Gamification Considered Harmful?

Points, levels and leaderboards are often perceived as the bread and butter of gamification. Gabe Zichermann and Christopher Cunningham (2011) even call them “the heart of any gaming system” and “an absolute requirement for all gamified systems”. Game designer Margaret Robertson (2011) on the other hand decries this practice as pointsification and deems it “the thing that is least essential to games”. Similarly, Chris Hecker (2010) warned game designers not to blindly resort to achievements (or points, levels and leaderboards for that matter), because they could stifle players’ intrinsic motivation, that is, their desire to engage with a game (or gamified service). Intrinsic motivation denotes the pursuit of an activity, because it is inherently interesting or enjoyable, whereas extrinsic motivation is defined as doing something due to external events, such as money, deadlines or “good player badges”. However, psychological studies have shown that such extrinsic motivators are often ineffective and will eventually decrease people’s intrinsic motivation for even the most fun activities. While Hecker provided a comprehensive summary on current research on human motivation, he also stressed the need to specifically study the effects of game design elements, such as points, levels, and leaderboards, on intrinsic motivation.

Presentation at Gamification 2013

We conducted a controlled online experiment, where we compared people’s performance and intrinsic motivation in four different versions of a simple image annotation task:

1) Points only vs. 2) Levels vs. 3) Leaderboard vs. 4) Control condition, which featured no game elements whatsoever.

Study participants received 100 points for every tag they entered and could always see how many points they needed to score to reach the next level / next rank on the leaderboard. Following previous research on the effects of rewards on intrinsic motivation, we assumed that these game elements would decrease users’ self-reported intrinsic motivation, compared to the control group.

Surprisingly, our results revealed that not only did points, levels and leaderboards drive users to generate significantly more tags than users in the control group, but no differences concerning intrinsic motivation were found between the experimental conditions. Put differently, the implementation of these common gamification elements did not harm people’s enjoyment of the image annotation task. It seems that for certain types of activities – short, simple and voluntary tasks, – the oftentimes much maligned points, levels and leaderboards are indeed a viable means to increase productivity, without compromising users’ intrinsic motivation.

However, our study only just scratched the surface of the whole debate. For instance, why did points, levels and leaderboards enhance people’s productivity without boosting their intrinsic motivation? The jury is still out on that one and there are several possible explanations, but as Robertson (2010) puts it: “They’re great tools for communicating progress and acknowledging effort, but neither points nor badges in any way constitute a game.” In fact, research in psychology suggests that games are so intrinsically motivating, because they provide some form of voluntary challenge for players to take on (Przybylski et al., 2010). The image annotation task in our study could hardly be considered “Nintendo Hard” and offered little opportunity for skill mastery. In an ongoing research project, we aim to gain a better understanding of whether and how gamification may actually increase students’ intrinsic motivation and facilitate the mastery of more demanding curricular activities.

The full paper was published in the Gamification ’13 Proceedings and is available here.

Abstract: It is heavily debated within the gamification community whether specific game elements may actually undermine users’ intrinsic motivation. This online experiment examined the effects of three commonly employed game design elements — points, leaderboard, levels — on users’ performance, intrinsic motivation, perceived autonomy and competence in an image annotation task. Implementation of these game elements significantly increased performance, but did not affect perceived autonomy, competence or intrinsic motivation. Our findings suggest that points, levels and leaderboards by themselves neither make nor break users’ intrinsic motivation in non-game contexts. Instead, it is assumed that they act as progress indicators, guiding and enhancing user performance. While more research on the contextual factors that may potentially mediate the effects of game elements on intrinsic motivation is required, it seems that the implementation of points, levels, and leaderboards is a viable means to promote specific user behavior in non-game contexts.

Elisa Mekler
PhD Student @ HCI Research Group, Center for Cognitive Psychology and Methodology, University of Basel

Hecker, C. (2010). Achievements considered harmful.

Mekler, E. D., Brühlmann, F., Opwis, K. & Tuch, A. N. (2013). Do Points, Levels and Leaderboards Harm Intrinsic Motivation? An Empirical Analysis of Common Gamification Elements. In Gamification ’13 Proceedings of the First International Conference on Gameful Design, Research, and Applications, 66 – 73.

Przybylski, A. K., Rigby, S. C., & Ryan, R. M. (2010). A motivational
model of video game engagement. Review of General Psychology 14, 2, 154–166.

Robertson, M. (2010). Can’t play, won’t play.

Zichermann, G. & Cunningham, C. (2011). Gamification by design: Implementing game mechanics in web and mobile apps. O’Reilly Media, Sebastopol, CA.

(Note: For a followup, read “Gamification Absolved?“)

About Elisa Mekler

10 thoughts on “Gamification Considered Harmful?

  1. I didn’t read the full paper because you have to pay for it, so my apologies if something I’m confused about is explained in the paper. I’m quite confused by your use of “intrinsic motivation” in the context of this study. Specifically:

    “no differences concerning intrinsic motivation were found”
    “did not harm people’s enjoyment of the image annotation task”

    Because when you said:

    “The image annotation task in our study could hardly be considered “Nintendo Hard” and offered little opportunity for skill mastery” this alone makes it seem very unlikely that there *was* ANY intrinsic motivation for the task.

    By the SDT definitions of “intrinsic motivation”, wouldn’t that require the participants to first *find* the task “intrinsically rewarding”? i.e. something they’d do *for its own sake* because of its inherent pleasurable enjoyment. It does not appear that the image annotation task fits the criteria for intrinsically rewarding.

    I’d really like to understand this better. If this task was *not* intrinsically rewarding (by the definition of SDT), then this study isn’t testing the impact of gamified elements on *intrinsic motivation*. Again, I didn’t read the paper, but from the statements in this post, it sounds like you found a perfect example of the “safe” use of gamification: gamification can help increase performance on a NON-intrinsically-rewarding experiences. Unless the image annotation task was something people would do on their own, and purely for the enjoyment, in the same way people would *choose* to play a game, read a good book, etc. then it does not meet the criteria for intrinsic motivation. Which means there was nothing for the gamified elements *to* harm.

    1. Thanks so much for your comment, Kathy! You’re absolutely right about external motivators decreasing intrinsic motivation for activities that people find interesting and enjoyable. I should have been more clear about the initial motivational appeal of the image annotation task. In our study, the image annotation task was rated somewhat above average on intrinsic motivation (mean intrinsic motivation = 4.78 on a scale from 1 to 7) and several study participants also explicitly mentioned that they did enjoy the activity. However, your point is still valid on whether the task was interesting *enough* to be harmed by gamification elements. We definitely need more research about when gamification can be detrimental (or beneficial) to intrinsic motivation for activities that people find already enjoyable and interesting as such.

  2. Hi Kathy, the study *did* indeed measure intrinsic motivation (it used the established Intrinsic Motivation Inventory self-report scales), thus showing that adding points, leaderboards, or levels to a task does not automatically increase intrinsic motivation – as Mekler and colleagues suggest (and is plausible from SDT), for feedback to be competence-inducing, it better be about something considered challenging :) — that’s to me the really interesting (and plausible) finding of the study.

    However, the study arguably does not test undermining effects, because undermining effects have not been found (nor theoretically argued for) for informational verbal feedback to begin with, only for tangible rewards or controlling verbal feedback. The leaderboard, point, and level design of Mekler and colleagues arguably constitutes neither – no monetary payout was attached to how well people did, nowhere did anything spell out “you must/should/ought achieve X”.

  3. Really cool study. Thanks Elisa!

    IMHO, Deci & Ryan set the SDT theory on a continuum, which means that no activity is 100% intrinsically regulated, all the time, in all the circumstances. If we are to take Amabile & Kramer’s The Progress Principle approach, any feature connected to an experience that is geared to provide feedback and show progress within the framework of that experience has the potential to reinforce the intrinsic motivators. The sole experience of earning rewards as a digital expression of progress is interconnected with the context of the activity, and it becomes particularly difficult to discern whether the activity foster intrinsic motivation, or is fueled by extrinsic ones.
    I agree with Sebastian here. Without the competence factor, progress becomes either irrelevant or shortlived.

    1. Hi Raul, totally agree with you. As you probably know, organismic integration theory, a sub-theory of SDT, describes motivation on a continuum including intrinsic motivation, as well as several different types of extrinsic motivation (Gagné & Deci, 2005), going from highly controlled and unrelated to the task (e.g., external: most forms of monetary incentives), to highly autonomous and task-relevant (e.g., integrated motivation: engaging with a task because it is meaningful, that is, it supports one’s own values). Besides competence need satisfaction, I believe providing meaning also has the potential to foster intrinsic (or integrated) motivation. See also Nicholson’s (2012) framework for meaningful gamification.

  4. Hi Elisa,

    I just skimmed over the article, sorry if I ignore something you wrote.
    I think the problem of undermining intrinsic motivation could/should emerge in a follow-up or long-term study: Will (the same) participants be equally intrinsically motivated and/or productive if they don’t get their rewards anymore? I think that’s the problem in using extrinsic motivation: it has a short-term effect, but can’t foster behavior in the long run. As you discuss in limitations chapter, this would explain increased motivation/productivity in short-term, “one-time” tasks and motivating / gamifying these tasks extrinsically would be ok. On the other hand, long-term changes in behavior could only be achieved through intrinsical motivation (even if it’s started externally). I think the study of Thom, Millen and DiMicco is a perfect example of what I’d assume will happen most of the time when extrinsic motivation is removed or has lost its novelty-factor. Till now, I’m not that good in sdt, but wouldn’t the removal of extrinsic motivation be the point where overjustification effect kicks in?

Leave a Reply