Determinants of Wealth Inequality in the Game of Diplomacy: An economic and computational perspective

<p>This paper presents the first comprehensive analysis of wealth inequality within the strategic gameof Diplomacy, addressing a significant gap in existing literature, which typically focuses on tactics,cooperation, and deception rather than distributional outcomes. Utilizing an anonymized dataset of over1 million records corresponding to players’ winnings (supply centers and all territories) at the end of thegame, we calculate Gini coefficients to quantify inequality. The analysis reveals that Diplomacy exhibitsmoderate-to-high inequality (Gini = 0.56 for supply centers and 0.52 for all territories), comparable toreal-world economies such as Spain or South Korea.Fractional logit and beta regressions reveal that wealth inequality in strategic environments is notmerely a result of individual skills but is structurally shaped by institutional rules, communication, andpath-dependent accumulation. In particular, higher player count and longer game duration are associatedwith increased inequality. Moreover, the possibility to engage in private communication significantly in-creases wealth disparities compared to no-communication environments, as it facilitates complex coali-tions and betrayals. However, high levels of overall communication activity tend to reduce inequality,though inequality in communication itself (dominance by one speaker) correlates positively with wealthinequality at the end of the game. Finally, preliminary results suggest the presence of AI agents has asignificant impact on wealth distribution, confirming that Diplomacy is a relevant laboratory for studyinghuman-AI collaboration and competition</p>