Metagaming and Metagames in Esports

The paper argues for a unified definition of the metagame in esports as the perceived optimal or dominant playing strategy at a given moment, with metagaming describing the use of information external to the game to shape play. Its authors contend that accurate esports analytics, casting, and machine learning depend on understanding the prevailing metagame, since the same character or statistic can carry different meanings as the dominant strategies shift. The concept is traced from its origins in game theory through leisure-game scholarship, with the authors favoring a view of the metagame as the surrounding ecosystem of strategies, expectations, and player knowledge.

The analysis makes the case that strategies such as deliberately losing to gain a later advantage and queue dodging are legitimate metagame tactics, supported by examples from chess, Dota 2, League of Legends, traditional sports, and the 2015 LPL Oscar Night incident. A central claim is that the esports metagame is shaped by perceived rather than actual strength, and the paper distinguishes several forms of metagaming, ranging from exploiting opponents to rules lawyering.

Empirically, the work demonstrates that win rates alone can mislead: Dota 2's Death Prophet, for example, shows a 68% win rate despite a pick rate of only about 5.6%. Drawing on API data from more than 35,000 matches, the authors show that metas vary by skill bracket and shift across patches, using the Fano Factor as a measure of choice variability. They describe metagame evolution as a cycle of ignorance, understanding, and stability, disrupted by patches, novel strategies, regional differences, and developer economic incentives.

The paper identifies three developer motivations for altering the metagame: revenue generation through new characters and buffs that drive cosmetic sales, player retention through novelty and mastery motivation, and ecosystem preservation through competitive balance, as illustrated by Dota 2's nerfing of the Io core. The authors acknowledge that the revenue claims remain largely anecdotal and that the scope of their psychological and economic framing is limited. The bibliography draws on patch histories across several titles, sports labor economics, game theory, and psychology research on boredom and mastery.