The reading corner

A practical method for puzzles with several totals

Updated

Turn a crowded selection puzzle into a sequence of smaller decisions using bounds, remainders and careful swaps.

A selection puzzle can look like an invitation to try combinations until something works. You have a group of objects, several numbers on each object, and a target for every number. A selection that looks excellent in one column may miss badly in another. The difficulty comes from meeting all the conditions together, rather than from any individual addition.

In Parcel Prism, you choose four of nine cards to match Heat, Lift and Glow totals. There is no timer, and selecting a card does not submit an answer. That gives you room to reason before checking. The method below works with that board, but you can also use it for other small puzzles where a fixed number of choices must satisfy several constraints.

Write the conditions before making choices

First separate the rules from your guesses. How many objects must you choose? Can an object appear twice? Must the totals match exactly? In Parcel Prism the answers are four, no, and yes. A set of three cards with perfect totals is still incomplete. Four cards that match only two of the three totals are still wrong.

Keep a consistent order when reading values. Say “Heat, Lift, Glow” each time, rather than switching between whichever number catches your eye. If you make notes, label the columns. A surprising number of apparent dead ends are simple reading mistakes: a correct subtraction applied to the wrong column, or the same card accidentally used in two places.

Find the tightest condition

An unusually low target limits how many large values you can include. An unusually high target limits how many small values you can include. Start with the column that seems hardest to satisfy. This is a way to reduce the possibilities; it is not a reason to ignore the other columns.

Suppose an invented board asks for four cards whose Heat totals 9. A card with Heat 7 would leave only 2 for the remaining three cards. If every card has a Heat value of at least 1, that is impossible. You can reject the card immediately, without trying combinations that contain it. The conclusion follows from a lower bound, not from a guess about what looks likely.

Use the remaining cards to make stronger bounds

After choosing one or two tentative cards, subtract their values from the targets. Then ask whether the remaining slots could possibly supply those remainders. Use only cards that are still available. If two slots remain, add the two smallest available values in a column and the two largest. The remainder must fall within that interval.

For example, suppose two available slots need 13 Glow. If the largest two Glow values left are 5 and 6, that branch cannot work. Undo a tentative choice. Passing this test does not prove a solution exists, because the best cards for Glow might have unsuitable Heat. Failing it does prove that the current branch cannot finish.

Let a remainder identify the final card

When you have three tentative cards, calculate exactly what the fourth must contribute. Imagine targets of 18 Heat, 20 Lift and 15 Glow, with current totals of 13, 14 and 12. The missing card must be exactly 5, 6 and 3 in that order. A card with 5, 6 and 4 does not count as almost right.

Look for the whole triple among the unselected cards. If it is absent, the current three cannot be part of a completed set together. Change one of them and recalculate. This is often easier than repeatedly adding four complete cards. On paper, record the missing triple beside your tentative three so you can distinguish a tested branch from an untested one.

Think about swaps as differences

Sometimes a set already matches two columns but misses the third. Resist replacing a card simply because its third value looks better. Work out the difference the replacement must make. If Heat and Lift are correct while Glow is 2 too high, a single replacement must preserve Heat and Lift and reduce Glow by exactly 2.

If no available card makes that difference, one swap cannot fix the set. You may need to replace two cards. Compare the combined values of the pair leaving with the pair entering. This turns a vague search for better cards into a precise question. It also explains why a promising set may be a poor starting point despite matching two targets.

Keep a simple search order

If you need to search systematically, label cards by their displayed positions. Consider groups in increasing order, so a group containing cards 1, 2, 4 and 7 is considered once. Reordering those same cards produces no new selection. You do not need to list every possible group upfront; just use an order that prevents repeating work.

Before checking, confirm the count, the three totals and the identity of each selected card. If you discover an arithmetic error, correct the method rather than treating it as evidence that the puzzle is unfair. A useful finish is being able to explain why the chosen set works. The check button then confirms a conclusion you have already reached.

Browse all articles