Mornington Crescent

A long-running British radio comedy game called Mornington Crescent — in which players name London Underground stations while pretending to follow complex, mostly fictional rules — inspires both nostalgia and elaborate in-jokes among fans. Commenters unpack how the “game” actually functions as a social improv framework rather than a solvable puzzle, compare variants and pseudo-rulebooks, and note parallels to everything from esoteric programming languages and AI attempts to play it, to modern corporate rituals and children’s social games.

What Mornington Crescent “is”

  • Several commenters clarify that the “game” is essentially improvisational: players pretend there is a deep, complex ruleset while making up plausible-sounding moves and justifications.
  • The spoken goal (to be first to say “Mornington Crescent”) is deliberately at odds with the real goal (entertaining banter, clever fake rules, and social play).
  • Some participants remain baffled; others explicitly explain that the confusion is part of the joke.

Teaching, kids, and social dynamics

  • One thread describes using a stripped-down version with children to teach anticipation, hidden rules, and reading social cues.
  • The key learning: realizing that “winning” by saying the target station immediately ruins the fun.
  • Commenters compare it to party games where the point is lively interaction, not the final score.

Rule variants and mock “theory”

  • Large subthreads parody an elaborate competitive scene: Atlantic variants, Perridale, forfeiture rules, trackage limits, “nib” and “Nid,” St. Hillary changes, Windsor Appendices, etc.
  • There is faux-serious debate over classic vs. modern rules, “gatekeeping,” time controls, and the effect of new lines or station renamings.
  • Some push back jokingly against “prescriptivism,” arguing the game should evolve.

Computers, esolangs, and AI

  • An esoteric programming language based on Mornington Crescent is linked; someone considers writing an interpreter in Rust.
  • Others mention a TikTok about training an LLM, and share an example of GPT-4 generating convincing in-universe moves and lore.
  • There is humorous talk of “Stockwell”-like engines, Twitch meta-strategies, and even a speculative DeepMind-style solver.

Related games and cultural references

  • Comparisons are made to Numberwang, 43-Man Squamish, and improv games like “Questions,” plus Whose Line Is It Anyway? and Pokémon / corporate Agile as Mornington-Crescent-like social constructs.
  • Multiple links to ISIHAC episodes and YouTube videos show the game in practice.

Reactions and anecdotes

  • Many express affection for ISIHAC’s mix of silliness and intelligence.
  • Several recount surprise at discovering Mornington Crescent is a real Tube station and visiting it as fans.