The feedback loop
Every good puzzle gives you information. Mastermind: coloured pegs tell you correct colour and position. Wordle: green, yellow and grey tiles constrain the solution space. The 1940s Bombe machines did the same at scale: each rotor setting eliminated millions of Enigma possibilities until only one remained. The loop is identical — guess, evaluate, eliminate.
Constraint satisfaction
Sudoku is pure constraint: each digit must appear once per row, column and box. Code-breaking adds the constraint of adversarial design — the encoder tried to make your job hard. The resulting puzzles feel richer because the constraints are not arbitrary; they were placed by an intelligence. Games like Cryptogram and Codeword capture this directly.
Pattern recognition training
Bletchley Park hired chess masters, crossword champions and linguists because pattern recognition transfers across domains. Modern brain-training apps sell the same idea under different branding. Puzzle games are low-stakes practice for high-stakes recognition: the same visual cortex that spots a Wordle green tile spots anomalous radar tracks or fraudulent transactions.
The gamification of intelligence
Today, satellite imagery analysis is gamified on platforms like Zooniverse. Protein folding was crowdsourced through Foldit. The UK GCHQ runs an annual Christmas puzzle that thousands solve. The boundary between professional analysis and recreational puzzle has never been thinner — and the skills are the same.