There is a question I have been asking in talks for fifteen years, and the answer is always the same. I ask: what is gamification? And the room answers, almost in chorus: points, badges, leaderboards. It is a remarkably stable misconception, because it is not merely imprecise, it is exactly backwards. Points and badges are not the heart of gamification. They are the thing you most reliably use to destroy it.
That sounds provocative, so let me make it precise. Gamification can work beautifully. But it works the moment it is built non-Skinnerian, meaning not as a reward machine that trades behavior for small external receipts. It works when it does what good games have always done: make progress experienceable, keep challenge and skill in balance, and give feedback that does not consist of a number but of access, of new possibilities, of the next, slightly harder step. And precisely this mechanism can be dismantled with a single, well-intentioned addition.
Two projects, one pattern
I am not telling this from theory. I am telling it from two cases I will never forget, because they show the same pattern. The first we built ourselves, the second we learned about firsthand from people involved.
In the first, we had built a system for a company that rested entirely on progress. There were no points. There were balanced difficulty levels, and the feedback for progress was that you gained access to new things that made the activity more demanding and more interesting. Whoever got better got more to do, not more to collect. The system worked well. Then management changed, we had long been off the project, and the new team looked at it and said the sentence that always comes up in stories like this: cool, let's improve it. We'll add a reward system on top. Points. Badges. And with that, in a single step, they reinstalled exactly the extrinsic reward we had painstakingly designed out.
The second case we did not run ourselves, we learned about it firsthand from people involved who witnessed the trajectory before and after the change directly. It concerned an internal community at Adidas that worked on its own. There was engagement, there was exchange, people contributed because contributing was itself worth something. Here too, at some point the thought arrived: let's fuel this further. More incentive. We'll hand out points and badges for actions and interactions. And from that point on, according to those directly involved, you could watch, fairly precisely, how it went downhill. Not because people were suddenly less kind to one another, but because the meaning of the action had shifted. Whoever had contributed because it mattered to them now did it for the points. And once the brain has made that reinterpretation, the rule becomes: I'm only doing this to get something.
In both cases the worsening was well intended. No one set out to destroy engagement. They wanted to amplify it, and in doing so they pressed the one button that reliably produces the opposite.
The brain lives in the future
To understand why this goes wrong so reliably, you have to go briefly to where our work has its origin: the neuroscience of prediction. Over the past years, Lisa Feldman Barrett has helped shape perhaps the most important shift of perspective in brain research. The brain is not primarily reactive. It does not react to the world, it predicts it. It lives, to put it pointedly, permanently a few hundred milliseconds in the future and continuously generates predictions about what will happen next. Learning, attention, and what we experience as tension or interest arise where a prediction does not come out exactly, in what is called the prediction error.
Now picture what a classic points and badge system does to this machinery. It makes the prediction certain. You do X, you get a hundred points. Every time. The brain predicts: I do this, and the expected thing happens. The prediction comes out exactly, the prediction error goes to zero, and with it the reason to stay engaged with attention and tension. Why stay with something when the system permanently confirms to you that nothing unexpected will happen anymore? You have systematically designed away the condition for engagement, which is productive uncertainty.
Dopamine is the chase, not the catch
The dopamine research paints the same picture, and here things are often miscited, so let's take it in order. In his lectures, Robert Sapolsky describes an experiment that turns the popular notion of dopamine on its head. A monkey has to press a lever ten times to get a reward. The dopamine release does not begin when the reward arrives. It begins with the signal announcing that the task is now starting, and it lasts as long as the work goes on. More striking still: if you give the reward only half the time, the dopamine release doubles. Uncertainty produces more drive than the guaranteed reward.
The precise neuronal mechanics behind this were described by Wolfram Schultz: dopamine neurons encode a reward prediction error. They fire for a reward that turns out larger than expected, and they stay silent for a fully predicted reward. A guaranteed reward simply produces no dopamine signal. That is the same logic as Barrett's, just one floor down, at the level of individual neurons.
Points and badges are predictable by construction. They turn the reward into a guarantee, and the brain responds to guarantees with silence. The reward is not the chase. The reward is the chasing after it.
This also dispenses with the most common objection. But people do collect the points, they want the badges, you can see it in the numbers. Yes. In the short term, every new layer of reward produces a spike, because it is unexpected at first. That is exactly the trap. What gets measured as an uptick is the beginning of habituation. As soon as the reward has become expectable, and that happens fast, the system drops below its starting level, because now the original, intrinsic drive is missing and the external one has already flattened out.
Flow runs in the wrong direction
That leaves the third pillar, and it concerns not the wanting but the difficulty. With his flow research, Mihaly Csikszentmihalyi showed that the state of deepest absorption and highest performance hangs on a moving condition: the challenge must grow along with the skill. If you get better and the task stays equally hard, you tip into boredom. Flow demands that it gets harder when you get better.
And now look at what points and badge systems almost always do in practice. They help you get the points. They reduce friction, they give hints, they make the next point more attainable, because the measured success, the point balance, would otherwise look bad. So they pull difficulty downward, in exactly the moment when skill is pointing upward. That is the precise opposite direction to flow. You optimize for collecting and thereby make rising impossible.
Put the three findings together and a consistent picture emerges. Barrett shows that certain predictions dry out engagement. Sapolsky and Schultz show that guaranteed rewards silence the drive system. Csikszentmihalyi shows that point logic pulls difficulty downward, while flow needs the opposite. Three different disciplines, the same diagnosis.
What this does not mean
It does not mean you should do without any feedback. On the contrary, feedback is central. It means feedback must not take the form of a guaranteed external receipt. Good playful feedback opens something up: a new area, a more demanding task, a visibly grown ability. It does not merely confirm that the expected thing has occurred. It shows that something has become possible that was not possible before.
That is the difference between gamification that holds and gamification that collapses. Not the question of whether game elements are used, but whether they keep the activity worthwhile or turn it into a means of collecting. If you want to build engagement, do not build points. Build progress that keeps the brain in the future, uncertainty that keeps the drive system awake, and rising challenge that makes flow possible. Points and badges are the easiest solution. And that is exactly why they are usually the wrong one.