Try the Game Here

Why my own game failed the kids it was built for

I built a math game for kids called Brain Coins. The concept is simple: solve math problems, earn coins, level up. It's fast, it's arcade-style, and kids love the energy of it. But I discovered a problem I didn't design for, and the children I was studying are the ones who showed it to me.

Several of them play Brain Coins well below their actual grade level.

These aren't kids who struggle with math. In other contexts, I've watched them build math problems above their grade level for fun. I've seen them solve multiplication on a playground without hesitating. One child created his own math quiz and then voluntarily extended the problem by calculating comparative ratios. Another figured out a mental math shortcut I hadn't considered and started teaching it to a sibling.

But inside Brain Coins, these same kids drop one, sometimes two full grade levels. And the reason is always the same. The timer.

The speed-accuracy tradeoff

Brain Coins gives players a fixed amount of time to solve each problem. The timer doesn't adjust based on the complexity of the question. A kindergarten addition problem and a third-grade subtraction problem get the same window. And that window is where everything falls apart.

These children can do grade-level math. But they can't do it fast enough under time pressure to succeed in the game. The cognitive steps required for multi-digit subtraction are fundamentally different from the steps required for single-digit addition. One requires regrouping, borrowing, and multi-step processing. The other is pure recall. Giving them the same timer doesn't measure the same thing. It measures processing speed for simple problems and penalizes complexity for harder ones.

So the kids do what any smart person would do. They drop to the level where they can win. Kindergarten math. First-grade math. Problems they can solve by recall, fast enough to beat the clock, easy enough to keep a streak alive. The game registers them as performing at a kindergarten level. In reality, they're performing at a strategic level, optimizing for the system's constraints rather than demonstrating what they actually know.

The game isn't measuring their ability. It's measuring their speed. And those are two very different things.

The frustration spiral

There's a second layer to this. When these kids do attempt problems at their actual grade level and get stuck on consecutive problems, the reaction is consistent across almost every child I've observed. It's not quiet defeat. It's a flash of frustration, sometimes a mini meltdown, sometimes throwing hands up and quitting the round. Not because they don't understand the math. Because they know they should be able to do it and the system won't give them enough room.

The emotional response isn't "I can't do this." It's "I should be able to do this and this isn't fair."

That frustration is a signal. It's telling me the game is creating a gap between what these children know and what they can demonstrate. And that gap feels, to the child, like failure. Repeated failure on problems you know you can solve is one of the fastest paths to disengagement. The kids don't think "this game's timer is too short." They think "I'm bad at this." And once that belief sets in, they stop trying the harder levels entirely.

I watched this pattern repeat across multiple children. After a streak of missed problems at grade level, they'd drop back to the easy rounds and stay there. Not because they wanted easy problems. Because easy problems were the only ones where the game let them feel competent.

What the game should have taught me sooner

I found a workaround during one of our sessions. I taught a few of the kids some mental math shortcuts, techniques for breaking larger problems into smaller steps they could execute faster. It helped. Their speed improved and they were able to handle some of the harder problems within the timer. But it also revealed something important: the problem was never their math ability. It was the interface between their ability and the system's constraints. The shortcuts didn't teach them new math. They taught them how to translate what they already knew into a format the game could recognize.

That's a design failure, not a learning failure.

When I stepped back and looked at Brain Coins through the lens of my broader research, the issue became clearer. I've been studying how children learn best, and every finding points in the same direction: kids engage deeper when the environment works with how they naturally process, not against it. Brain Coins, despite being a game I built specifically for kids, was replicating one of the worst patterns of traditional education. It was using a fixed, inflexible constraint as a proxy for understanding and punishing kids who couldn't perform under that constraint, regardless of whether they actually knew the material.

The classroom version of this is a timed test. Same problem. A child who needs an extra thirty seconds to work through a multi-step problem gets marked wrong, not because they got the wrong answer, but because they got no answer. The system records a failure. The child internalizes it. And the actual ability sits there, untested and invisible.

I built Brain Coins to be different from the classroom. But the timer made it the same.

What I'm rethinking

Brain Coins needs to change. Not the core concept. The energy of the game is right. Kids love the arcade feel, the coins, the leveling. But the timer mechanic needs to be rethought from the ground up.

A few directions I'm exploring:

The timer could flex based on problem complexity. A single-digit addition problem and a multi-digit subtraction problem require different amounts of cognitive processing. The timer should reflect that. Not by making harder problems feel slow and easy, but by giving the player enough runway to actually think without being punished for attempting difficulty.

Bonus time could function as a reward mechanic. Solving a streak correctly could earn extra seconds on the next problem. This keeps the arcade pressure alive but gives kids a pressure valve. Speed is still rewarded, but it compounds into breathing room rather than just points.

Strategy hints could appear before a round. A quick tip on mental math shortcuts, breaking a problem into smaller steps, or recognizing patterns could serve as a coaching moment that's part of the game itself, not separate from it. Kids learn how to work faster without being told to just "try harder."

A wrong answer could lead somewhere instead of nowhere. Right now, missing a problem in Brain Coins just costs time and coins. What if a missed problem branched into a quick breakdown of how to approach it, then re-presented the problem? The miss becomes a learning moment instead of a penalty. The child still wants to get it right because coins are on the line, but getting it wrong doesn't feel like a dead end.

The bigger principle

This experience reinforced something I keep finding across all of my research: the learner's ceiling is set by the environment, not by their ability. Brain Coins was supposed to be a better environment for learning math. In many ways it is. But the timer mechanic created an invisible ceiling that pushed capable children down to levels far below what they could actually do. The game didn't know they were capable of more. It only knew they were slow.

I built Brain Coins to make math feel like play. For the easy levels, it does. For the harder levels, it accidentally made math feel like a timed test. And timed tests are exactly the kind of environment my research keeps arguing against.

The fix isn't to remove the timer entirely. Kids like the urgency. It makes the game feel like a game. The fix is to make the timer serve the learner instead of gatekeeping them. Speed can be part of the challenge without being the only measure of success.

Children who drop two grade levels in a game they love are not children who can't do math. They're children whose ability is trapped behind a design decision that doesn't account for how they actually think. That's on me, not on them. And it's fixable.

Try the Game