Double Jeopardy in digital entertainment: a replication corner study using Steam data
本研究使用Steam平台2350款游戏的数据，复制了双重危险定律，发现动作、策略和角色扮演类游戏存在该效应，而模拟和冒险类游戏不显著，并探讨了游戏年龄的影响。
Abstract We replicate the Double Jeopardy (DJ) law using Steam data and a loyalty measure aligned with the DJ construct. Using 2350 games across five genres, loyalty is defined as user-level share of category requirements (SCR) within genre, with multi-genre playtime allocated by equal weights. Spearman correlations with bootstrap confidence intervals show DJ for Action, Strategy, and RPG ( $$\rho $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>ρ</mml:mi> </mml:math> = 0.215–0.281, CIs excluding zero), while Simulation and Adventure are inconclusive. Results are robust to alternative loyalty measures and to restricting SCR to users with multiple games in a genre. Age stratification indicates weaker effects for newer releases. These findings extend DJ to experiential digital products while delineating boundary conditions by genre and market maturity.