The Psychology of Video Games

Why did Activision take an already expensive game and release an even MORE expensive version without adding a whole lot to it? And what does it have to do with “Ozark wild mushrooms served with a brandy demi glaze?” I’ve got an idea. Let me share it with you.

As you may know, DJ Hero, a relatively new rhythm game from Activision and FreeStyleGames, includes its own controller in the form of a fake plastic turntable. Many gamers thought that the initial price tag of $120 was high, but were outright boggled by the $200 –TWO HUNDRED DOLLAR!– list price on the special “Renegade” collector’s edition that only had a few paltry perks relative to the base model. Why would Activision take an already expensive game and put out an even MORE expensive version?¹

$200? Ehh....

They’re stupid? Nope. According to one recent press release, DJ Hero was 2009′s highest grossing new video game IP. So something worked.² At the very least, the Renegade Edition pricing is an example of what economists call “price targeting” and what author Tim Harford likens to getting turkeys to vote in favor of thanksgiving.³ In essence, Activision is putting the Renegade Edition out there so that people who are cavalier about price self-identify themselves and allow themselves to be sold basically the same product for more money. It’s the same trick restaurants use to find patrons willing to pay more for food when they charge an extra $.80 for a slice of cheese on your burger when it really only costs them a few cents.

But I don’t think that’s the whole story. I think there’s something called “the contrast effect” at work, and I think Eminim and Jay-Z would be appalled.

You may remember the contrast effect from my discussion how the cover art for Borderlands might have gotten approved. Basically it’s a cognitive bias that kicks in when focusing on the magnitude of one bit of information affects your perception of another piece of information. Like a colored tie appearing brighter when it’s contrasted against a white shirt. Here’s an extremely simple example where the same color of gray looks lighter or darker depending on the background it’s contrasted against:

Which circle is lighter? Trick question! They're the same.

This is why I think the Renegade Edition of DJ Hero was put out there: to activate the contrast effect and to make the regular, $120 edition look cheaper in comparison. Man, $200 for a video game? Forget that. I’m gonna be a the smart shopper and only buy the $120 version. What restraint I have! Quick! Somebody congratulate me!

This kind of thing is done to you EVERYWHERE around you, as the following quote from a New York Times story on the science of restaurant menu writing illustrates:

Some restaurants use what researchers call decoys. For example, they may place a really expensive item at the top of the menu, so that other dishes look more reasonably priced; research shows that diners tend to order neither the most nor least expensive items, drifting toward the middle. Or restaurants might play up a profitable dish by using more appetizing adjectives and placing it next to a less profitable dish with less description so the contrast entices the diner to order the profitable dish.”

Those “Ozark wild mushrooms served with a brandy demi glaze” may be the most expensive side dish listed, but the restaurant only listed them first so that the cheaper mashed potatoes, which have a higher profit margin, look more appealing.

If you’re like me, you see attempts at the anchoring effect everywhere once you know about it. Now that you know about it, think back on the last time you saw a regular version of a game advertised next to the collector’s edition. Does the regular one seem so cheap now?

1. DJ Hero Renegade photo credit: j.reed on Flickr.com [↩]
2. But as Joystiq points out, that’s not the same as “most profitable” [↩]
3. Harford, T. (2006). The Undercover Economist. Oxford: University Press. [↩]

I wrote just the other day about how loss aversion could be used to increase conversion rates on trial games. You can read that article for more details and a neat experiment illustrating the effect, but the gist of it is that people hate to lose things more than they like to gain them. Losing $10 is more painful than gaining $10 is pleasurable because “losses loom larger than gains.”

I kind of hate to dip back into the loss aversion well so soon, but Funcom recently provided such a textbook example that I couldn’t resist. Many players who had unsubscribed from the Age of Conan massively multiplayer game got an e-mail from the publisher stating, in part:

Dear customer,

Thank you for playing Age of Conan.

As part of our maintenance your account is now flagged to have your characters below level 20 deleted as part of maintenance. Please re-activate your account now to ensure that your characters progress and names stay intact.

In other words, “come back or your low level alt¹ gets taken out back and shot.”

A Funcom database administrator gets ready to subject your character to "maintenance."

I’d be fascinated to see what this did to Age of Conan’s resubscription rate. If I were in charge of these things at Funcom, I would have randomly separated that mailing list into two groups and sent the above e-mail to the first half. The second half would have gotten something along the lines of:

Dear customer,

Thank you for playing Age of Conan.

As part of our maintenance your account is now flagged to have your characters below level 20 saved as part of maintenance if you resubscribe. Please re-activate your account now to ensure that your characters progress and names stay intact.

And then I would have looked at the differences in resubscription rates between those whose message was phrased in terms of losing their character and those whose message talked about saving it. Which of those two messages would you, as a MMO player, respond to more strongly? My guess would be the former, especially if you weren’t the handsome and well educated person you are on account of reading about loss aversion here.

Note: A combination of this and my previous post on loss aversion appeared on GameSetWatch. Look for more of my writing to appear there in the future!

1. not to mention your bank and your mule characters [↩]

One day when going to check on my friends’ status updates on Facebook, this jumped out at me:

Attention Facebook friends: Please for the love of God stop sending me gifts and invites for Farmville, Mafia Wars, Vampires, and whatever other crappy THING you’ve been playing. DO NOT WANT. Just …STOP. GOD.

Those of you on Facebook or MySpace can probably sympathize. How many times have you checked your notifications and thought “Gee, you sent me a …virtual goat. THANKS. I guess. Guess I should click on your link.” Indeed, developers of these social games have gone to great lengths to make “gifting” of imaginary stuff a core element to the gameplay, and they’re even starting to offer pixilated gifts for real money. Why is that? Why do people do that?

The answer has to do with one of the most powerful habits in social psychology: the reciprocity effect. When people give you something, you feel the need to give something back; it’s that simple. Or possibly if you’re like my friend quoted above, you yell at them. But usually you want to reciprocate. Some evolutionary psychologists think that this is an evolutionary advantage in that it encourages societies to form –and enforce– mutually beneficial norms. Adhering to the norm is seen as a good deed, and others want to return that deed; breaking the norm is an attack, and will earn you a misdeed in return, like shunning or a punch to the neck.

Thank you for the sheep, Guy I Knew in High School.

The reciprocity effect is put to use by marketers and savvy businesspeople all the time. Every year the March of Dimes charity sends me a lovely set of return address labels for use with my Christmas cards¹. The labels are a free gift, but not coincidentally, they come in the same envelope as a plea to donate. The message is clear: “Dude, we totally just gave you some free stuff. You should return the favor with a donation.” Psychologist Robert Cialdini explained in a 2001 article in Scientific American how the Disabled American Veterans organization used this same trick to increase the success rate of their appeals for donations from 18% to 35%.²

The same technique is used by supermarkets giving you free samples of new cheese crackers, or the video game developer who gives out free tee shirts to the press or buyers during a trade show. I’m not saying that you’ll be mind controlled and compelled to return the favor by buying the crackers or giving a favorable writeup, but you’ll at least think about it more than you would have otherwise. Many organizations even invoke “no gifts” codes of conduct to guard against things like the reciprocity effect.

But what about Farmville? That’s a free game, right? And most of the gifts are free, too, right? For the most part, but Zynga, the makers of Farmville and other social games like Mafia Wars, nevertheless want new players to come in and existing players to stick around. The gifts in these games are useful to their recipients within the game, so seeing a notification that you’ve gotten one encourages you to log into the game and put it to use. And actually just clicking on the link will start you down the path to installing and playing the game, which increases Zynga’s numbers. Then the reciprocity effect then encourages you to return the favor by sending a gift back, which creates a cycle of reciprocating fruit plants, livestock, and penguin statues flying back and forth. Even worse is when you realize that if you DON’T perpetuate the gifting loop, you’ll hurt your friends by making them waste in-game money for things they were hoping to get from you as gifts, you heartless bastard.

This is an effective mechanism for getting people to perpetually log back in to Farmville, for example, instead of moving on to other games. There’s the notification telling you that you need to log in to reciprocate the gift, and while you’re there you might as well play for a while. You can even send gifts to people who don’t play the game yet, encouraging them to pay you back by starting up a game as your neighbor or teammate. Farms everywhere in an unholy amalgamation of psychology and agriculture!

But wait, there’s more. The real money for companies like Zynga comes when you feel compelled to spend REAL money to reciprocate a “premium” gift. In fact, let’s see what Mark Pinkus, CEO of Zynga, had to say in a recent interview with Charlie Rose³:

We are excited about the future of social games and virtual goods as a revenue model within social games. What I mean by that is …these are free games, and one to two percent of the users will spend money in the games. And they can spend it on virtual goods, virtual gifts we just started selling, and that has been a revenue model that has enabled our company to be profitable for eight straight quarters.

So, enjoy your free sample. But don’t underestimate its effect on you.

1. Tip: if you try to use them with e-mail greeting cards, it gets your monitor all sticky [↩]
2. Cialdini, R. (2001). The Science of Persuasion. Scientific American. February 2001 [↩]
3. And if there’s one thing I know it’s that the kids, they can’t get enough Charlie Rose [↩]

I’m not sure where I first noticed Bayonetta, Sega’s crazy action game starring a witch of the same name (“Bayonetta,” not “Sega”). But I can tell you that I immediately wrote it off as something I wasn’t interested in based on very little actual information. Why? Well, on the 12/25/09 edition of the 4 Guys 1 Up podcast, GamePro’s John Davison gave a pretty amusing assessment:

There’s a guy in our office that you could put a picture of him up on the wall and the guys at Bayonetta would be like “We are making this game for THIS guy. It’s got boobs and hair and guns, and crotch shots, and she sucks on a lollipop, and her hair is her PANTS, and and it’s just AWESOME and things explode and and and and …she’s wearing librarian glasses just in case that does it for you as well!”

My first glance at the game gave me pretty much the same impression and I felt that it was all I needed to know. Hey, I appreciate sexy ladies, but Bayonetta’s appearance immediately made me think it was nothing but fan service for 14 year old boys.¹ From that, I inferred that the rest of the game would be childish, shallow, and cliche so I drop kicked it off my radar.

Clearly, my fears that Bayonetta would be half-assed were unfounded.

So imagine my surprise when Bayonetta was recently released and actually started getting good word of mouth and reviews, scoring in the very impressive high 80s to low 90s on Metacritic. While my snap judgment of the “sexy witch fan service” was dead on (and then some), apparently the mechanics are both deep and fun, not to mention the ridiculously awesome presentation of it all.

So why did I assume the whole game shared traits with its protagonist’s character design?

Psychologists call this “the halo effect”² or sometimes the “affect heuristic.”³ It’s a cognitive bias that happens when your evaluation of one trait unjustifiably bleeds over into your perceptions of others. It’s basically the same reason that interviewers tend to think that if you’re a sharp dresser or have a firm handshake, you’ll be a good employee.⁴ Perhaps more infuriatingly, it’s also the reason that beautiful people often get preferential treatment, even by members of the same sex, when people subconsciously assume that because someone is physically attractive, he/she must also be smart, competent, and likable. Research has shown that this even happens in courts, with uglier defendants receiving less lenient sentences.⁵

Two researchers named Richard Nisbett and Timothy Wilson used a clever experiment to reveal how powerful the halo can be.⁶ They twice filmed a professor ostensibly answering students’ questions. In the first film, the proff acted kindly. In the second film, he was a jerk. Two groups of subjects were then shown one film or the other and then asked to describe how appealing they found his appearance, mannerisms, and even accent. I bet you can guess which version of the professor people thought was better dressed, more suave, and possessed of a more charming accent.

As another example, crack open a history textbook (books are like if Wikipedia were made of trees) and consider the first Presidential campaign debates that were broadcast simultaneously on radio and television in 1960. Richard Nixon looked like crap for the cameras while John F. Kennedy looked well rested and snappy. Later polls showed that radio listeners (who obviously couldn’t form impressions of the candidates’ appearances) thought Nixon had won the debates while those who had watched the EXACT SAME exchange on the television thought Kennedy had nailed it.

So when I first saw Bayonetta, her lascivious appearance formed a halo in my mind (or horns, if you prefer⁷) which led me to believe that every other aspect of the game would be so shallow, base, and uninteresting. In a market with SO many games and so little time, this may actually be an adaptive strategy since first impressions can often be right. But in this case I was surprised to find out that I’m wrong, at least according to Metacritic.

Of course, one man’s horns are another boy’s halos, and maybe I’m just not in the target demographic after all.

1. Which is totally accurate, at least from a character design point of view, though I still fail to see why a woman covered in thick black hair is supposed to be sexy [↩]
2. No relation to Master Chief [↩]
3. I like “affect effect” myself, but nobody listens [↩]
4. Mack, D., & Rainey, D. (1990). Female applicants’ grooming and personnel selection. Journal of Social Behavior and Personality, 5, 399-407. [↩]
5. Stewart, J. E. II. (1980). Defendant’s attractiveness as a factor in the outcome of trials. Journal of Applied Social Psychology, 10, 348-361. [↩]
6. Nisbett, R. E., & Wilson, T. D. (1977). The halo effect: Evidence for unconscious alteration of judgments. Journal of Personality and Social Psychology, 35, 250-6. [↩]
7. No “horny witch” jokes PLEASE [↩]

Fanboys. You don’t have to be a very experienced browser of gaming-related forums to see your share of discussions fouled by flames between people hysterically defending their favored game/console/genre/whatever and attacking everything else in sight. Some of it is deliberate trolling, for sure, but not always. There were (and still are) way too many Xbox owners ready to point and laugh at the Playstation 3′s lack of games. Or good luck trying to find someone who will stand up as a fan of BOTH Halo AND Killzone.¹

Why do gamers do this? One word: “social identity theory.”

Typical Fans

This theory explains (or at least predicted) the Great Console Wars of today and tomorrow. In one study, psychologist Henry Tajfel and his colleagues brought together teenage boys² and asked them to express preference for one of two sets of paintings, saying that their choice would place them in one of two otherwise arbitrary groups. Thus sorted, the boys then participated in a separate study where they distributed (fake) money to their fellow subjects under a variety of conditions. But here’s the key: each boy was told whether those to whom he was doling out the virtual cash were in “his” group or in the “other” group.

I’ll bet you can guess the results: subjects showed stark favoritism for people who had liked the same set of paintings as they had and who were thus in “their” group. Remember that like messageboard denizens, these kids had absolutely no self-interested reason to do this –they weren’t rewarded for favoring their group and they weren’t given any reason to expect their fellow group members to return the favor and be best friends forever. They just did it because they considered those strangers to be “us” and –perhaps more importantly– the rest to be “them.”³

Tajfel and his collaborators theorized that people have a natural tendency to construct identities based on group membership. Part of who you are –and how you communicate that to others– is defined by what groups you belong to. And we naturally want to belong to high-status groups, right? Okay, fine, but everything is relative; a group isn’t high status unless there’s a low status group for it to be contrasted against. So not only do some people identify themselves as Xbox fans, they attack Playstation owners in order to raise their status. This tendency is human nature, the researchers concluded, and a lot of other data support them. What’s more, we’re perfectly willing to do it at the drop of a hat.

Some savvy game designers even build this kind of thing into their game, the biggest example being Blizzard’s long-standing “Horde vs. Alliance” rivalry in World of Warcraft. Some folks will roll toons on either side of the divide, but many hardcore players will vigorously stick to just one side, and Blizzard happily plays this rivalry up in the player versus player aspects of the game.

One of the most interesting uses of social identity theory I’ve seen, though, was pulled off by Valve Software during their recent “Demoman versus Soldier” event for Team Fortress 2.

Valve uses social identity theory for fun and profits

Harnessing their flabbergasting ability to track gameplay stats through Steam, Valve promised a new in-game weapon for the class (Demoman or Soldier) that scored the most overall kills against his opponent during a certain time frame. The results were nuts as people chose sides, let rockets/stickies fly, and created renewed buzz for the game.⁴

I think the Soldier explained it best on the Official Team Fortress Blog:

Gentlemen, I have NO IDEA what this weapon is. I don’t even know if I’ll WANT it. But BY GOD, I know what’s IMPORTANT, and it’s that WE get it and the DEMOMAN DOES NOT.

This is psychological warfare at its finest.

1. And don’t even get me started on the whole “table versus jetski” debacle. [↩]
2. There’s the bulk of your console war soldiers right there, in fact. [↩]
3. Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorization and intergroup behavior. European Journal of Social Psychology, 1(2), 149-178. [↩]
4. In case you were wondering, Soldier won the shootout, but just barely –6,372,979 Soldiers gibbed vs 6,406,065 Demomen. [↩]

For those of you who somehow don’t know, pro golfer and occasional video game star Tiger Woods has recently been in trouble over not keeping his club in his bag, so to speak. The casualties include endorsement deals with companies who capitalized on Woods’s previous reputation as focused and reliable, but interestingly game publisher EA isn’t dropping him. EA Sports President Peter Moore recently said that they intend to stick with the athlete on properties like the Tiger Woods PGA Tour series:

Our relationship with Tiger has always been rooted in golf. We didn’t form a relationship with him so that he could act as an arm’s length endorser. Far from it. We chose to partner with Tiger in 1997 because we saw him as the world’s best, most talented and exciting golfer. We struck that partnership with the assumption that he would remain near or at the top of his sport for years to come.

By his own admission, he’s made some mistakes off the course. But regardless of what’s happening in his personal life, and regardless of his decision to take a personal leave from the sport, Tiger Woods is still one of the greatest athletes in history.

This is surprisingly rational behavior, because Moore (and EA by extension) is explicitly avoiding a very fundamental bias in human psychology: the “fundamental attribution error,” or “FAE.” It’s so fundamental, it’s got “fundamental” right there in its name. You can’t get more fundamental than that!

Tiger doing his best Darth Vader impersonation.

FAE is people’s tendency to rely too much on internal attributes (like personality) to explain others’ behavior and to underestimate external attributes (like the situation or environment).¹ The classic example is to imagine that someone is talking to you and that they cross their arms. FAE would lead us to believe that this body language is because they’re defensive or dislike you. But maybe the thermostat is set low and the person is just cold.

The other classic example of FAE was provided by Edward Jones and Keith Harris in an experiment where they asked subjects to read essays that were either in favor of or against Cuban dictator Fidel Castro.² Even when told that writers were randomly asked to manufacture a pro or anti Castro essay on the basis of a coin toss, subjects STILL tended to believe that the authors wrote what they wrote because they held the views espoused in their essays. In other words, they attributed the essays to internal factors (beliefs) instead of external ones (instructions by the experimenter).

Where does this tie in to Tiger? While browsing Jonah Lehrer’s excellent blog, I came across his post on the Woods controversy, which quoted from a New Yorker article on the same subject:

Woods’s appeal was based, ultimately, not on his physical abilities but on his mental toughness, his extraordinary capacity for focus and discipline. He was the man who always made the key putt, who never cracked under pressure. That’s why Gatorade, introducing a new drink with his face on the label, called the drink Tiger Focus. And it’s why the most powerful Nike ad about him is the one in which his father, in a voice-over, says, “I’d say, ‘Tiger, I promise you that you’ll never meet another person as mentally tough as you in your entire life. And he hasn’t . . . and he never will.’”

In other words, Woods has been presented as the embodiment of bourgeois virtues: dedication, hard work, single-mindedness. Indeed, when, in 2008, Woods won the U.S. Open while essentially playing on one leg, the Times’ David Brooks devoted a column to his extraordinary ability to block out distraction and focus on the matter at hand, dubbing him “the exemplar of mental discipline” for our time. For millions of people–many of them, to be sure, affluent middle-aged white guys–Woods embodied an approach not just to golf but to life.

Lehrer goes on to expand on how this relates to the FAE. Before the scandal, we attributed all of Tiger’s success to his fundamental, internal qualities –his drive, talent, focus, and personality. When the headlines hit the fan, everything we heard about Woods was that he was unfaithful, dishonest, and unreliable. We were shocked (and thus interested in the story) because we couldn’t figure out why someone with such focus and drive could do what he did. It did not compute, because the FAE makes it hard to consider how his behavior on the greens could have more to do with the fact that he was ON THE GREENS and that his behaviors could vary so much across situations.

But squishy human brains tend not to work that way; we think that everything he does is based on the little slice of character we’ve seen on TV, so everyone drastically re-evaluated who he “is.” Which is not to say, of course, that infidelity is excused in any circumstances; it’s just a matter of how shocking it is to us when we held someone in high regard based on seeing him mostly in just one situation and can’t resolve the “two Tigers” we think we see, when it’s really just one Tiger who acts differently when he’s golfing.

This is why I think EA’s decision (for now, anyway) to stick with Woods for their PGA Tour series is commendable, or at least rational. Woods may be a philanderer in some situations³, but on the golf course, real or virtual, he still embodies all those things that EA wants to associate with its brand –competence, focus, reliability, and top-tier play. Despite his cheating on his wife, Woods is still a great golfer.

Well, he would be, if he hadn’t dropped out of the game. But you know what I mean.

1. Curiously, we do just the inverse when attributing causes to our own behavior [↩]
2. Jones, E. E. & Harris, V. A. (1967). The attribution of attitudes. Journal of Experimental Social Psychology, 3, 1–24. [↩]
3. such as when he’s philandering [↩]

How could publishers get way more people to buy an Xbox Live Arcade or Playstation Network game after trying the trial version? Let me glue on my goatee and practice my maniacal laugh a few times and then I’ll tell you my idea.

But first, let me ask you a couple of hypothetical questions made famous in certain circles by Amos Tversky and Daniel Kahneman:¹

Imagine that the U.S. is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientific estimate of the consequences of the programs are as follows:

If Program A is adopted, 200 people will be saved.

If Program B is adopted, there is 1/3 probability that 600 people will be saved, and 2/3 probability that no people will be saved.

Which of the two programs would you favor?

Which would you pick? The researchers found that most people chose Program A: 72% versus the 28% who chose B.

So then the researchers asked the following version of the same question:

If Program C is adopted 400 people will die.

If Program D is adopted there is 1/3 probability that nobody will die, and 2/3 probability that 600 people will die.

Which of the two programs would you favor?

Which would you pick? Most of the experimental subjects picked Program D by a wide margin — 78% versus the 22% for Program C. The thing is, both sets of choices are identical. Look closely. Programs A and C both result in 400 people dying and 200 living. Programs B and D both have a 1/3 chance of saving everyone and a 2/3 chance of killing everyone. The ONLY difference is that Programs A and B are phrased in terms of lives saved and Programs C and D are described in terms of lives lost.

Tversky and Kahnaman said this points to “loss aversion,” which is one my favorite kinks in the human brain. In short, loss aversion is our willingness to go to great lengths to avoid losses –much farther than we’ll go to get an equivalent gain. In other words, losing $10 is more painful than gaining $10 is pleasurable.

Consider another quick question and suppose that a company were offering two subscription plans for an online MMORPG.

• Option A gives you a $5 credit
• Option B lets you avoid a $5 monthly surcharge

Assuming both options were otherwise identical, which do you think would be more popular? In all likelihood it would be Option B, since people prefer not losing $5 to gaining a $5 discount. This despite the fact that the monthly costs would be identical. This is also one of the reasons you’ll more often see “$10 late registration fee” advertised instead of “$10 discount for early registrations” for events where the organizers want you to register early.

So what does this have to do with getting people to buy a Xbox Live Arcade or Playstation Network game after they play the trial version? Right now, it’s not uncommon for such trials to pop up a message saying something to the effect of “You would have just gotten an achievement/trophy just now! Buy the full game to get it!”

And that’s pretty good. Pretty sneaky. Pretty psychological. Because we obviously like getting things we value.² But the phenomenon of loss aversion suggests a way to be better, more sneaky, more psychological. Instead of saying that you will get the achievement or trophy if you buy the game, actually give it to them and then say you’re going to take it away if they DON’T buy the game. And I mean really give it to them –have it show up in their gamer score and on their achievement/trophy list. Just take it away if they exit the trial version of the game without buying the full thing, and make sure they know it.

So, to all the game developers/publishers out there, I guarantee that your conversion rate will go up, because while people like the promise of getting something, they hate the promise of losing it way more. Just don’t tell the gamers that you got the idea from me.

1. Tversky, A. & Kahneman, D. (1981). The Framing of Decisions and the Psychology of Choice. Science, 211, 453-458. [↩]
2. The “I don’t care about Gamerscore” folks can just put a cork in it, here [↩]