Psychologists Should Use Video Game Data

Summary: Role playing video games have become more and more realistic and complex over the past several decades. These games have unique attributes that make them particularly and potentially useful for study, including points at which major emotional and cognitive decisions are made by the player, ease of data extraction, and diverse sample size. Psychologists, behavioral economists, and sociologists should consider investigating these ever more popular games as they provide a source of enormous and relatively high quality data.

The relationship between popular media, psychology, and video games have been rather tumultuous. Seeking to either identify a relatively new and salient variable in our society, or to divert attention from deeper or more contentious factors, pundits often claim that video games have caused destructive behavior in our society – including mass shootings. The research has not really supported that conclusion. Some studies have connected various abilities with gaming, including but not limited to: creativity, cognitive development, persistence and motivation, health, spatial awareness, hand-eye coordination, reaction speed, emotional regulation, and social competence.[1]

When one thinks of video games in our day and age, one typically thinks of the popular online ones that are continually played. In 2019, the most popular online games are almost all first person shooters as identified below:

Clearly shooter-type games dominate the scene, which affects the most salient image of a video game (the availability heuristic in action). In turn, because there is violence therein, it is simple [yet fallacious] to suggest that violence in society and video games are causally correlated. And yet, a whole other set of video games exist that, in my opinion, could yield fascinating troves of data.

Let us step away from the media’s focus on first person shooters and redirect our attention to a wholly different sort of video game that has developed incredibly in the past several decades: role playing games (RPG’s). Note that some RPG’s hold a spot in the lists above. An RPG features a typically elaborate fictional story in which players assume the role of a character. In a sense, an RPG is a way of playing out a story. As time passes, some RPG’s – such as Mafia II/III – have become more akin to playing a role in a movie than a pixelated video game, they’ve become extremely realistic. In many of these games, the character is often given the opportunity to make complex decisions. I argue that these decisions can provide an extremely large repository of data for psychologists. I will briefly discuss two specific examples from CD Projekt Red’s The Witcher 3 (2015) to highlight some of the available data for study.

In The Witcher 3 (forthwith The Witcher), money has an important exchange value used to upgrade and purchase equipment and consumables; money is the standard reward for completing tasks and quests. There is a quest in the game in which the main character (Geralt) fist fights his way across the entire game’s world to become the preeminent world champion. Periodically, an opponent approaches Geralt outside the ring, before the fight, and asks him to throw the fight for some personal reason. The opponent offers Geralt an equivalent amount of money, which is not a paltry sum. The minute the first fight is lost, the fight can be re-prompted and won, and the overall quest will continue unimpeded. Losing the fight has no negative effect save the 20-30 seconds or so that it takes to actually lose. Furthermore, the reason the opponent requests help is typically deeply sympathetic, for instance a fighter named “Stan Fishgulper” (a fisherman) needs the nonmonetary reward of the fight to feed his family. The point is that in about a minute, Geralt can, risk-free, double his income with no negative repercussions and performing a potentially kind act. And yet, nevertheless, when I played I often found myself denying these kinds of requests! Was it pride? Was the perception of lost time too great? Did my perception of the monetary gain decrease based on the amount of money I currently held affect my decision (which would of course be expected according to microeconomic theory – and yet I found myself throwing more fights later in the game because the items I wanted to purchase had radically increased in price, commensurate with my total wealth)? These are the personal questions for me, but perhaps I am not an outlier. What is the proportion of people who throw the fight? Is there a correlation with how long they’ve played the game generally, or how long they’ve been playing for that specific session? Does their actual geographic location affect their decision process? (i.e. does an American throw less fights than a Belgian?) Do male gamers behave differently than female gamers? This unexhaustive list of questions is reflective of highly important behavioral economic of psychology questions already being researched.

My second example from The Witcher is a bit more nuanced. Geralt runs into a friend and colleague named Lambert, who asks for your help to kill a merchant named “Jad Karadin.” Jad was once an assassin and Lambert believes that Jad killed one of Lambert’s close friend (and perhaps lover). When you find Jad, he has become a merchant and has begun a new life with his wife and adopted children, and that Lambert’s friend’s death was an enormous accident that plagues his consciousness – in an interaction somewhat akin to the confrontation with the evangelist Alexander Jerrod at the end of the movie Faster (2010). Here is the dilemma: Lambert really wants your help to kill Jad, no matter the truth. Lambert is an important character, is an in-group individual whose opinion matters (and one does not want to generally create dissonance with ones friends, you also require Lambert’s aid yourself, resulting in an appearance of a case of reciprocity), and you know that killing Jad will result in a substantial reward. On the flip-side, killing Jad results in no negative repercussions whatsoever and appears to improve your relationship with Lambert, Jad is a neutral character, and he offers no rewards for his life. But, killing him is morally wrong. So who chose to kill Jad? How many were purely utilitarian and are there any other factors that contributed to this decision?

Limitations

These two examples highlight the richness and depth of available information that could be gathered. There are however several major limitations to using these games, the three most salient are: cheating, initial sample, and lack of reality. I will briefly discuss these three in this paragraph.

Cheating: cheating is potentially a major issue as it would really warp how people played their games. If one knows prima facie what will happen ex post facto, it will inevitably change how one behaves. In the second example with Jad for example, learning that siding against Lambert with Jad might be more probable if you knew that it would not affect your long term relationship with Lambert. However at the same time, I somehow doubt that people often cheat in these RPG’s during their initial play through; an experiment or set of experiments could be run to determine this and adjust for this bias. In particular it is especially unlikely that one would cheat when it came to seemingly insignificant encounters that might be the most profitable source of experimental data; if it appears that the outcome of an action is unimportant, I am unlikely to go online to find out which path is the right one to take.

Constrained Sample: Not everyone who can access a video game actually plays a video game. I know, for instance, of my friends growing up, the girls rarely played video games – and never RPG’s. Access to video games also constrains the sample further, for the cost, free time, and infrastructure (i.e. electrical costs) are all substantial barriers to individuals across the world. The sample size would be overwhelmingly tilted towards rich and developed countries. And yet while the sample is skewed, it is still representative of a large swathe of the world: the Xbox One is directly sold in over 40 markets, while the PlayStation is sold in over 70. Over 40 million Xbox’s have been sold, while over 100 million PlayStation 4 items have been sold.[2] That is still hugely better than many experiments currently conducted on tiny samples of WEIRD people (Western, educated, from industrialized, rich, and democratic countries; oftentimes US undergraduate psychology majors doing experiments for class credit).[3] Given how many psychology experiments have been refuted or whose results were found difficult to replicate, increasing the sample size available in studies is a great first step.[4]

Unreal: Lastly, a massive apparent reason is that the video game is not realistic, it’s a video game! The repercussions that occur to your character in a game are simply not equivalent to repercussions in real life. Almost anyone who has played Grand Theft Auto, for example, randomly will begin to wreck havoc to see how many police assets (increasing your “Wanted Level”) they’re able to draw unto themselves before being eliminated. At the same time however, I am not proposing that video games are representative of real life generally, I am proposing that there are specific instances and decision points that should be analyzed. Those decisions in RPG’s are often incredibly realistic and nuanced; the closeness one feels with one’s character and the time dedicated to the game thus far means that one is heavily invested and decisions do carry a lot of weight. Even in video games with morally tainted main characters, one feels a strong connection to that character – just like audiences cherished rather disreputable actors in shows like Breaking Bad, Mad Men, or The Sopranos. By the end of both Red Dead Redemption’s for example, I would be shocked if someone did not feel a deep affinity towards John Marston and Arthur Morgan (the main characters for the first and second games, respectively), even though aspects of their character are rather wicked and vile.

Benefits

There are enormous benefits to leveraging video game data that have not yet been discussed. I’ve already mentioned that video games would provide voluminous and comparatively diverse data (not only WEIRD samples). To this list I also wish to add three elements that bypass major limitations in many psychology studies.

Response Bias: this is a general term stating that a subject in an experiment might answer or act in an untruthful manner, in a way that is not reflective of their true selves. People often do not actual report what they believe or feel for a wide variety of reasons, including but not limited to a desire to conform to societal expectations, the choice architecture of questions, acquiescence bias (just preferring to say yes) , and – above all – demand characteristics (altering responses because one is part of an experiment). But in a video game, those factors diminish or disappear, there is no authority figure standing right above you, you don’t actually see any Research Assistants with their checklists or the (real or perceived) whisper of judgment in their eyes. Video games are also so carefully crafted to create a good user experience that there is little cognitive load associated with decision-making, options are clear, and it is extremely easy to go with your gut.

Realism: In stark contrast to my last point as a major limitations, many modern video games are actually often quite realistic, especially when compared to lab settings. There is a major difference between sitting in a shiny clean lab to answer questions, and actually being in the real world making those decisions. In famous dictator games, for instance, several decades of experiments were invalidated when researchers requested subjects perform work before allocating or appropriating any resources. They found that if one works to earn resources, one is unlikely to give those resources – or to take them from another party. In their attempt to replicate real life, video games have embedded many of these factors. In The Witcher or Red Dead Redemption, for instance, you work for your money, and it can take considerable time and effort to accumulate any wealth.

Cheap: Conducting a laboratory experiment is enormously expensive. The workers and test subjects must be paid or rewarded, the fixed and variable costs of maintaining a physical location must be borne, and the variable fees of the experiment must be considered. Logistically, extracting data from a game is a marvel: a snippet of code can be added and transmitted online since modern stations are connected to the internet. All that would be necessary is a repository and data analysis, an overall cost far less than the current tens of thousands of dollars requisite in classic psychology experiments.

Conclusion

Millions of experiments, however violent and convoluted they might be, have been created and are performed every day. They are just at our fingertips. Why not look at what the live experimental data is actually saying?

[1] See for instance, Bediou, B., et al., “Meta-analysis of action video game impact on perceptual, attentional, and cognitive skills,” Psychological Bulletin (2018), 44, pp. 77-110; Ewoldsen, D. R., et al., “Effect of playing violent video games cooperatively or competitively on subsequent cooperative behavior,” Cyberpsychology, Behavior, and Social Networking, (2012) 15, pp. 1-4; American Psychological Association, Resolution on Violent Video Games (2015)

[2] Matt Peckham, “Xbox One to Launch in 29 New Markets: Why It Matters,” Time, September 3, 2014, https://time.com/3262663/xbox-one-markets/, Accessed 11/17/2019.Derek Strickland, “Microsoft’s family of Xbox hardware has sold about 41 million units globally,” Tweaktown, January 24, 2019, https://www.tweaktown.com/news/64602/xbox-one-sales-hit-41-million/index.html, Accessed 11/17/2019; Tom Warren, “Sony has sold 100 million PS4s,” The Verge, July 30, 2019, https://www.theverge.com/2019/7/30/20746712/sony-playstation-4-sales-100-million-milestone, Accessed 11/17/2019.

[3] Bethany Brookshire, “Psychology is WEIRD,” Slate, May 8, 2013, https://slate.com/technology/2013/05/weird-psychology-social-science-researchers-rely-too-much-on-western-college-students.html, Accessed 11/17/2019; Eric Michael Johnson, “the WEIRD Evolution of Human Psychology,” Scientific American, December 7, 2011, “https://blogs.scientificamerican.com/primate-diaries/the-weird-evolution-of-human-psychology/, Accessed 11/17/2019.

[4] Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), aac4716. Doi: 10.1126/science.aac4716