My 2.5-year-old plays Tempopo. Sort of.

He's got the mechanics — he can move things, place things, get through the first levels. He goes back to it on his own, which is more than most things manage. But the difficulty becomes a wall he can't climb, and past that point he isn't playing it anymore. He's just handling it.

That's the good end of what's out there. The other end is what most "toddler games" are: click the balloon, it pops, forever. Two seconds of game on a loop. No wall, because there's nothing to climb.

So the gap isn't the one people usually name. It isn't that toddler games lack depth — depth exists, it just starts above their heads. What's missing is a difficulty curve that begins where a small kid actually is, and rises slowly enough that they can ride it up.

One week. Browser. Aimed squarely at two- and three-year-olds — if you catch yourself designing for a four-year-old, you've drifted. Four reads a little, knows some conventions, is a different player. Two and three is where the gap is.

The one assumption

They understand nothing, and the game has to.

Not the words. Not the icons. Not that a red X means close, that a gear means settings, that an arrow means forward, that a key on the keyboard means anything at all. None of it is installed yet. A small child brings almost no conventions to your game, and every one you rely on is a place it quietly stops working.

This is a description, not a list of bans. You can use anything — you just have to build the understanding for it inside the game rather than assume it's already there. That's the interesting half of the problem, and where the best entries will come from.

The two rules

1. They can play it after being shown. Not blind. A grown-up will sit down, demonstrate, and say "now you do it" — several times, probably across days. That's normal, that's how small kids learn anything, and your game doesn't have to teach itself in one sitting.

What matters is what happens after: can they then do it alone? The actions have to sit inside their motor skills, and nothing can hinge on reading — that one's not negotiable, because no amount of showing fixes it.

2. They can beat something. Then something harder. This is the jam. Making a game get harder is easy. Making it get harder starting from the right place is the whole problem.

Your first real challenge should be one a two-year-old beats on their second or third go — actually beats, not "the game quietly lets them through." Then the next is slightly harder. Then the one after that. Your ceiling can be as high as you like; nobody gets the ceiling wrong. Almost everybody gets the floor too high and the steps too big.

The failure state to avoid: a kid who interacts with your game for a week and never once wins.

Failing is fine. Unbeatable isn't.

Let them fail — difficulty is where the whole thing lives, don't sand it off. A small kid will retry without much drama. Failure isn't what stops them; a wall is.

The cost of a failed attempt should be the attempt, and nothing else. No losing progress, no being sent backwards, no lives, no countdown running out, no scolding or scary feedback. Retry is one click away, or just happens. And nothing bad should follow from them putting it down and wandering off for twenty minutes — that isn't quitting, that's a nap.

Requirements

• Playable in the browser. HTML5, embedded on your itch page. One click to open — no downloads, no installs, no accounts.
• Fully playable with a mouse. Click and drag are both fair game; two-year-olds can drag, and most games for this age assume otherwise. Other inputs are welcome on top — keyboard, touch, whatever — as long as the game teaches whatever it asks for. A game whose actual goal is teaching a small kid the keyboard is very much wanted here.
• No text as a required signal. Letters and numbers are great as content — counting, ABCs, naming things. Never as instructions.
• No ads, no IAP, no data collection, no outbound links. Someone's kid is driving this.
• Sound optional and gentle. Assume it might be muted. Assume it might be on for two hours straight.

Strongly encouraged, not required

• Playtest on a real small human, and watch for the difference between playing and handling. Ten minutes of that teaches you more than the other six days. Post what happened — those devlogs are the best part of a jam like this.
• Say on your page what age you're aiming at and what a real kid did with it.
• Built during the jam is preferred; reusing your own engine, tools, or assets is fine — just note what's pre-existing.
• Any team size including one. Any tool.

One real baseline

Not a rulebook — a snapshot of one actual 2½-year-old, so you have something concrete to design against instead of guessing. Yours will differ.

He can:

• Click. Clicking always works.
• Click and drag, reliably.
• Hit big, obvious targets.
• Follow a demonstration and repeat it days later.

He can't, yet:

• Use the keyboard — individual keys don't mean anything to him.
• Double-click, right-click, or click-and-hold for a duration. All timing, all out of reach.
• Hit small or moving targets, or anything needing precision.
• Release a drag in exactly the right spot. Be generous — snap it, or let it land somewhere sensible.
• Recover when a drag leaves the browser window. Handle that, don't get stuck.
• Wait through a sequence of actions before anything visible happens.
• Avoid stray clicks. There will be a great many; none should be destructive.

The best thing that could come out of this jam is a game that moves something from the second list to the first.

One rule about AI

No generated art or audio in entries. Everything else is fair game — engines, asset packs, libraries, your own older work, anything with a licence that allows it. itch will ask you to disclose AI content when you upload; please just answer it honestly.

When you submit

Set your project's kind to HTML and tick "This file will be played in the browser." If it doesn't run embedded on the page, it can't be judged.

Please tag your game with all four of these, plus whatever else fits (itch allows ten):

toddler · kids · no-text

These are how a parent finds your game a year from now, long after this page has gone quiet. That's the whole point.

Voting

Ranked by community vote on four criteria — see below. Voting is open to everyone, not just entrants, so parents can weigh in. The Climb counts as much as everything else, on purpose. It's the point of this jam.

Prizes

None. The prize is a folder of games you can actually hand to a kid.