P1: GG! P2: Bad Game: losing, trash talking and aggression (Breuer et al., in press)

Last summer, I played League of Legends and I immediately got hooked, but as I progressed I encountered more toxic players, they became increasingly hostile when they realized that they would be losing and then the blame game goes around. That’s when I stopped, I still have the game installed, but never updated it. Johannes Breuer (University of Münster), Michael Scharkow (University of Hohenheim), and Thorsten Quandt (University of Münster) published a study in Psychology of Popular Media Culture regarding how participants behave when they lose or win.


The impact of video game play on player aggression continues to be debated within the academic literature. Most of the studies in this area have focused on game content as the independent variable, whereas the social context of gaming is largely neglected. This article presents an experimental study (N = 76) on the effects of game outcome and trash-talking in a competitive colocated multiplayer sports video game on aggressive behavior. The results indicate that an unfavorable outcome (i.e., losing) can increase postgame aggression, whereas trash-talking by the opponent had no such effect. We also tested the frustration– aggression hypothesis for video games and found that the effect of losing on aggressive behavior is mediated by negative affect. The results suggest that the frustration–aggression hypothesis can be applied to the use of digital games and that game characteristics alone are not sufficient to explain effects on aggression.

I have uploaded an excel sheet regarding the number and type of articles I have collected this year, I update it every Thursday.

Much of the research on violent videogames focused on the content of the videogame, the authors noted, however, the social context of videogames has not received much attention. This is understandable as the internet a little more than decade ago was not as fast or integrated with our social lives. Today, the third author edited a book, Multiplayer: The Social Aspects of Digital Gaming, on areas that we currently know and don’t know (dibs on chapter 19). A good number of experiments examined two-person play affected aggression, cooperation, competition, enjoyment, etc.

The authors sought to examine how winning or losing in a competitive game would affect aggressive behavior as explained by the frustration-aggression hypothesis. I believe the frustration hypothesis is the most popular among gamers in explaining why they get angry, trashed their computer or any other aggressive acts. The authors noted that frustration, as proposed by the original theorist John Dollard, is “an event or action that complicated or obviates the accomplishment of a subjective relevant task”. The negative emotions from such event is often misattributed as frustration which is actually the effect of being prevented from accomplishing your goal. These negative emotions can lead to aggressive behaviours.

A source of frustration is competing with another for the same thing, a win can only be given to one party. So, losing is “not an option” for the poor players as they are frustrated and would feel quite badly about the results, perhaps they would lash out. Besides the competition, players can push other players’ emotional buttons, it can go positive and/or negative, so trash talking might push other players to feel negatively and increase aggression.


Participants: 76 participants from a German university. Originally, they had 91 participants, but 15 were dropped because of technical issues and/or suspiciousness about the study. The 76 participants consisted of 48 women and 28 men. Average age is 22.6 (SD=3.2) and 63% of them were university students. They also received 10€ compensation for their time.


Negative affect: measure for negative emotions using 4-items from the PANAS. There are angry, irritated, ashamed and frustrated.

Aggressive behaviour: The competitive reaction time task. The basics of the task has a participant to compete how fast they can press button when given a signal. They are partnered with an ostensible partner. The losing party would be punished by a loud noise set by the winning party. The loudness and duration is also set by the winning party. The authors created their own version of the task citing various criticisms from other researchers. Their version consisted of having the participant to compete with their partner (who played with them earlier), the participant set how long the noise will be, the loudness is set at the same volume throughout the task. The original task consisted of 25 trials, theirs consisted of 10 trials. As with other studies, their analysis of the data will only examine the first trial.

Videogame used: FIFA World Cup 2010 on the Xbox 360. The game is non-violent, but very competitive and Germans are pretty big fans of soccer. Play time is set for two matches of 5 minutes, total play time is 10 minutes. Also they all played the same German national team to avoid any gameplay imbalances.


Participants were asked if they played FIFA before, if not they are given a 5 minute tutorial. Afterwards, participants are randomly assigned to play with a confederate (their only one and is trained with the game) who will either win or lose the game and who will trash talk or be helpful with the participant. After play time, they completed some questionnaire, including the negative affect scale, then they went into the competitive reaction time task with the confederate, both the participant and the confederate complete the task in separate rooms, so they would not peek at each other.

The trash talking is described, by the authors, as moderate which consisted of commenting sarcastically, or saying “Oh! This is going to be easy” whenever he scores. This is done to protect the confederate from physical harm. I’ve spoken with the second author when he visited Ohio State a few years ago, he mentioned in passing about how Germans were quite polite and were not used to insulting each other.


The manipulation checks checked out, participants in the lose condition felt more overstrained and felt less superior than the confederate. Participants in the trash talk condition felt less sympathetic to the confederate and less interested in playing with him in the future.

Their first ANOVA revealed that those in the lose condition were significantly more aggressive (M=3.3, SD = 2.1) than those in the win condition (M = 2.4, SD = 1.8), slightly that is. Statistically speaking, there were no difference in aggression for the trash talking (M=3.1, SD=2.0) versus helpful (M = 2.7, SD 2.0) conditions. No interaction effects were found.

To test whether the frustration hypothesis explained aggression, they conducted a structural equation modeling. Participants’ sex and video game play per week were entered as covariates. The results revealed that losing predicted negative emotions, trash talking, however was not a significant predictor. The effect of negative emotions on aggression is significant, but small. The indirect effect of losing on aggression is significant as mediated by negative emotions.


The take home message is that when you lose in a competitive non-violent game with another player, you’re more like to be aggressive. Between the Germans, the trash talking isn’t much of a factor in predicting negative emotions or aggression. Of course, we don’t want the confederate to be too harsh lest they face the potential wrath of a participant.

The authors proposed that game content, game mode and social interactions operate through different psychological mechanism. Aggressive social interactions might increase aggression through negative emotions, or being helpful increases prosocial behaviors. Game content might prime or bring up aggressive concepts into one’s consciousness and prosocial content bring up prosocial thoughts. Another alternative is excitation-transfer theory which posited that arousal leads to aggression. They argued that losing might more arousing, although they don’t know why.

I am not familiar with the frustration-aggression hypothesis, but I gather that it might generate other outcomes. As the frustration event elicit negative emotions and aggressive cognition, is it possible that it might elicit some cognitive biases? One cognitive bias is whether frustration can cause the losing player to rationalize their loss in order to avoid cognitive dissonance, they might try to blame or rationalize that the other player might be a cheater, overly powerful, playing unfairly.

Some limitations are the low strength of the trash talking, using a sports game that is non-violent, yet competitive, their operationalization of frustration, their oeprationalization of aggressive behavior. The authors argued that future studies should investigate and disentangle the relationships between aggressive emotions, thoughts and behaviours. Varying the skill level of the confederate between being incompetent to superior,, varying the violent content, comparing cooperative and competitive play, among other things.

The study generated some thoughts I’d like to share. I happened to read an article by Dominique Muller and colleagues where they found that winners were more aggressive to their loser partners. The studies differ on many aspects, but they share some aspects, such as competition and aggressive behaviour based on whether they won or lost. Muller and colleagues’ theoretical arguments are based on social power, when you win you feel powerful which can be disinhibitive to do things you’d normally won’t do because of consequences. So, when you win you feel that you can send the middle finger, gloat or humiliate the losers. There is anecdotal evidence where some players trash talk even more harshly or were more aggressive when they won a match or scored a kill. But this is something for a survey. I also happened to found a recent article on the psychology of competition if anyone is interested (I haven’t read it).

Most experimental studies, so far, have been dyadic and co-located (i.e., two players and physically close). My impression of the literature is that there are very few studies that examined computer-mediated dyadic play (i.e. play online with another player or separate rooms). Having a CMC game play might be enough to have a more foul-mouthed confederate to trash talk the participants and is more naturalistic as trash talking occurs more online than offline (but there is no empirical basis in that argument). Second, if the participant do get angry, at least, there is some physical barriers in reaching the confederate. Or we can do a field experiment and see how players react to trash talking players, such as reciprocation, banning, ignoring, etc. (Better talk to James Ivory, if there are any data).

Another methodological concern is the choice of reciprocating for the participants. The participants are given a choice to act aggressive, nothing else. I forgot the citations, but prosocial and aggressive behaviours are not necessarily on the same spectrum. What if they are both aggressive and yet prosocial? The competitive reaction time task can be reconfigured for prosocial behaviours, ask Dr. Bushman. On a more naturalistic and social interaction side of things, players can call “GG” to show their appreciation for good sportsmanship and being courteous to each other or follow-up with more hateful messages. This brings up another study on sportsmanship and courtesy, how players behave if they are told to end a game courteously (e.g., shaking hands), harshly (e.g, flipping the bird?) or silently. I suppose ending a game on a polite note might inhibit aggressive tendencies for a bit. I believe Jeffrey Lin of Riot Games may have something about that.

Other behaviours to consider on what players can do online are revenge seeking (The blue shell from Mario Kart), ragequiting, cheating, or calling the other player a cheat. I suppose we can do that in an experimental setting.

Breuer, J., Scharkow, M., & Quandt, T. (in press). Sore losers? a reexamination of the Frustration–Aggression hypothesis for colocated video game play. Psychology of Popular Media Culture. DOI: 10.1037/ppm0000020

2 thoughts on “P1: GG! P2: Bad Game: losing, trash talking and aggression (Breuer et al., in press)

  1. A very good summary of our study and some useful comments on its limitations and opportunities for further research in this area. Thx! Indeed, we recently ran a second study on gaming and frustration in which we looked at prosocial behavior as the DV (using a modified prisoner’s dilemma task). And more follow-up studies are in the making…
    There’s also another recent publication by Przybylski et al. on gaming and frustration in the JPSP that I can strongly recommend to anyone interested in this topic: http://psycnet.apa.org/psycinfo/2013-45445-001/

  2. Pingback: The 2014 International Communication Association Annual Conference – Day 3 | VG Researcher

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s