Back to the article

Historical play records

Evidence review: AI Campaign Journal editorial operator. Source windows mark where the review stopped; they do not establish matching versions, settings, causes or current behavior.

  1. Node, 11:10–11:31. Collected October 22, 2026 (Beijing time). Choosing to stay leads to starvation and a game-over message. Observation stops at that message; another input was not observed.
  2. OneyPlays, 17:44–18:24. Collected October 21, 2026 (Beijing time). The death line does not end the recorded encounter. Observation stops after play continues.
  3. Jerma985 (upload: Golden_Winged), 18:02–18:29. Collected November 7, 2026 (Beijing time). A traffic action leads to a bus hit, death narration, waking again and a loop. Observation stops at the announced restart.
  4. Nookrium, 17:15–17:31. Collected October 22, 2026 (Beijing time). Combat continues after the missile injury. Observation stops within the healing decision sequence.

Google organic results checked November 7, 2026

Query: text adventure. Google English / United States parameters. Comparison checked: One historical AI Dungeon death followed by an explicit Game Over message, set against a separate historical AI Dungeon death line after which the players continued the same encounter. Only the ranked destination itself was counted; advertisements, AI Overview, video and image blocks, and question modules were excluded.

  1. Text Adventures — No. Game catalogue and descriptions.
  2. AI Dungeon — No. Product landing page and updates.
  3. Iron Realms — No. Genre guide.
  4. Mobius Digital — No. Outer Wilds prototype.
  5. Steam Curator — No. Curated game listings and reviews.
  6. Wikipedia — No. Interactive fiction history.
  7. Medium — No. ChatGPT adventure creation example.
  8. Reddit — No. Recommendation thread and visible comments.
  9. QuestViva — No. Authoring and play platform.
  10. Perchance — No. Generator interface, starter scenarios and instructions.

Results without the comparison ÷ results inspected = 10 ÷ 10 = 100%. This is a page-level result for the captured search, not an article word share or a claim about the rest of the web.