Activities that move: apples you can count, frogs that hop
Until today, a counting question in the app was a row of emoji characters glued together, like "πππππππ + πππ", and a child had to squint at it like text. Now those questions are drawn as little scenes instead. Apples pop onto a plate one by one, dots appear in a ten-frame and then vanish, and a frog hops along a number line. When the answer is right, the screen bursts into confetti.
Why drawing matters, not just decoration
Young children don't count text, they count things. A long string of identical emoji is hard to count even for an adult: there is nothing to anchor the eye, and it's easy to lose your place halfway. So the new scenes are laid out the way a teacher would lay out counters on a table:
Before: πππππππππππππ
Now: βββββββββββββββββββββββ
β π π π π π β rows of five, so "ten and three"
β π π π π π β can be seen, not just counted
β π π π β
βββββββββββββββββββββββ
Rows of five are a deliberate choice. They let a child see "five, ten, and three more" at a glance, which is the first step from counting one by one towards understanding tens.
Three scenes, each matched to a skill
- Counters on a plate, for counting, for "how many altogether / how many left" with small amounts, and for "which group has more". Objects pop in one after another, so a child can follow along. In a comparison, the two groups sit side by side with a bright "VS" between them.
- A ten-frame, the two-row box of ten cells that classrooms everywhere use. It's used for recognising small amounts without counting them, a skill teachers call subitizing. The dots appear for a moment and then disappear, so the child has to see "four" rather than tap through "one, two, three, four". Once they answer, the dots come back so they can check themselves.
- A number line with a frog, for counting in twos, fives or tens. The known numbers are joined by little arcs that draw themselves one hop at a time. The frog waits on the last one, and a yellow "?" bubble marks where it will land next. On a right answer the frog jumps onto the bubble and the number appears. On a wrong answer the number is revealed too, so the question still teaches something.
Feedback that feels like a game
Every question now slides gently into place. A correct answer gets a short burst of confetti, and the objects in the scene do a little bounce. A wrong answer gets a soft wiggle of the card. There is no red flash and no buzzer, just a nudge, followed by the right answer.
Some children (and adults) find on-screen motion uncomfortable, and most phones and computers have a "reduce motion" setting for this. When it's switched on, the app now respects it completely. There's no confetti, wiggle or hopping, just the same scenes, drawn still. The one thing that stays is the disappearing ten-frame, because that's part of the exercise, not decoration.
Tap the chicks: answering inside the picture
Later the same day, we looked hard at one of the youngest activities, "Which has more?", and weren't happy with it. It showed two rows of objects as two big buttons, so a four-year-old was really picking a button, not comparing groups. It now works the way a teacher would do it on a table. The balloons and chicks are tipped out together, all jumbled, into a bigger play area. The child taps a chick (or a balloon) to say "these ones". Children who use a keyboard or a screen reader get one big round button per kind of object.
The best part comes after the answer, right or wrong. Every object glides out of the jumble into two neat lines, a balloon above each chick, like children pairing up in the playground:
π π π π π π
π€ π€ π€ π€ π€ π€ π€ π€ π€ <- these three have no partner, and they glow
The chicks without a partner light up. That's the idea of "more" made visible: not "the row looks longer", but "some are left over when you pair them up", which is exactly how early-years teachers introduce it. A wrong answer becomes a small lesson rather than a red cross.
A frog that counts (and sometimes falls in)
The last activity we reworked today was "Count in Order": does the child know the counting words come in a fixed order? It used to show four near-identical buttons like "1 2 3 3 4 5" and "1 2 4 5", which only an adult reader could tell apart. Before redesigning it we asked two questions: what would a teacher want a child to practise here, and what can a three-year-old actually use on a phone?
The answer is a pond. Each possible counting is a lane, and each lane has a frog on the bank and a row of lily pads with one number per pad. For short counts the pads also carry dots, so the number can be seen as well as read. The child taps a lane, and that lane's frog sets off:
right lane: πΈ β 1 β 2 β 3 β 4 β 5 every pad lights up, the lane bounces
wrong lane: πΈ β 1 β 2 β 4 π¦ the frog slips where 3 should have been
( 3 ) a ghost pad shows the missing number
When the counting is wrong, the frog stumbles at the exact spot where it goes wrong: a skipped number, a number said twice, or two numbers in the wrong order. Then the frog in the correct lane hops the right sequence, one lit pad at a time. The mistake isn't hidden behind a red cross. The child sees it happen and then watches the right way.
Built once, reused everywhere
The bigger change is behind the scenes. Instead of drawing each activity by hand, the app now has a small "activity engine": a shared set of building blocks (plates of counters, ten-frames, number lines) that any activity can describe and the app knows how to draw and animate. Moving an activity onto it is now a matter of saying what to show ("three apples plus two apples"), not how to draw it. That's why three activities could switch over today. Many more can follow, and new building blocks are already planned: base-ten blocks for place value, a ruler for measuring, a clock face, fraction bars, and later, activities where children drag and tap objects themselves instead of typing an answer.