How Gamification Can Transform Learning

Why games are so good at teaching

Watch a kid play a good video game for ten minutes. They fail, try again, look closer and try something new. Nobody has to tell them to keep going. Good games are learning machines. They give a clear goal, quick feedback and a next challenge that sits just past what you can already do.

Gamification means borrowing those parts for something that isn’t a game, like a math lesson. Points, levels, badges, stories and quests are the usual tools. Done well, they can turn practice into something kids look forward to. Done badly, they turn learning into a slot machine. The difference comes down to how motivation actually works, so that’s where I’ll start.

What dopamine actually does

Dopamine gets called the brain’s pleasure chemical. That’s not quite right. It’s a chemical messenger, and some of the best-known research on it comes from monkeys learning to expect a sip of juice.

In the 1990s, Wolfram Schultz and his colleagues recorded dopamine cells while monkeys learned that a light or a sound meant juice was coming. At first the cells fired when the juice arrived. Once a monkey learned the pattern, the cells fired at the signal instead. If the juice then didn’t show up, their activity dipped below normal.

So dopamine tracks surprise about rewards. Scientists call this a reward prediction error: the gap between what you expected and what you got. Better than expected sends a burst. Just as expected sends almost nothing. Worse than expected sends a dip. It works like a teaching signal that says, “Pay attention. Update your guess.”

Kent Berridge and Terry Robinson added a twist. Their work suggests dopamine has more to do with wanting something than with liking it. Wanting is the pull that gets you moving. Liking is the enjoyment once you get there. The two can come apart. That helps explain how people keep scrolling a feed they aren’t even enjoying.

Curiosity runs on the same wiring

This matters for learning because curiosity uses the same reward system. In a 2014 study, Matthias Gruber and his team showed people trivia questions while scanning their brains. When people were really curious about an answer, reward areas linked to dopamine became more active. They also remembered those answers better.

The surprising part came next. While people waited for an answer they were curious about, the researchers flashed an unrelated face on the screen. People remembered those faces better too. Curiosity seemed to put the brain in a state that’s ready to learn, even things that weren’t the point.

That gives game designers a better target than points. A good mystery, a story that makes you wonder what’s next, a puzzle you can almost solve: these tap the same system that prizes do. But they point attention at the thing you want learned, not at the prize.

When rewards backfire

Rewards can also do real harm. In a classic 1973 study, Mark Lepper and his colleagues found preschoolers who already liked drawing with markers. Some were told they would get a “Good Player” certificate for drawing. Others got no reward, or got the same certificate as a surprise.

Later, the markers came out again during free play, with no reward on offer. The children who had drawn for the promised certificate spent less time drawing than the others. Being paid had turned something fun into a job. Psychologists call this the overjustification effect.

In 1999, Edward Deci, Richard Koestner and Richard Ryan pulled together 128 experiments on this. Expected prizes, like money or certificates, tended to lower people’s later interest in the task. Positive feedback didn’t have that effect.

Ryan and Deci’s self-determination theory explains why. People stay motivated when they feel three things. They have some say in what they do (autonomy). They’re getting good at it (competence). And they feel connected to other people (relatedness). A reward that feels like control works against the first. Feedback that shows you’re getting better feeds the second.

So does gamification work?

Mostly yes, with a big “it depends.” In 2014, Juho Hamari and his colleagues reviewed the research so far. Most studies found good effects, but results depended a lot on the setting and on the people using it.

In 2020, Michael Sailer and Lisa Homner combined the results of many experiments on gamified learning. They found small but real gains in what people learned, in their motivation and in how they behaved. Results varied a lot from one design to the next.

The pattern across all of this is pretty clear to me. Gamification works when the game parts support the learning. It fails when they’re stuck on top of a boring drill, or when they train kids to chase the reward instead of the skill.

Engagement isn’t the same as learning

Many apps measure success by minutes spent and days in a row. Those numbers are easy to grow with tricks that have nothing to do with learning. Streaks make you afraid to miss a day. Random prize boxes work like a slot machine. Countdown timers add pressure. Endless feeds never give you a natural place to stop.

These tricks lean on the wanting side of the dopamine system. For adults they’re annoying. For a three-year-old they’re worse, because young kids are still learning how to stop on their own. I didn’t want any of that in a game for preschoolers.

So I flipped the usual question. Instead of asking how to keep a child playing longer, I asked how to make each minute count, and how to make stopping easy.

Why start with math?

Early math matters more than most people expect. In 2007, Greg Duncan and his colleagues looked at six large studies that followed children over time. Of the skills kids had when starting school, early math was the strongest sign of how well they’d do later. It beat early reading and attention.

A practice guide from the U.S. Institute of Education Sciences gives the basic recipe for young kids. Teach numbers in a steady order, from small amounts up. Use play, everyday talk and real objects. Watch how each child is doing and adjust. That’s a recipe a well-designed game can follow.

Nova Numbers: the math is the game

Nova Numbers: The Lost Starlight is my attempt to put all of this into practice. It’s a 3D space adventure for kids about 3 to 5. The galaxy’s little lantern-stars, the Starlets, have fallen asleep. A harmless cloud called the Sleepy Swirl has mixed up the shapes, numbers and paths of six small worlds. The child flies a little ship with a friendly crew and wakes the Starlets one mission at a time.

It takes its spirit from Math Blaster Jr., a 1990s computer game with a spaceship to explore, planet missions and math toys to play with. Everything in Nova Numbers is new, from the characters and art to the music and voices.

There are six worlds with twelve missions each, 72 in all. They build up from groups of one to three, all the way to counting to 20 and adding and taking away within 10. Ten kinds of activities carry the math. Kids pop gems into baskets, build rockets from shapes, route marbles through gates, hop across comet stones in order and connect stars into pictures.

A child who can’t read can play the whole thing. Every instruction is spoken, and every button is one big tap. It runs in a phone’s web browser and keeps working offline after the first visit.

How the game uses rewards

The main reward in Nova Numbers is the math working. Pop the right number of gems and a flower blooms. Put the right shape in the right socket and a rocket part locks in. Pick the right path and the marbles roll home. Each reward shows the child exactly what their answer did. That kind of feedback builds a feeling of being good at something, not a hunger for prizes.

Surprise and story do the rest. Each world has something new to see, and the story gives a reason to find out what happens next. That points curiosity at the next problem instead of at a prize box.

There are prizes, but they’re small and predictable. Finishing missions earns stickers for a sticker book, 36 in all. Each world also adds a new outfit for Nib, the crew’s pet, and an upgrade for the ship. None of them depend on speed, and none of them can be lost.

What I left out on purpose

Nova Numbers has no timers, lives, scores to lose, streaks, leaderboards, ads, purchases, accounts or tracking. Progress stays on the device. The game never says “wrong.” A wrong tap gets a gentle wobble, and Pip, the child’s little astronaut, never falls off a comet stone.

Hints come in steps. First the game teaches a way to work it out, like counting together one gem at a time. Only after that does it show the answer. Moving on depends on finishing the story, not on being perfect, so a child who needs hints still reaches the next mission.

Stopping is built in too. Grown-ups can turn on a short break about every ten minutes, and set a gentle reminder when it’s time to finish.

Help that adapts, without labels

Kids this age vary a lot, and not in a neat line. A child might count to ten easily but still mix up which group has more. So the game tracks 17 skills separately, like counting objects one by one, comparing groups and spotting patterns. It never guesses a level from a child’s age.

Only a child’s first try counts as doing it alone. Answers given after a hint are kept separate. The game marks a skill as Consistent only after at least six first tries with five right, across two kinds of activities and two play sessions.

Grown-ups can open a parent area behind a simple gate. It shows informal notes for each idea: Emerging, Practicing or Consistent. There are no grades and no test scores.

What I still don’t know

Everything above is a design choice backed by research on motivation and early math. It isn’t proof that Nova Numbers works. No study has measured whether kids learn more from it. The real comparison is a tough one: playing with blocks on the floor next to a grown-up.

If I ever test it, I’d want to know a few things. Do kids use what they learned away from the screen? Do they come back because they want to, not because something nags them? And do they stop easily when it’s time? Those answers would tell me more than minutes played ever could.