Under the gaming influence
Yesterday I blogged about this study that has been reported in the press as “racing games can make you drive recklessly in real life.” While that does seem to be a bit of a sensationalistic headline for the actual study, and while the study also seems to have some obvious limitations, I am nonetheless a bit surprised at the reaction among the gaming population.
Basically, I think that sometimes in our zeal to defend our favorite medium, we unwittingly rob it of its power.
The power of media lies in their ability to convey information. Often, it lies most specifically in their ability to convey information from particular perspectives: we value much nonfiction because it manages to present events dispassionately, whereas we other nonfiction because it so powerfully conveys a given experience through another person’s eyes.
With games, we have more: not only can we rely on the tricks of other media to accomplish this, but we also get to literally train the player’s perceptions in a way that is very direct. We get to build reflexes, we get to alter the player’s cognitive take on the flow of data, we get to construct mental models for the player to apply.
This is the core of gameplay, this building of mental models.
The whole point of mental models is that you can then apply them elsewhere. You leave the Holocaust movie having learned to empathize in a way you hadn’t before. You leave the history class with a sense of the broad sweep of repeated patterns. And in games, you leave with fresh understanding as well. What’s more, you can actually leave with those reflexes that you have trained up: actual physical and cognitive tools that you can then apply elsewhere.
In fact, were games not to have this effect, then they would truly be a waste of time, entertainment trifles. But instead, what we see is that games can help train doctors, assist in repairing brain injuries, and so on.
Here’s the thing though: if you accept that they have this power, then you must acknowledge that games can teach both good and bad habits.
And this is why I am not dismissive of studies like this racing one. I absolutely believe in the power of games to teach, and I know that I personally attribute elements of my driving to things I learned in racing games. So it’s not a stretch for me at all to say that were my teacher teaching me bad habits, it’s likely I would pick them up.
This doesn’t mean that racing games are bad — on the contrary, I think that most people are plenty smart enough to recognize the difference between a racing game and a freeway commute. But reflexes are reflexes. Once, when I was a teen I was mock-fighting with a buddy, and it was all good until some long-disused karate kick surfaced from the dim recesses of muscle memory and I laid him out with a mai geri to the nuts.
When a study like this appears, the questions we should be asking are not “is this valid?” but instead ones like these:
- Does the realism of the game affect how it teaches?
- How long does the teaching linger?
- Which lessons does the game teach most effectively?
- Which lessons does the game teach poorly?
- Which lessons are taught as direct reflexes, and how much does the control mechanism mimic reality? (I doubt that something that taught sharp swerves with a joystick would translate to sharp swerves with a steering wheel, for example).
- How do we build our games to teach the right things?
What I am not willing to do is to throw out the baby with the bathwater. Games have power, and power can be used for good or bad purposes, purposely or inadvertantly. The responsibility lies with us.
37 Responses to “Under the gaming influence”
Sorry, the comment form is closed at this time.
The Marriage – Mar 20, 2007 – Raph Under the gaming influence – Mar 20, 2007 – Raph FlowPlay, Maple money, and racing – Mar 19, 2007 – Raph The Sunday Song: This Don’t Groove – Mar 18, 2007 – Raph King Lud IC’s on fire
links diggados (digg.com/users/nandop) del.icio.us/nandop (bookmarks) » Raphâs Website » Under the gaming influence » Web 3.0 = (4C + P + VS) » “Incantations for Muggles: The Role of Ubiquitous Web 2.0 Technologies in Everyday Life” » Philosophia Perennis : HomePage » twittervision » Summoning Things Into Oneself
links diggados (digg.com/users/nandop) del.icio.us/nandop (bookmarks) » alpha (beta) » Raphâs Website » Under the gaming influence » Web 3.0 = (4C + P + VS) » “Incantations for Muggles: The Role of Ubiquitous Web 2.0 Technologies in Everyday Life” » Philosophia Perennis : HomePage » twittervision » Summoning Things Into Oneself
[…] Original post by Raph […]
[…] Under the gaming influence […]
Agreed wholeheartedly.
One thing that the study missed was that this transference goes both ways. For example, in NFS:Carbon i always brake before the sharp turns, a habit from RL driving that serves me very well. I also tend to follow the rules of the road when in Free Roam and thus avoid extra police attention although the fact that it’s all about driving on the wrong side of the road is irritating 😉
I disagree about the mechanism for turning, that’s about the only point we differ on. A racing game encourages you find a good line and follow it through a curve and that doesn’t change regardless of the mechanism for achieving it.
Once you’re as familiar with an XBox controller as you are with a steering wheel, they become essentially interchangeable. This is why F1 drivers can still drive street cars.
Hallelujah. Thank you.
Yehuda
I strongly disagree. Science is credible precisely because of the scientific method that requires peer review. Now, many people could hardly claim to be “peers” but scientists more often than not consider public perception just as they promote policies to support their research. When any research is published, the validity of the research should always be questioned, first. If there were no skepticism in science, as Ursula Goodenough once wrote to a colleague in a peer-reviewed journal, “The Sun would have taken to orbiting the Earth some time ago.”
Once the validity of the research is established, then we should allow the research to instigate further inquiry. We should allow the research to inform our analysis. Until the research is validated, however, the research is moot and should not inform decisions.
The questions you suggested are disconnected from the research. Those questions are what designers should be asking while they are designing a game or discussing when not. Those questions are what game critics should be asking anyway. But they are certainly not questions connected with the research in any way until the research is proven valid and can legitimately inform analysis.
Most particularly, just about none of us, particularly not those who commented in the threads on it. 🙂 I quite recognize that science is built on skepticism — what I am commenting on is the fact that our kneejerk reaction is to pooh pooh results we don’t like, or to immediately attack the validity of the results. That isn’t good science either.
I disagree with the notion that these questions are ones that cannot or should not be asked by researchers until the base study is proven valid. To my mind, these are questions that could be asked in the very first pass.
But that wasn’t what I was saying. What I’m saying is that the answers to the questions you suggested (as well as other questions) should not be assumed to be supported by research that has not yet been proven valid. Should researchers ask those questions? Sure. They should. But when research is published, should we question its validity? Yeah, definitely.
Immediately attacking the validity of the results may not be good science, but it’s certainly good feedback, especially if the attacks stem from questions about the research methodology. For example, the methodology used in this case had people play two games with different rulesets: one for racing, one for driving. The researchers discovered that those who played the racing game were more aggressive and more willing to perform riskier maneuvers in the driving game, but they concluded that people who play racing games are more aggressive and more willing to perform riskier maneuvers on the actual road. This is certainly not a trivial objection. I’m certain that their conclusion fits the description of at least one logical fallacy, and I’m sure that allowing unquestioned research to inform inquiry would also produce hasty and irrelevant answers.
[…] I’m still due for my post on general models to follow up my post on mental models. It’s mostly written, but I haven’t had any time to put into that kind of thought recently. However, to keep things moving in the ongoing side discussion on this blog about models, and to foreshadow the eventual tie-in with games (which I may have already started with my Digital Games and the Ultimate Tool post), I thought it would be fitting to introduce you to Raph Koster’s blog. Starting with his recent post Under the gaming influence. […]
I played racing games for years before I first drove a car, and I was rather delighted to discover that it was much easier to control a real car at normal driving speeds. Up until then, what racing games had taught me was that I was a horrible driver and that I was going to careen off the highway, slam into a wall, and flip over in a fiery wreck. 😛
…maybe this was why I avoided driving on the Beltway for years!
Does the teaching involved in play have to be literal? I’m guessing that a lot of the backlash I’ve been seeing lately about my recent article on the Columbine RPG is that the game is percieved to be straightfowardly teaching you how to be alienated and massacre your class-mates. What I’ve been argueing is that games can teach at both a content level and a subtextual level, you can teach a bad habit and then reflexively teach something about how such bad habits are learned. I know that sounds meta, and it is kinda, but I think there’s a lot of power there.
1: Games can teach both good and bad things.
2: Most games trivialize sex and violence.
3: Therefore, most games teach people to trivialize sex and violence.
4: Game designers make games.
5: Therefore, most game designers make games which teach people to trivialize sex and violence.
6: Trivialization of sex and violence is bad.
7: Therefore, game designers are evil.
8: Ban games immediately!
———————
Fix #2, then we’ll talk.
I can’t help myself, I thought that Woody’s take on the report was funny.
I have to agree with Raph. The question is worth asking. But at the same time, with the amount of studies that seem to come out and then are disputed (not to mention the studies that are created/funded by groups with their own agenda) does concern me.
As someone that does research and publishes in peer reviewed journals[although not on gaming] I’m sympathetic to the comments of both Raph and Morgan. I do have 1 quibble with Morgan though. You don’t prove things in science, you attempt to disprove things. That’s true for validity as well. There are subjective measure for evaluating validity [face and content validity] but subjective measures have obvious problems. Criterion and construct validation are more well received but they still don’t ‘prove’ validity. Strictly speaking, ‘proving’ only happens in the domain of formal systems such as formal logic and mathematics, not in science.
Why am I being pedantic about this?
“Once the validity of the research is established, then we should allow the research to instigate further inquiry. We should allow the research to inform our analysis. Until the research is validated, however, the research is moot and should not inform decisions.”
Since theories can never be proven, only disproven, the above is very problematic, imho.
So I guess this is why I got carded buying God of War II (best first hour of gameplay ever btw).
So do most movies, TV shows, and these days, quite a lot of songs. 😛
While I agree with this concern in general, I do think that as research comes in, we’re going to find that what games teach is much more on the cognitive level and much less on the “dressing” level.
I can’t resist putting in this quote from Galiel in the Terra Nova thread on Kiddification that elaborates on Raph’s “The responsibility lies with us.”
“Every affordance we facilitate, every behavior we inhibit or seek to thwart, is a moral choice we present to our gamers, and collectively this set of affordances and inhibitors constitute a moral universe in which we hope human players will invest a great deal of time and emotion.
Or is it the case that the ones who need to mature and take on the adult responsibility for their behavior are not our players, who only play according to the intrinsic rules of the worlds we design for them, but rather we as designers, who make the cultural choices that strongly affect their behavior?”
Popperian falsifiability is familiar to me. My grandfather was very much a scientist, one of my uncles is a top NOAA scientist, and several of my relatives are also scientists. I’m not ignorant of the ways of science, as I grew up with science as my religion.
I’m not suggesting that scientists and researchers attempt to prove their conclusions, and I’m not suggesting anything else about how scientists and researchers do their jobs. I am suggesting that consumers of scientific research hold off on applying research to their decision making until the research is no longer misleading, inconclusive, unfairly influenced by experimenter bias, or questionable because of sample bias.
Can you name a game that includes sex and/or violence, but does not trivialize them, in some fashion? I am undecided as to whether it’s possible. Some would argue that it is wholly impossible to do these things justice — that they are trivialized simply through the act of play. There’s some philosophical connection with drama (the other “play”) here, I feel.
Ah, yes, thank you Tom Stoppard. This is the idea I want to express:
Don’t get me wrong: We can treat these matters with dignity and respect. We can be very solemn about it all. We can lend them as much gravity as we are capable of in our respective arts. And yet, as our good friend Guildenstern points out above, it’s not quite real. That is a good thing. We don’t want our real lives to be violent. We don’t really want to die, or to be reminded of our own mortality, in the midst of our play. Play is difficult when one is overcome with existential angst. One might as well pack up one’s toys and go home.
Children — especially boys — love to play violent games, and have played violent games for as long as there have been children. Their games all trivialize violence. In all fairness, it is a far better thing for them to trivialize violence through play than it is for them to stop trivializing violence, by engaging in authentic violence. We might wish for them to simply do neither — to abstain from all manner of violence, play or otherwise. Yet, you would find it as easy as keeping a kitten from chasing a string. We learn survival skills through play, and so long as our species still faces the threat of real violence, violence will always be a part of our play.
In fact, some would argue that violent play becomes more important as our society becomes less violent, because it provides a useful outlet for those who no longer have a culturally acceptable channel for violent tendencies. Organized sports, likewise, take the place of tribal battles. It’s an interesting thought, but I’m still chewing on it a bit, before I know for certain what I think.
“I am suggesting that consumers of scientific research hold off on applying research to their decision making until the research is no longer misleading, inconclusive, unfairly influenced by experimenter bias, or questionable because of sample bias.”
Excellent, we’re on the same wavelength. The question is, how do consumers do that? In most cases [policy makers for example] don’t have the expertise to pull it off; they have other skills and do other things. One very common method [again very popular with policy makers] is to look at the evaluations made by people with those skills.
Here’s a technique so simple that even deans can do it.
1. Was a peer evaluation conducted?
Any schmoe can write a book, especially if he or she is willing to foot the bill and go to a ‘vanity’ press. The criteria used by publishers for nonfiction is pretty much ‘will it make some money’ not ‘is it even close to being true’. A peer reviewed outlet has people make an evaluation of the quality of the work [among other things].
“The study appeared in the Journal of Experimental Psychology: Applied, published by the American Psychological Association.”
This is a peer reviewed journal.
2. Was the peer evaluation conducted by high quality people?
Not all journals are of equal quality and have equal quality editorial boards and cadres of ad hoc reviewers. The first thing that springs out is that the journal in question is published by the dominant professional association of psychologists in the U.S. A good pedigree.
On a more quantitative note [and deans can’t read but they can count] there is journal impact factor, the metric typically used by policy wonks and deans. The Institute for Scientific Information journal impact report for 2005 [the most recent report] shows the journal in question has an impact factor of 1.293 and ranked 10th out of 49 journals devoted to applied psychology. Not the best but top 20%
The missing step here is and equivalent look at the researchers; pedigree, impact analysis etc. But lunch time is about over and I don’t have the time…
Here’s my quick take. A cursory but reasonably objective look suggests that the research is likely to pass the sniff test. The internet is full of trolls and idjits who won’t bother to look at the source of course but the people that hang out here are a better sort 😉
Re: JuJuTsu’s pairing of Raph’s quote with mine:
Raph is the Miles Davis (or Count Basie) of our industry:
A point that took me 103 words to express, he said in 5.
[…] think Raph Koster got it right here: https://www.raphkoster.com/2007/03/20/under-the-gaming-influence/ Rather than hide behind disingenuous claims that games have no impact upon the world (because if […]
Rubbish! Yes, operant conditioning works. HowEVER, what one learns from a video game is not flying, shooting, or driving skill, it is Videogame Playing Skill. Notice for example, in this very study, how the tendency to drive recklessly was tested. NOT on the highway, oh no. It was tested by observing the subjects behavior during ANOTHER VIDEO GAME. The proper conclusion of this study is that those who tend to drive recklessly during a video game, will also tend to drive recklessly during a video ‘simulation.’
Hrmph. I find the logic of this response to be, well, ugly.
It is the logic used, all the time, by people who want to rationalize a belief in something that isn’t justified. “Who cares if X is valid, it’s interesting to me and it might have been valid therefore I say we keep talking about it anyway.”
No. This isn’t about whether they should consider questions like this. Obviously such questions are being considered. It’s about whether we should allow statements of scientific results to pass on the basis of “that sounds right” instead of “that’s a proper and well-tested result of scientific merit”.
I am published, I have peer-reviewed, I have studied cognitive science and from what I have read of this study I do call “shenanigans”. There are obvious causation/correlation issues at stake (as one poster put it: “this just in, playing Madden causes you to get drunk more frequently on Sunday afternoons”) and the methodology of the game vs. simulation has obvious flaws.
None of this says that we can’t talk about how game playing impacts congnitive development/structure. It just says that poorly-executed science IS poorly-executed science and not the sort of stuff that should get public attention. If we want to have a conversation about games and learning then we could do that and we can leave the poorly done studies aside.
I am not defending this study — I am saying that all too often, the kneejerk reaction is “no possible study would ever be valid,” and the gaming community dismisses all research with facile comments about how it’s just a game.
While there have been plenty of critiques of the study on the basis of questions about methodology, the majority of the commentary I have seen have been more on the order of sticking one’s fingers in one’s ears and shouting “la la la.” We just don’t want to hear it, in large part because we as a community are hypersensitive to criticism precisely because we’re under so much attack.
First of all, that’s not what you are saying with this:
You are saying there that it doesn’t matter if the study is valid. Of course it matters! This is just the sort of rationalization used by smart people the world round to say: well it may be wrong but I want to talk about it as though it were right anyway.
Secondly, I don’t hear the “la la la”. Quotes from “disagreement” you cited:
While none of these people likely have PhD’s they are, all the same, giving valid criticism of the study and seem to be making an effort to engage the actual merits of the study and not “put their fingers in their ears.” There have been a lot of poorly done studies on video games and so it’s not like taking a critical view from the outset is unwarranted.
The study is available at
http://www.apa.org/journals/releases/xap13122.pdf
As on of StGabe’s quotes point out, what the press says and what the study says are not exactly the same.
@StGabe
What is ‘invalid’ about the research?
@Raph
“We get to build reflexes, we get to alter the player’s cognitive take on the flow of data, we get to construct mental models for the player to apply.”
Where is the valid experimental research to back up that claim? 😉
Uh, all over the place. Didn’t I just point to yet another study about doctors improving in surgery after playing games? I cite several studies along these lines in the book; my favorite is the one about mental rotation of 3d objects. There are also places where it is just blatantly obvious: typing games, for example, for the reflexes; and mental models, sheesh, all over the place.
These days, we not only regularly see negative studies, we also see positive ones pretty consistently. I don’t see people from the gaming community jumping up to attack the methodology on them, either.
I am surprised that StGabe doesn’t see the “la la la.” I see it all the time.
Sorry Raph, that surgery study won’t cut it by gamer research standards. It’s not based on an experimental design. It only has 33 nonrandomly selected subjects. It’s primary findings are “Past video game play in excess of 3 h/wk correlated with 37% fewer errors (P
hmm format issues I guess
Sorry Raph, that surgery study won’t cut it by gamer research standards.
It’s not based on an experimental design. It only has 33 nonrandomly selected subjects. It’s primary findings are “Past video game play in excess of 3 h/wk correlated with 37% fewer errors (P less than .02) and 27% faster completion (P less than .03).” so “There are obvious causation/correlation issues at stake”. The regression analysis is shakey with only 33 degrees of freedom…clear problems with statistical power.
I’m forced to conclude la la la
Ok no more on this thread, I promise to behave.
As I said, that was “yet another” study — I didn’t even read that one, to be honest. There have already been several studies on the medical side starting a few years ago. I would have to go dig up the actual references…
First of all, the conclusion of the study is:
(a) and (b) aren’t that controversial and actually I think that you’ll even find several gamers in the link that Raph posted in his first post agreeing with these (which further leads me to be confounded by the talk of “la la la”). (c) is the conclusion that seems to be largely unfounded or “invalid”. Not that the study also goes on from (c) to make suggestions for public policy including age restrictions on driving games.
There are three studies.
Study 1: finds correlation between playing driving games, owning a powerful car, wishing to express “power” when driving, engaging in risky driving situations and having auto accidents.
Largely a useless result. Yes, people who like to have powerful cars and enjoy risky driving also have a tendency to like driving games that let them drive powerful cars in risky situations. Looking at their procedures they are doing lots of fancy stuff but nothing that would lead us from correlation to causation.
Study 2: finds that playing a driving game that rewards risk tends to prime someone for word associations that skew towards “risk”.
Interesting but also largely useless as a result. Reading a book about NASCAR drivers could easily have the same effect. Thinking about risk or having risk narrative associations in your head has in no way been proven to cause you to go out and perform risky acts. This sort of fallacy is at the crux of a lot of bad video game literature. This is a very weak result but it is fruit for justifying all sorts of strong judgments and ideas.
Study 3: finds that playing a driving game that rewards risk tends to prime someone to then play a driving simulation as though it rewarded risk.
Again interesting but largely useless. As has been pointed out all over the place you are really just having the player play two driving games in a row in a way that will provide confusion as to what their goal is.
A study that would be valid would be to take take separate, randomly chosen groups of people, have them play the neutral or driving games on a regular basis over a period of time and plot their driving records in the time following their exposure to driving games. Of course such a study is very unlikely to yield exciting results.
From Slashdot….
A study mentioned in the Syndey Morning Herald suggests that playing a videogame with violent content generally has no affect on most kids.
http://www.smh.com.au/news/National/Most-kids-unaffected-by-violent-games/2007/04/01/1175366055463.html
Can I expect the gaming community to start savaging the research methodology? 😉
[…] steps to keep sleazy, spineless retailers from selling it to anyone with a pulse.That being said Koster pretty much nails this one, and his conclusion sounds unusually conservative. Not saying Raph is gonna go out and vote […]