🧩What Building a Puzzle App Taught Me About How Kids Think

I have loved puzzles for as long as I can remember. Logic grids, word games, the kind of riddle that makes you groan when you finally see it. Retro games that asked you to figure things out without a tutorial. Board games where the fun is in the thinking, not the dice.

So it was probably inevitable that, once I started building software myself with an AI coding assistant, one of the first things I would make was a puzzle app.

It is called Yosi, or Budhipur Daily Puzzles. Every day there is a new puzzle. There are eight different puzzle types: grouping puzzles where you find the hidden connections between words, logic deduction grids, drag-and-arrange puzzles, lateral-thinking challenges, and a few more. It is a web app, and I built it myself. A version for Orchids students is what I plan to build next.

I expected building it to be fun. I did not expect it to teach me so much about how children think, and about how badly I had misjudged difficulty.

Why one a day

The first decision was the most important one: one puzzle a day, and then you are done.

Almost everything on a phone today is designed to never end. There is always another video, another post, another level. The business model depends on you not putting it down.

I wanted the opposite. A daily puzzle is a small ritual. You sit with it for a few minutes, you solve it or you do not, and then you go and do something else. Tomorrow there will be another one. The limit is the point.

As a parent, this matters to me. I would much rather my kid spent ten focused minutes on something that makes them think, and then walked away satisfied, than an hour drifting through an endless feed. A finite thing has a shape. You can finish it. Finishing is a feeling children deserve more of.

There is also something social about a daily puzzle. Everyone gets the same one. You can compare notes at the dinner table. "Did you get today's?" is a better conversation than most.

The playtesting shock

Here is where I got humbled.

I built the first batch of puzzles, played them myself, and was quite pleased. Then I watched other people play them.

Several of them were far too easy.

Some of the drag-and-arrange puzzles were solved almost immediately, with barely a pause. A few of the lateral-thinking ones, which I thought were clever, turned out to be completely straightforward once you read them. The "aha" moment I had designed never came, because there was no struggle before it. People just got it.

That was a useful lesson about my own blind spots. When you design a puzzle, you already know the answer. You cannot un-know it. So you cannot truly feel how hard it is. You imagine a solver going down the wrong path, but you are guessing, and your guesses are biased by the solution sitting in your head.

The only real test is watching someone else.

Difficulty is the hardest part

I now believe that calibrating difficulty is the single hardest thing about making puzzles. Harder than coming up with ideas. Harder than building the app.

Too easy, and there is no satisfaction. You solve it, shrug, and forget it. Too hard, and there is no satisfaction either. You give up, feel a bit stupid, and are less likely to come back tomorrow.

The sweet spot is narrow. It is the puzzle where you are stuck for a while, try a few things, feel a flicker of frustration, and then something clicks. That click is the whole reward. And the window in which it happens is different for every person, every age, every mood.

With eight puzzle types, it gets more complicated. A logic grid and a lateral-thinking riddle are hard in completely different ways. One rewards patience and careful elimination. The other rewards letting go of your first assumption. Someone who flies through one may be stumped by the other. So "difficulty" is not one dial. It is several, and they do not line up neatly.

Puzzles are like children's books

The closest thing I have found to puzzle design is the editing that goes into a good children's book, which I saw up close running a publishing house.

A good children's book is pitched precisely. The vocabulary has to stretch a young reader slightly, but not so much that they lose the thread. The story has to leave room for them to predict what happens, and then occasionally surprise them. Every word earns its place. And you only find out whether it works by reading it aloud to an actual child and watching their face.

Puzzles are the same. Every clue has to be there for a reason. Every red herring has to be fair. The solver has to feel, at the end, that they could have seen it sooner, not that they were tricked. And just like a book, you cannot judge it from the author's chair. You have to hand it over and watch.

The editor's instinct, to cut, to simplify, to remove anything that does not serve the reader, applies almost exactly. Except that with puzzles, sometimes you need to add a little friction rather than remove it.

The joy of productive struggle

Underneath all of this is something I care about deeply as an educator: productive struggle.

There is a kind of difficulty that is good for children. Not the kind that crushes them, but the kind that makes them work a little, try again, think differently, and eventually succeed through their own effort. That success feels completely different from being told the answer. It belongs to them.

Schools talk about this a lot. It is genuinely hard to create in a classroom, because every child is in a different place, and a lesson moves at one pace. A daily puzzle is a tiny, low-stakes way to practise it. Nothing depends on getting it right. It is just you and a problem, and the quiet pleasure of cracking it.

Watching people find my early puzzles too easy was, in a way, watching them be denied that pleasure. I had taken away the struggle, and with it the joy.

Where it goes next

I am still tuning. I keep a closer eye now on which types tend to drift too easy. A version for Orchids students is next, which raises a whole new set of questions about pitching puzzles for different ages without making anyone feel left behind.

What I have learnt, more than anything, is respect for the craft. The logic puzzles, retro games and board games I grew up loving were not just clever. They were carefully calibrated by people who watched a great many players get stuck in exactly the right places.

I am trying to learn to do the same. One puzzle a day.