99 Club Studio · Games & Puzzles Backlog

Last updated: 14 September 2026

This document captures promising puzzle types for future expansion of the Maths Games & Puzzles section.

Selection rule

A puzzle should only be added when it has:

  • a clear mathematical purpose;
  • a proper deterministic/seeded generator;
  • a validity check and, where appropriate, a uniqueness solver;
  • meaningful Easy / Standard / Challenge progression;
  • a pupil-friendly explanation with minimal unnecessary text;
  • strong browser-preview and PDF/print parity;
  • layouts that remain readable on ordinary A4 school printers.

The aim is not to maximise the number of puzzle types. Each new engine should add a genuinely different reasoning mechanic or useful curriculum application.

Curriculum-coverage rule

Puzzle mechanics and curriculum coverage must now be planned together. We should not keep adding attractive logic puzzles while leaving important primary maths content weakly represented.

A puzzle may only claim a curriculum topic when solving the puzzle genuinely practises that mathematical idea. Merely using prime numbers as cell values, for example, does not make a puzzle a prime-number activity unless pupils must recognise or reason about primality.

Particular gaps / under-used areas to target deliberately are:

  • fractions — equivalence, comparison, fraction arithmetic, fractions of quantities, mixed/improper forms;
  • number properties — primes, factors, multiples, common factors/multiples, divisibility, square/cube numbers;
  • decimals and percentages — equivalence and conversion, ordering, arithmetic and percentage relationships;
  • ratio and proportion — equivalent ratios, scaling and sharing;
  • measurement — area/perimeter relationships, units and conversion where a puzzle mechanic genuinely fits;
  • coordinates and geometry — route, transformation and spatial puzzles rather than only isolated questions;
  • statistics/data — generated chart/graph interpretation puzzles where the visual itself is part of the task.

The existing Custom Worksheet curriculum map should be used as the source of truth for year appropriateness. Games & Puzzles should complement it, not duplicate every direct-question family.

Coverage-led variants to build into suitable engines

  • Sumplete: whole-number mode first, then fraction and decimal modes where age-appropriate.
  • Divisibility / Property Maze: primes, factors, multiples, square/cube numbers, divisibility, equivalent fractions, FDP equivalence.
  • Factor / Number Chains: factors, common factors, multiples, prime relationships and multiplication fact families.
  • Fraction Match Grid: fractions ↔ equivalent fractions ↔ decimals ↔ percentages.
  • Honeycombs / Hex Grids: fraction, multiplication and sequence variants.
  • Coordinate Treasure: coordinates, translation/reflection and route reasoning.
  • Region / tiling puzzles: area, perimeter and fractional parts when the geometry is honest and visually generated.

Wave 1 · Highest priority puzzle mechanics

1. Sumplete / Cross-Out Sums

A grid of numbers with a target for each row and column. Pupils cross out numbers so the remaining values make every target.

Why it fits

  • very arithmetic-heavy;
  • simple visual language;
  • easy to scale by grid size and number range;
  • lends itself to exact solution checking;
  • highly printable;
  • suitable for later fraction and decimal variants.

2. Skyscrapers / Towers

A Latin-square style puzzle where each value represents a building height. Edge clues tell how many buildings are visible from that direction.

Why it fits

  • genuinely different logic mechanic;
  • combines ordering, comparison and deduction;
  • natural 4×4, 5×5 and 6×6 progression;
  • deterministic solver can guarantee uniqueness.

3. Binary Puzzle / Takuzu

Fill a grid with 0s and 1s so each row/column contains equal numbers of each, no three identical values are adjacent, and no two complete rows/columns are identical.

Why it fits

  • concise rules;
  • highly generatable;
  • strong logical reasoning;
  • visually clean for print.

4. Killer Sudoku / Sum Sudoku

Sudoku combined with outlined cages whose cells must total a given number.

Why it fits

  • reuses existing Sudoku and Arithmetic Cages concepts;
  • arithmetic + logic in one puzzle;
  • relatively efficient to implement using existing infrastructure;
  • familiar format for teachers and pupils.

5. Bridges / Hashi

Numbered islands are joined with one or two horizontal/vertical bridges. The number on an island gives the total bridges connected to it; bridges cannot cross.

Why it fits

  • very different spatial reasoning mechanic;
  • strongly visual with little text;
  • scalable difficulty;
  • suitable for a dedicated uniqueness solver.

Coverage track · bring forward alongside Wave 1

These should not wait until every Wave 1 logic engine is complete because they address important curriculum gaps.

A. Divisibility / Property Maze

Navigate a grid by selecting cells that satisfy a mathematical rule, for example:

  • prime numbers;
  • multiples of 6;
  • factors of 48;
  • square/cube numbers;
  • numbers divisible by a specified value;
  • equivalent fractions;
  • fraction/decimal/percentage equivalents;
  • values greater/less than a target.

This should be distinct from the existing Correct-Answer Maze because the reasoning is classification/property based rather than a sequence of separate questions.

B. Factor Chains / Number Chains

Arrange or follow numbers so neighbouring values satisfy a rule, such as:

  • share a factor;
  • one divides the other;
  • share a common factor greater than 1;
  • sum to a prime;
  • form a multiplication fact family;
  • satisfy a stated multiple/factor rule.

C. Fraction Match Grid

A logic/grid puzzle linking equivalent fractions, decimals and percentages rather than a simple matching-card exercise. Variants should include fraction-only equivalence for younger pupils and FDP equivalence for Years 5–6.

Wave 2 · Strong follow-up candidates

6. Coordinate Treasure Puzzle

Use coordinates to reveal a route, shape, picture or hidden message. Could build on the existing coordinate-question infrastructure while offering a much more game-like activity.

7. Missing Number Honeycombs / Hex Grids

Hexagonal number structures using horizontal and diagonal rules. Variants could target:

  • addition;
  • multiplication;
  • fractions;
  • sequences;
  • missing-number reasoning.

8. Domino Sum Placement

Place a supplied set of dominoes into a grid so row/column or region targets are satisfied.

This should be a true placement puzzle, not simply the previous domino matching/chain format.

Further candidates

9. Maths Mines / Neighbour-Count Puzzle

A child-friendly Minesweeper-style puzzle where clues show how many hidden stars/gems/marked cells surround each square.

10. Alphametics / Number Code Addition

Letters or symbols stand for digits. Start with compact primary-friendly forms such as A7 + 2B = 95; later consider full alphametics if a robust uniqueness solver is practical.

11. Number Partition / Region Sums

Shade, divide or assign cells into regions so numerical conditions are satisfied. Could support addition, multiplication, area, fractions or target sums.

12. Equation Loop / Arithmetic Loop

A continuous route through numbers/operators where each section forms valid arithmetic relationships. Intended to feel like a puzzle rather than isolated missing-operation exercises.

Possible additional variants / later exploration

  • prime-pair puzzles;
  • ratio/proportion chains;
  • percentage conversion paths;
  • multiplication-table logic grids;
  • arithmetic word wheels / radial number structures;
  • codebreaker variants using number properties;
  • area/perimeter tiling puzzles;
  • coordinate battleship-style activities;
  • symmetry completion puzzles;
  • transformation-route puzzles;
  • graph/data interpretation puzzles where the graph itself is generated;
  • more advanced number-code and cryptarithm variants.

Near-term development order

The next work should deliberately interleave new mechanics with curriculum gaps:

  1. Sumplete / Cross-Out Sums — whole-number engine first; design data model so fractions/decimals can follow without a rewrite.
  2. Divisibility / Property Maze — explicitly cover primes, factors, multiples and related number properties.
  3. Skyscrapers / Towers.
  4. Fraction Match Grid — make fractions a first-class puzzle topic rather than an afterthought.
  5. Binary Puzzle / Takuzu.
  6. Factor Chains / Number Chains.
  7. Killer Sudoku / Sum Sudoku.
  8. Bridges / Hashi.
  9. Coordinate Treasure Puzzle.
  10. Missing Number Honeycombs / Hex Grids.
  11. Domino Sum Placement.

After each engine is implemented, review it in browser preview and PDF before starting the next one. Do not batch many unreviewed engines together.

Regression / quality rules inherited from previous work

  • deterministic for a given seed;
  • no fake difficulty: Easy/Standard/Challenge must change meaningful puzzle properties;
  • all generated answer keys must be validated;
  • unique-solution puzzles must be proven unique by the engine, not assumed from construction;
  • clue placement should be deliberate, not merely random deletion;
  • mixed difficulty must continue to respect pack-wide weighted allocation;
  • engine settings must survive switching modes and re-rendering;
  • browser preview and direct PDF should tell the same mathematical story;
  • printed supporting text must remain readable at normal A4 size;
  • grids/graphics must stay inside their activity frame at every supported size;
  • worked examples must use the same rules as the real engine;
  • do not silently weaken an engine just to make generation faster: use bounded attempts and a valid fallback or fail cleanly;
  • do not tag a puzzle to a curriculum topic unless the solving process actually exercises that topic;
  • preserve existing engines and layout fixes unless a change is specifically required.