Immersion Level and Bot Player Identification in a Multiplayer Online Game: The World of Warships Case Study

This study investigates whether a player's level of immersion in a multiplayer online game affects their ability to distinguish human opponents from bots, using World of Warships as the setting. Thirty participants, evenly split between experienced and novice gamers, played matches against either bots or human players. The central hypothesis held that higher immersion would reduce accuracy in identifying bot opponents, but the results did not support it. The research draws on the Turing test tradition and on Strojny and Strojna's (2014) immersion questionnaire to frame its inquiry into perceived human likeness in game agents.

The immersion measure yielded a mean score of 96.87 (SD = 15.61), with experienced players (M = 95.53) and casual players (M = 98.20) showing no meaningful difference. Immersion levels did not differ significantly between groups (Mann-Whitney U = 89.5, p = 0.34), and identification accuracy likewise showed no significant group difference (p = 0.534). Descriptively, experienced players identified opponents correctly more often than novices (73.33% versus 53.33%), a trend in the direction the authors predicted for identification, though not statistically reliable.

The authors interpret the null findings with caution, attributing them largely to the small, purposively selected sample, the brief gameplay sessions, and participants' unfamiliarity with the game. The immersion trend ran opposite to the prediction, suggesting that the assumed inverse relationship between immersion and bot detection may not hold under these conditions. The study therefore contributes a tentative, methodologically constrained picture of how player experience relates to human-bot discrimination, rather than a definitive answer.