The impact of Deci & Ryan’s self-determination theory in social psychology is difficult to overstate. This is the source of the frequently argued point that intrinsic motivation (motivation to do something based on inherent properties of the task that make it interesting or enjoyable) is undermined by extrinsic rewards (i.e., points, money, badges, etc.). The theory is not universally accepted: Extrinsic rewards offered for performing boring tasks have been demonstrated to lead to greater intrinsic motivation, and two (controversial) meta-analyses have argued that rewards that reflect competence increase inherent interest in a task. Chris Hecker has sorted through the literature to illuminate the two points on which most researchers seem to agree:

For interesting tasks,

1. Expected, tangible rewards that are contingent on task completion (generally) reduce free-choice intrinsic motivation, and
2. Verbal, unexpected, informational feedback (generally) increases free-choice and self-reported intrinsic motivation.

Let’s focus on the first point. Why is this true? Some argue that it’s because we tell ourselves stories about why we do what we do: when rewards are present, our story is more likely to become “because I’ll get a reward” than “because I want to.” Others say it’s because we feel that we are being cajoled or bribed, which causes us to devalue the task—if someone has to reward us for doing it, it must not be worth doing on its own. In either case, it is our beliefs about the inherent value of the task that drive intrinsic motivation. The danger in both cases is that when the rewards are removed or become uninteresting, our motivation for the task will vanish as well.

However, things get complicated in the case of games, in which some tangible, expected, contingent rewards also happen to make a game more inherently interesting or enjoyable. Furthermore, rewards that are extrinsic to one aspect of a game can provide intrinsic motivation for another, often in an indirect or unpredictable way. A thought experiment from Jesse Schell’s 2010 Gamasutra interview illustrates how easily things can get hairy:

"But this is where it gets tricky -- is that intrinsic and extrinsic are tangled in complicated ways. So, for example, I may set up a system of giving out points, right, that's totally extrinsic. And you would say, "Well, therefore, in the long run, it won't work."

Well, but what if me and my friends all kind of get into it, and like we start this kind of social thing about one-upping each other, and we're now doing it not because we care about the points for the sake of the points, but it now becomes like a little social ritual with us, which is intrinsically rewarding?

So, these extrinsic systems can sometimes become an anchor for something that has intrinsic power, and that part is where I think our brains get a little tangled up, because it's difficult to predict and it's difficult to plan for."

And another from Ted Castronova (reprinted with permission):

"If a man does not want to do a task, but does it in return for $10, we say that he was extrinsically motivated, but not intrinsically motivated. OK.

If the man plays a game and is not paid to do it, we say that he was intrinsically motivated. He "wanted" to.

Suppose a man is playing a game and does not want to perform a task, such as, mining. Yet if he mines ore and sells it in the game, he gets a thing called "gold pieces." So he performs the task. Is he intrinsically or extrinsically motivated? Which if any of the following are true?

1. The man is intrinsically motivated to play the game, since he wants to do it without any reward other than just playing the game.

2. The man is extrinsically motivated to play a part of the game, the mining.

3. The mining is nested inside the game. Thus you can have extrinsic motivations nested inside intrinsic motivations.

4. The man is motivated partly intrinsically and partly extrinsically to play the game.

Here's where it cuts. According to the theory, a game is more engaging if it is more intrinsically rewarding… These games could take out all the mechanics that use extrinsic motivation, all the grinding elements. It seems to me that that would be a pure candy environment, and boring."

The research on intrinsic motivation suggests that if the mining task is “interesting” (in the absence of reward), then receiving a reward is likely to make it seem more dull; if the mining task is dull, receiving a reward might make it seem more interesting, or it might not, depending on whose theory you subscribe to.

If we assume the mining task is dull, then fewer people are likely to put in the effort to perform it, making the resulting gold pieces objectively more valuable. Also, independent of the actual economics, self-perception theory suggests that the duller the task, the more valuable the reward will seem (I wouldn’t be swinging this heavy pick, the player tells herself, if I weren’t getting something totally awesome in return). I’m speculating here, but it doesn’t seem too implausible that having an in-game currency of high psychological value enables a more interesting economy, more difficult and compelling choices about how to allocate one’s resources, and other high-level phenomena that make the game inherently enjoyable. So here, as in Schell’s example, eliminating a mechanic that seems superficially extrinsic is likely to undercut an intrinsic motivator that supports the experience of playing the game as a whole.

If we can’t be sure about the effects of eliminating seemingly extrinsic motivators, a logical alternative strategy would be to amp up seemingly intrinsic ones. In the next post, I’ll talk about how multifaceted theories of motivation provide game designers with more actionable ways to do just that.