It's Wordle, but for equations. There's a hidden equation, and you have six guesses to find it. Every guess has to be a valid equation itself — it has to actually balance — and after each one the tiles tell you which numbers and symbols are in the right place, which are in the equation but misplaced, and which aren't there at all. You narrow it down clue by clue until you crack it.
A fresh daily puzzle lands each midnight, plus unlimited practice play whenever you want it. Send a friend the exact puzzle with the challenge link.
Green, yellow, and grey feedback applies to digits and operators alike, so a well-built guess tests arithmetic structure, not just numbers. Guesses must balance — the equals sign is a hard rule, not a suggestion — which is what separates a probe from a wasted row. Seven difficulty levels stretch the equation length from Beginner to Cyborg, and the daily equation is one shared secret across every player on the site.
Math Match is Wordle for equations, and everyone describes it that way — including me, just now — but the description hides the design problem that made this game genuinely hard for me to build. Wordle gives feedback on letters. This game has to give feedback on two fundamentally different kinds of symbol at once: digits, which carry value, and operators, which carry structure. Colour them with one undifferentiated scheme and the feedback turns to mush — a yellow that might mean "right digit, wrong place" or might mean "this plus sign exists somewhere" tells you almost nothing you can act on.
So the call that makes the whole game work: operators and digits are coloured as separate systems. Right symbol in the wrong place reads differently from a wrong symbol entirely, and the operator feedback stands apart from the digit feedback, so every guess returns two clean channels of information instead of one muddy one. It sounds like a small presentation choice. It's the difference between a deduction game and a mood ring. Once the channels were separated, testers stopped squinting at the colours and started actually reasoning — which is the moment I knew the game existed.
Now the mistake, and it's a beautiful one because it's imported directly from Wordle expertise: the probe guess. Strong Wordle players open with sacrificial words — burn a turn, harvest letter information, no intention of being right. Players carry that instinct here and it fails for one structural reason: every guess in Math Match must be a valid, balancing equation. You can't play the equation equivalent of ADIEU — a stack of untested symbols that doesn't need to make sense — because the game rejects anything that doesn't compute. An invalid probe isn't a clever information play. It's a burned turn that taught you nothing, a quarter of your budget spent on a rejection message.
The discipline that replaces it: every guess must do double duty. Valid equation first — it has to balance, always — AND maximum new information second. Build guesses that test unproven digits and untried operators while still computing cleanly. It's a genuinely harder constraint than Wordle's, and it's exactly what makes this game deeper than its inspiration: your information-gathering has to be disguised as correct arithmetic, every single turn.
So think like a scientist with a lab budget: a handful of experiments, each one required to be internally coherent, each one designed to eliminate the most possibilities. The daily equation lands at midnight, the same for everyone. Balance first, probe second — inside the same guess. That's the whole art of it.
A last habit worth stealing: track the operators separately in your head from the digits, exactly the way the colours do. Most stuck positions resolve the moment you realise the structure is already proven and only the values remain in question — or the reverse.
Your first guess shouldn't try to win — it should try to learn. I open with an equation that uses a wide spread of common digits and an operation or two, purely to find out what's in play. Burning the first guess on information instead of a hopeful answer is what sets up a clean solve, the same way it works in Wordle.
After that, let the math do the narrowing. Once you know a couple of numbers are present, there are only so many ways they can combine into a valid equation that fits the green positions — and most candidates get ruled out by simple arithmetic before you ever type them. So I work it on paper in my head: which equations are even possible given what's confirmed, then which of those fit the clues. The constraint that every guess must balance is a gift; it shrinks the field fast if you let it.
Math Match makes you reason backwards — from clues to the answer that has to fit them — while keeping the arithmetic honest the whole way. It's deduction and calculation working together. And a note on why this exists at all: we built the harder variants with more to crack precisely because we love maths and wanted the puzzle to push you, not just pass the time. Cracking a hidden equation from clues is real step-by-step reasoning — rule something out, confirm something else, close in on the only answer that fits. That's the same disciplined thinking that solves problems far bigger than a number row, and the bigger puzzles are built to stretch it.
You're learning to use partial information efficiently and to eliminate possibilities with math rather than guess at them — a genuinely useful habit of mind. The daily keeps it fresh and the practice mode lets you drill the technique. No download, no sign-up — play in your browser on any device. If you like this, try Emoji Math next.