Cut Share

A row of twelve numbers runs left to right. Some of them are gains and some are losses. Drop three cuts into it — or two, depending on the round — so that every piece between the cuts adds up to exactly the same amount. The pieces are never the same length, so counting cells gets you nowhere; the only thing that decides where a cut belongs is the running total. The board tells you something this puzzle usually hides: how many different cuts actually work, out of every way the cuts could have been placed. Sometimes six arrangements come out level. Sometimes one does, and finding it is the whole point.

It runs entirely on your device: 0 network requests, works offline. No ads, no account, nothing to buy, and no betting, money or prizes of any kind.

1. Focus the board — click it, or press Tab until it is focused.
2. Read the target. The status row says how many shares the row must be cut into and what each one has to total. That figure is just the row's total divided by the number of shares, and it is the number every decision is measured against.
3. Move along the gaps with Left and Right. The cursor sits between two cells, which is the only place a cut can go. Twelve cells means eleven gaps.
4. Cut with Up, or press Enter to cut and un-cut the same gap. Pressing a gap with the pointer does the same thing.
5. Take a cut back with Down or Esc, or clear the row with Z. All three are free.
6. Watch the share totals under the row. Each piece shows what it adds up to and how far that is from the target, so a wrong cut tells you which direction to move it. Moving a cut one cell hands exactly that cell from one share to its neighbour.
7. Submit with F once every cut is down. Submitting with cuts still missing is refused rather than charged — an unfinished answer is not a wrong one.
8. Win when every share totals the target. Each wrong submission spends one attempt, and the round ends when they run out.

How many attempts you get, and what the round is worth, come from the same measurement: how many placements you would have to sift to find one that works. A round with a single answer among 165 placements is generous with attempts and pays up to eight times; a round where a dozen arrangements work is tighter with attempts and pays once.

How do you cut a row of numbers into groups that all add up to the same amount?

Add the whole row up first and divide by the number of groups. If it does not divide exactly, stop — there is no answer, because every group total is a whole number and so their common value must be one too. If it does divide, that quotient is your target. Now walk the row from the left keeping a running total, and this is the entire method: the first cut belongs where the running total equals the target, the second where it equals twice the target, the third where it equals three times the target, and so on. The last group needs no check at all — if every earlier group is exactly the target and the row's total is fixed, whatever remains is the target too. With only positive numbers the running total climbs steadily, so each of those figures is passed once and there is exactly one answer to find. Once losses are in the row the running total goes down as well as up, it can pass the same figure twice or three times, and picking the wrong one of those positions leaves the rest of the row impossible. That is where the puzzle lives.

Is every row actually cuttable?

Yes, by construction, and it is not a hope. The shares are built first: the game draws how many shares the round has and what each must total, draws a set of block lengths that fill the row and are never all equal, fills each block with values that add to exactly that total, and then concatenates them. The block boundaries are therefore a working cut before the row has even been shown, and there is no generate-and-reject loop anywhere that could stall or hand you an impossible row on an unlucky run. Every dealt row is then checked again by an exhaustive scanner that never sees the blocks the generator used, so the guarantee is verified rather than assumed. If you run out of attempts, the board moves the cuts to a working answer so you can see the arrangement instead of being told about it.

Why do some cells have losses in them?

Because without them there would be nothing to think about. A row of positive numbers has a running total that only ever climbs, so it reaches the target once, twice the target once, and so on: one sweep from the left finds the only possible answer the moment you have done the division. Adding losses makes the running total go up and down, which means it can arrive at the same figure more than once — and now the question is not where the running total reaches the target but which of the places it does leaves the rest of the row workable. It is also the only reason a round can have more than one answer, which is what the difficulty rating measures. A positive-only row would be the same round every time.

What does the "ways" number mean, and why does it change the score?

Before the round starts, the game walks every way the cuts could be placed — 55 arrangements for two cuts, 165 for three — and counts the ones that come out level. Both numbers are shown. It is a real measurement taken by enumeration, not a difficulty label someone typed in. Dividing the second by the first gives how many arrangements you would sift per working answer, and that single figure sets both the attempt allowance and the score multiplier. Measuring it that way rather than by the raw count of answers matters, because two answers out of 55 is a far easier round than two out of 165; comparing them honestly is what lets the number of shares change from round to round.

Is there any money, betting or in-app purchase in this?

None, and there never will be. The cells are plain numbers on tiles: no currency, nothing to buy, and no gambling mechanic of any kind anywhere in the game. The score exists so that a reasoned cut feels better than a lucky one, and it is not for sale.

Can I play it with a keyboard or a screen reader?

Both. Left and Right move along the gaps, Up cuts, Down and Esc take a cut back, Enter does both, Z clears the row and F submits. Every cell announces its value with its sign; every gap announces whether it holds a cut and which cut it is; and the board announces the target, how many cuts are down, what each share currently totals, what the next cut is looking for and how many attempts remain. A cut is drawn as a bar and an empty gap as a dot, neighbouring shares sit on alternating backings, and each share total carries a tick or a "not equal" sign, so nothing at all depends on telling one tint from another.

Does it work offline?

Yes. Once the page has loaded once, the whole game is on your device — the generator, the exhaustive scanner, the rules and this page. Nothing is fetched while you play, and no result leaves the device.

Twelve cells, each a whole number from -9 to 9, cut into three or four shares whose target is a whole number from 6 to 15. Blocks run from two to five cells and are never all the same length. The count of working cuts is exact rather than sampled, because 165 arrangements is small enough to enumerate honestly. Attempts run from two to five and the score multiplier from one to eight, both derived from the same measurement. The cells are plain numbers with no unit — nothing in the game depends on their being rupees or kilograms, and inventing a unit would suggest a precision the puzzle does not have. There is no hint button and no partial credit: a cut is level or it is not.

Weight Split is the same arithmetic without the row: you divide a set of weights into two groups of equal total, and because any weight can go on either pan there is no running total to follow. Sum Links asks for a set of touching cells that reach a target, so geometry does the pruning that the running total does here. Target 24 works the other way round again — a fixed set of numbers, a fixed answer, and the search is over the operations rather than over where to cut.

Bookmark this page (Ctrl+D, or ⌘D on Mac) or install the app — it works offline the next time you need it.