romero, & Munck, 2000). At moderate levels, cortisol enhances memory formation, by sharpening memory consolidation and retrieval (Abercrombie et al., 2006, McEwen & Sapolsky, 1995) and, in combination with testosterone, may facilitate the learning of competition-related competencies (Schultheiss et al., 2005).

In response to a rowing competition, Kivlighan, Granger, and Booth (2005) found that higher prematch cortisol was associated with more mental preparation, focus and competitiveness; inconsistent with Mazur's (1985) prediction, testosterone did not increase in anticipation of the competition and higher testosterone levels just before the competition were associated with poorer performance. In the abovementioned dominoes competition, Wagner et al. (2002) found that cortisol levels increased significantly before both matches and especially before the between-village one.

We used a video game to simulate within- and between-group competition in a controlled laboratory environment free of extraneous testosterone- and cortisol-eliciting stimuli. A video game is ideal for this type of study, because it is easily replicable and allows for control of the competitive (virtual) environment, which remains constant between conditions. It is appropriate to point out that a video game was used by Mazur, Susman, and Edelbrock (1997), but they concluded that the game (i.e., Pong) was not sufficiently engaging to replicate testosterone increases for winners nor was the game a simulation of male–male competition.

As noted, we predicted differences in testosterone response between tournament types. Based on the Gonzalez-Bono et al. (1999) finding, we did not expect a homogenous team win or loss effect for the between-group tournament (e.g., all members of the winning team increasing in testosterone) but did expect high-ranking players on winning teams to show a postwin increase in testosterone. For the within-group tournament, in comparison, we expected testosterone responses to be blunted. We were less certain about cortisol response, because both between- and within-group competitions have a social-evaluative component and thus both types of competition should result in cortisol increases (Dickerson & Kemeny, 2004).

Section snippets

Participants

The participants were 14 three-member teams of undergraduate men (n=42), with a mean age of 19 years (S.D.=.97). They were given class credit for participating and they competed for cash awards to make the outcome of the tournaments salient. For the between-group tournament, all members of the winning team were awarded $45 and members of the losing team $15. For the within-group tournament, the top-ranked player was awarded $45 and the two other players $15.

Video game

Unreal Tournament 2004 by Atari games

Results

Means and standard deviations for testosterone and cortisol are presented in Table 1; cortisol values were converted to nanomoles per liter to make them easier to read. Missing values — 3.96% and 2.82% for testosterone and cortisol, respectively — were replaced by the group means from the corresponding time of measurement. One team had incomplete data due to a lack of saliva and thus all of their data were discarded. The cortisol values were positively skewed, and therefore a square root

Discussion

We found qualified evidence for both of our hypotheses. The simulated male–male coalitional competition during video game play elicited a testosterone challenge response, but only when this between-group tournament was played before the within-group tournament. Men did not show a challenge response during the within-group tournament, whether assessed in terms of their contribution to the overall points scored or their within-group rank. If anything, higher ranking men had a muted testosterone

References (39)

  • OliveiraT. et al.

    Testosterone responsiveness to winning and losing experiences in female soccer players

    Psychoneuroendocrinology

    (2009)
  • SalvadorA.

    Coping with competitive situations in humans

    Neuroscience & Biobehavioral Reviews

    (2005)
  • SchultheissO. et al.

    Implicit power motivation moderates men's testosterone responses to imagined and real dominance success

    Hormones and Behavior

    (1999)
  • SchultheissO.C. et al.

    Effects of affiliation and power motivation arousal on salivary progesterone and testosterone

    Hormones and Behavior

    (2004)
  • WagnerJ.D. et al.

    Hormonal response to competition among male coalitions

    Evolution and Human Behavior

    (2002)
  • AlexanderR.D.

    Evolution of the human psyche

  • AndersonC.A. et al.

    Video games and aggressive thoughts, feelings, and behavior in the laboratory and in life

    Journal of Personality and Social Psychology

    (2000)
  • AnderssonM.

    Sexual selection

    (1994)
  • BoehmC.

    Egalitarian behavior and reverse dominance hierarchy

    Current Anthropology

    (1993)
  • Cited by (53)

    • Testosterone reactivity to competition and competitive endurance in men and women

      2020, Hormones and Behavior
      Citation Excerpt :

      Testing the influence of social presence on competitive behavior and hormonal relationships with competitive behavior is important because it is likely that these behaviors and related mechanisms evolved within the context of complex social groups (Crofoot and Wrangham, 2010; Flinn et al., 2012). Further, previous research has found that increased testosterone in response to competition is specific to between-group, but not within-group contests (Wagner et al., 2002; Oxford et al., 2010; Flinn et al., 2012), suggesting that testosterone reactivity to competition is attenuated in contexts in which within-group status dynamics are at play – where prosocial and cooperative behavior, in contrast to overt dominance, may also be required to simultaneously maintain within-group status (Diekhof et al., 2014; Reimers and Diekhof, 2015). In the present research, manipulating the physical presence of a competitor group was intended to shift competitive focus towards or away from the out-group by providing a relative reference for out-group when a competitor group was present.

    • The physiological and biochemical effects of gaming: A review

      2020, Environmental Research
      Citation Excerpt :

      Oxford et al. found that the most skilled player on a given team had higher cortisol levels after competing against teammates compared with competing with teammates against a team of strangers. The authors argued that it was an indication of male-male competition behavior in gamers (Oxford et al., 2010). Gray et al. had a group of 26 experienced League of Legend players (all male) compete against players from their own e-sports club or computer-controlled opponents to investigate stress response.

    • The impact of a competitive learning environment on hormonal and emotional stress responses and skill acquisition and expression in a medical student domain

      2019, Physiology and Behavior
      Citation Excerpt :

      Human competition promotes a complex set of hormonal and emotional responses linked to social status, bonding, affiliation, and reproductive success [1–4].

    View all citing articles on Scopus
    View full text