Worksheet generator

1.D — 19KA-0018Complementary counting

• AI counted the codes that avoid 3 entirely and gave that as my answer without subtracting.
• BI counted the codes with a 3 in the first position only and forgot the other two positions.
• CI counted the codes with exactly one 3 and forgot the ones with two or three of them.
• EI counted every possible code and forgot to remove the ones with no 3 at all.

2.C — 4KA-0040The pigeonhole principle

• AI answered with the lucky case, where the first two socks happen to match.
• BI used the number of colours as my answer without adding one for the sock that must repeat.
• DI worked from the number of socks of each colour instead of the number of colours.
• EI took the whole drawer, guaranteeing a pair but far more socks than are needed.

3.A — 12KA-0071Overcount, then correct

• BI divided the 120 arrangements by 6 rather than by the 10 movements that leave a bracelet looking the same.
• CI allowed for turning the bracelet but forgot it can also be flipped over.
• DI halved the 120 for flipping but forgot that turning also gives the same bracelet.
• EI counted every arrangement in a line, as if the bracelet had a fixed first bead.

4.C — 12KA-0090Counting cubes in a stack, including hidden ones

• AI counted the middle cube of each face, which has exactly one painted face rather than two.
• BI counted the corner cubes, which have three painted faces.
• DI counted every cube that is painted at all except the corners, without separating one face from two.
• EI gave the total number of small cubes rather than the ones with exactly two painted faces.

5.B — 2KA-0094Weighing and balance puzzles

• AI assumed one weighing could separate nine possibilities, but it has only three outcomes.
• CI split the coins into halves each time, which wastes the balance's third outcome.
• DI weighed the coins one against another in pairs rather than in groups.
• EI compared each coin with a known good one in turn, which always works but is nowhere near the fewest.

6.D — No, because the number of moves is oddKA-0097Parity arguments

• AI assumed enough moves can undo anything, without asking what each move preserves.
• BI tried a few sequences that nearly worked and assumed a better choice would finish the job.
• CI reached for the number of items rather than for what a single swap actually changes.
• EI thought the choice of swaps could change the outcome, when every swap has the same effect.

7.C — No, the last number is always evenKA-0098Invariants and monovariants

• AI found sequences ending in small numbers and assumed 1 was among the reachable ones.
• BI assumed a hard-looking question must have exactly one answer.
• DI found one sequence ending in 0 and assumed every sequence must end there.
• EI thought the order could change the outcome, when the quantity that decides it never changes.

8.B — 11KA-0112Mixtures and weighted averages

• AI averaged the two averages, ignoring that the classes are different sizes.
• CI weighted the averages but used the wrong class size against each one.
• DI leaned towards the larger class's average without computing the totals.
• EI added the two averages together instead of combining them.

9.A — 1/10KA-0125Spot the trick vs. decide to grind

• BI cancelled the chain but stopped one factor early, leaving a 9 on the bottom.
• CI worked out the first bracket and assumed the rest made no difference.
• DI gave the last bracket on its own instead of the whole product.
• EI assumed everything cancels completely, leaving nothing behind.

10.B — Answer it with your best guess, mark it, and move onKA-0126Time triage: now, later, or guess

• AI let the time already spent decide, but that time is gone whatever I do next.
• CI moved on but left a blank, which scores nothing if I never get back to it.
• DI spent time re-reading work I had no reason to doubt, while 21 unseen questions waited.
• EI assumed the hardest questions are the best use of time, when the unseen easy ones are worth the same each.

11.D — 49KA-0138Growth patterns and figurate numbers

• AI continued the pattern to the fifth square instead of the seventh.
• BI stopped one square short, giving the sixth instead of the seventh.
• CI added 6 to the sixth square's 36, treating the differences as constant.
• EI continued one square too far and gave the eighth.

12.B — 9KA-0158Last-digit behavior of products and powers

• AI used the first entry of the cycle because 2026 is even, without working out where in the cycle it lands.
• CI counted the cycle starting from zero, so a remainder of 2 put me one place too far along.
• DI assumed a power that large must end in 1.
• EI multiplied 3 by 2 and used the last digit of that.

13.A — 2 hours 24 minutesKA-0159Combined work rates

• BI averaged the two times, but two taps together must be faster than either one alone.
• CI added the two times, which would be right for filling two tanks one after the other.
• DI halved the smaller of the two times, guessing that two taps must simply be twice as fast as one.
• EI found 2 hours 24 minutes correctly and then halved it a second time.

14.A — 8KA-0175The pigeonhole principle

• BI used the number of possible remainders without adding one for the pair that must collide.
• CI doubled 7, thinking I needed two full sets of remainders.
• DI doubled 7 and added one, applying the pigeonhole idea to the wrong number of boxes.
• EI guessed from small cases without identifying what the boxes actually are.

15.A — 54KA-0177Overcount, then correct

• BI counted every line joining two vertices and forgot to remove the 12 sides.
• CI counted each diagonal from both of its endpoints and forgot to halve.
• DI used 12 x 10 without halving, double counting every diagonal.
• EI subtracted 24 rather than 12, removing each side twice.