iMath is a math-fact fluency world built for this classroom: students earn a daily "Green Light" by practicing facts inside seven mini-games, an adaptive engine tracks true fluency (speed of recall, not just accuracy โ with each child's typing speed calibrated out), and across the term students design their own games on paper, see them built, and iterate like real designers.
One click, no setup: the demo student skips the intro flow and has every booth unlocked so you can tour all seven games freely.
Sessions run ~15โ20 minutes, 3+ days a week. Wrong answers are never punished โ missed facts just come back.
Every picture below is a real screen from this build, in the order a child meets it. The whole first run โ sign-in through placement โ takes about five minutes, once, on day one.
Kids tap their own face on the class grid and type a 4-digit PIN. A brand-new student taps First time here? and makes a player in under a minute โ nickname and a face, nothing else.
New players pick one of eight cube colors. That cube is how they walk around the island, and it is the first thing on the whole site that belongs to them.
Axel, the island's resident axolotl genius, introduces himself, explains that the island is a carnival of math games with booths that open as you play, points at the Test Your Strength tower, and hands the child off to the warm-up. A skip link is on every screen for the kid who has seen it before.
The child types five numbers as fast as they can. The app measures milliseconds per digit and stores that as their personal typing baseline.
Why it matters: fluency here means speed of recall, so the engine has to know how much of a child's answer time was thumbs and how much was thinking. Every later "was that fast?" judgement adds their own typing overhead back in โ a slow typist is never mistaken for a slow mathematician.
The placement test is a carnival high striker โ "Test Your Strengths." Each swing of the mallet is one math fact; ringing the bell means the child has cleared a skill band. It searches five bands, and it does not march up them in order:
Two quick correct answers clear a band and the test jumps upward; two misses drop it back down. It is a binary search, capped at fifteen questions, so most students finish in about two minutes regardless of level.
Two design promises to tell your class: one slow-but-correct answer never counts against them โ a distracted moment is not evidence. And "don't know โ skip" is scored exactly like a wrong answer, no worse, so there is nothing to gain by guessing and nothing to fear in skipping.
Clearing a band credits every fact underneath it. A strong student never grinds 1+1, and a student who is still shaky on their tables never opens the app on 7ร8.
Credited is not mastered. Everything the assessment credits starts as practicing โ it still has to be answered quickly, several times, inside the games before the engine calls it fluent. Placement sets the starting line, not the grade.
A carnival midway the student walks around with arrow keys, WASD, or by tapping the ground. Booths are the games; Axel's School is one-on-one coaching on a fact family they keep missing; the tower is always there for a re-test.
Sky Hopper and Critter Garden are open on day one. The other five booths are padlocked and cost a key โ which is where the daily goal comes in.
๐ข Green Light is the daily goal: answer a target number of different facts correctly in a day. The target is 40 โ but it is capped at how many facts the student actually has in play, so a day-one child with ten facts can still win the day. Hitting it extends their goal-day streak.
๐๏ธ Keys come from Green Light days โ one key per streak day โ and one key opens any booth the student chooses. Booths are earned, never bought, and never gated behind a fixed order.
Three platforms are live at once, each labelled with a fact. Answer one and the frog hops up to it. A wall of fog rises from below the whole time; if it catches you, the climb ends.
The loss condition is announced before the game starts, and the card closes with the promise that makes the whole thing safe: "If the fog catches you, the climb ends โ but everything you learn is saved." No lives, no strikes, no lost progress.
No timer, no chase. Every correct answer grows a flower, and flowers bring visitors: Shelly the snail at 4, Dot the ladybug at 8, Flit the butterfly at 12, Buzz at 16, and Plume at 20 โ which throws the Garden Party.
The Garden Party tracker is the point: a child who finds Sky Hopper stressful gets the identical fluency engine wrapped in something that only ever accumulates. Every classroom needs both temperaments available.
One tab per student, five summary numbers, and a fact family map โ one cell per family, colored fluent / practicing / learning / not started / needs help. Below it, "Needs a hand" lists the families a student has missed three times running, each with the strategy hint to try one-on-one.
Per-student controls sit at the bottom: switch their fact set (full ladder, addโsubtract only, multiplyโdivide only), unlock all games, reset progress, or remove the student entirely.
Alongside the daily fluency work, every student designs a game of their own on paper โ and we build it. The packet is six phases; phases 3 and 4 repeat as often as you like.
โณ Phases 3 and 4 are the iteration loop โ repeat them and designers collect versions (v2, v3โฆ).
Sheets are one focused task each, sized for large 3rd-grade handwriting, with scan-anchor marks so completed packets can be photographed and read back automatically. Nothing assumes a game has coins. Phases 2โ4 work unchanged for a cardboard-arcade unit โ same design loop, cardboard instead of pixels.
| Standard | Title (abridged) | Where it shows up |
|---|---|---|
| 3.OA.C.7 | Fluently multiply and divide within 100; know products from memory | The daily games themselves โ the core fluency engine (add/sub facts likewise for students assigned them) |
| 3.OA.A.1 ยท A.3 | Interpret products; solve problems with equal groups and arrays | Phase 6 Math Brainstorm: planning collectibles as an array or as equal groups |
| 3.OA.D.9 | Identify and explain arithmetic patterns | Phase 6 Math Brainstorm: writing a pattern rule for their own game ("every 4th coinโฆ") |
| 3.G.A.1 | Categories of shapes; recognize and draw quadrilaterals | Phase 6 Math Brainstorm: quadrilateral hunt in their own game art |
| RI.3.3 ยท RI.3.8 | Describe a sequence of steps; connect sentences with first/second/third | Phase 2: the loop is a sequence โ the detective sheet and their own goal โ action โ tricky part |
| W.3.1 | Opinion pieces: point of view, reasons, linking words, conclusion | Phase 5 pitch ("You should play my game becauseโฆ alsoโฆ finallyโฆ") |
| W.3.2 | Informative/explanatory writing with illustrations | Phase 1 paragraph, with the hero drawing as the illustration |
| L.3.1 ยท L.3.2 ยท L.3.3a | Grammar, capitalization/punctuation/spelling, word choice for effect | Every written field; title capitalization; the feelings word bank |
| SL.3.1 ยท SL.3.4 | Collaborative discussion; report with facts and details, clear pace | Phase 4 playtest conversation; the Phase 5 one-minute pitch |
| CA NGSS 3-5-ETS1-1 | Define a design problem with criteria and constraints | Phase 2: naming the goal, the action, and what makes the game tricky |
| CA NGSS 3-5-ETS1-3 | Carry out fair tests and use results to improve a design | Phases 3 & 4 update cards โ observe, change one variable, predict, retest |
Questions or ideas? This project is actively built for this classroom โ feedback goes straight into updates. ยท back to the island