Klein-Goertz-Diehl Rivalry Dataset (RIV 5.10)Klein-Goertz-Diehl 对抗数据集 (RIV 5.10)

The Klein-Goertz-Diehl (KGD) rivalry dataset identifies rivalries by behaviour: recurrent militarized interstate disputes (COW MIDs) between the same pair of states, linked by issue and time density (Klein, Goertz & Diehl 2006). Each row is one spell of hostility between a country pair, 1816-2001: 1,212 spells across 871 country pairs, of which 881 are isolated conflicts (usually one or two disputes, but the authors also place some longer strings here when all of their disputes are tied to a single war, e.g. US-Cambodia 1964-75) and 331 are rivalries under the current coding (246 proto and 85 enduring under the authors' previous classification, carried as a separate column).

Each row carries the pair, begin and end dates, the number of disputes and the dispute density (disputes per year), plus the count of disputes that escalated to a COW war and how many separate wars that amounts to. Unlike Thompson's strategic-rivalry list, which codes rivalries from elite perceptions of mutual enmity and keeps only the most severe competitions, KGD applies no severity floor - hence isolated conflicts are part of the same continuum and low-severity rivalries are included.

Klein-Goertz-Diehl（KGD）对抗数据集以行为定义对抗：同一对国家之间反复发生的军事化争端（COW MID），由议题关联与时间密度串联（Klein, Goertz & Diehl 2006）。每一行是一对国家之间的一段敌对时期，1816-2001 年共 1,212 段、涉及 871 个国家对；其中孤立冲突 881 段（多数仅一两次争端，但作者也会把全部争端都挂在同一场战争上的较长序列归入此类，如 1964-75 年的美柬），按现行编码的对抗 331 段（按作者旧分类另分出雏形对抗 246 段、持久对抗 85 段，以单独字段保留）。

每行包含国家对、起止日期、争端数量与争端密度（每年争端数），以及升级为 COW 战争的争端数与去重后的战争场数。与 Thompson 战略对抗清单（基于精英对"相互敌人"的认知、只保留最严重的竞争）不同，KGD 不设严重度门槛——孤立冲突同属一个连续统，低烈度对抗也被纳入。