The theory of social games: outline of a general theory for the social sciences

The paper addresses the criticism that sociological grand theories are too abstract and cannot explain phenomena or guide empirical work. It advances the idea that a sociological general theory can be written around the concept of a social game, used not as a metaphor but as a heuristic center for theory-building. A general theory is conceived as a set of interlinked concepts that guide thinking and can translate into middle-range theory; it should not be self-contained or immunized from empirical falsification. The goal is to show that the theory of social games is as general as competing grand theories but provides a more straightforward link to middle-range theorizing and empirical research via a descriptive-interpretive heuristic, an explanatory heuristic, and formal and agent-based modeling. The author constructs the theory by abstracting from simple games-for-fun (e.g., noughts and crosses, chess) to generalize their properties to social games. The paper integrates insights from sociology (e.g., Goffman, Garfinkel, Elias, Coleman), as well as economics/mathematics (game theory), philosophy (Searle, Wittgenstein/Winch), biology/evolutionary social science (play in development), anthropology, and cultural/game studies (Huizinga, Caillois, Klabbers). It confronts standard criticisms that real-life social situations are too complex, rules are tacit/contested, and consequences are serious (not make-believe) by arguing for a broader definition of social game that accommodates such complexity. The contribution is an outline of a unifying scheme for qualitative, quantitative, and formal/agent-based modeling in sociology.

The paper situates itself amid longstanding debates on grand theory (Merton, Parsons) and critiques that such theories are too abstract and explanatorily weak (Münch; Van den Berg; Goldthorpe). It draws on: (a) sociological traditions—Goffman on performances/games in interaction; Garfinkel on layered/tacit rules and trust; Elias on game models linking individual and society; Coleman on emergent outcomes and micro-macro links; Boudon, Bourdieu, Fligstein & McAdam, Merton, Weber on fields, capital, unintended consequences, and rule legitimation; (b) economics/mathematics—von Neumann & Morgenstern’s game theory for strategic situations; (c) philosophy—Searle on constitutive rules and social reality; Wittgenstein/Winch on understanding via games; (d) biology/anthropology—play as learning/adaptation (Bateson; Gray); (e) cultural/game studies—Huizinga on culture as play; Caillois on game typologies; Klabbers on simulations and digital worlds. It also reviews criticisms of applying games to social life (Bourdieu; Garfinkel; Giddens; Rawls; Goffman; Maynard), arguing many games-for-fun are complex, ambivalent, and consequential, and that a broadened conception of social game addresses these critiques.

Theory-building approach using abstraction from games-for-fun to a general model of social games. The author specifies core elements: players (actors accepted as players, roles, attributes/resources), resources (all means aiding game goals), actions (moves recognized as part of the game), goals (final/intermediate, competitive/non-competitive, individual/team), rules (constitutive/regulative; prescriptions, preferences, permissions, proscriptions; legitimacy and sanctioning), representations (symbols and languages naming elements, reflexive discourse, and in-game language), objects (material entities symbolizing/constituting game elements), space and time (game-defined beginnings, internal temporal structure, bounded spaces), outcomes (states/events, statistics, covariances, dynamic functions), and context (external phenomena relevant to play). It posits an actor model, homo ludens, combining norm-following and bounded rationality, language capacity, need satisfaction, game recognition/internalization of goals/rules/representations, identity formation through play, and efficiency-seeking across multiple games. Empirical linkage is proposed via three methodological devices: (1) Descriptive-interpretive heuristic—systematic questions to reconstruct a social game’s elements, nesting/coupling, and context, implemented qualitatively (participant observation, interviews, documents), with a modeling prompt (what minimal elements would a board/computer game need to reproduce the observed dynamics?). (2) Explanatory heuristic—toolbox of mechanism-oriented hypotheses tied to game elements, illustrated with three rule-related mechanisms: H1 Rule change (enforced rule changes alter behavior but often yield unintended effects), H2 Absence/overuse of sanctioning (anomie and potential game collapse), H3 Rule advantage/social closure (incumbents erect barriers to retain benefits). It differentiates reconstructive (historical-genetic) and statistical explanations grounded in counterfactual causality. (3) Formal and agent-based modeling—use of classical/evolutionary/behavioral game theory and ABM to clarify deep structures (e.g., prisoner’s dilemma), produce ideal-type benchmarks, uncover assumptions, and simulate parameter-outcome mappings. The methodology is illustrated by reconstructing Blau’s study of a job-referral bureaucracy: a monitoring-rule intervention produced unintended competitive behaviors (dirty tricks) in section A, while norms and supervision emphasis in section B sustained cooperation and higher productivity; the underlying structure resembles a prisoner’s dilemma constrained by norms and supervision.

- Proposes a general theory: social life consists of interlinked social games—entities where players with resources engage in actions shaped by goals, rules, and representations, involving objects, within game-defined space/time and contexts, producing outcomes. - Ontology: social games are simultaneously real and socially constructed; their reality hinges on intersubjective acceptance (e.g., money’s value) yet yields real consequences. - Forms and levels: distinguishes games-for-fun vs serious games; identifies multiple levels/types (interactions, groups, organizations, networks, movements, milieus, markets, societal subsystems), with complex nesting/coupling across everyday life. - Actor model (homo ludens): integrates norm-following and instrumental rationality; actors learn and internalize games, align personal motives with game goals, form identities via performance/identification, and seek efficient satisfaction of needs across games. - Detailed specification of game elements: players/roles/attributes; resources (objects, cultural knowledge, social capital, positional/corporal/mental attributes, and game/context attributes); actions (moves); goals (final/intermediate; competitive/non-competitive; individual/team; motives vs goals); rules (constitutive/regulative; legitimacy; sanctioning and repair); representations (naming, reflexive discourse, in-game language; symbolic centrality); objects (as symbols and resources); space/time (bounded, sequenced, ritualized); outcomes (events, statistics, covariances, dynamics/functions; latent functions); context (external conditions relevant to play). - Empirical linkage and mechanisms: introduces descriptive-interpretive and explanatory heuristics plus formal/ABM as bridges to middle-range theory and empirical testing. Example mechanisms and data points: (1) Football offside rule change in 1925 increased Football League goals from about 4700 (1924–25) to 6373 (1925–26), with unintended tactical shifts (defenses deeper; more winger use). (2) Table tennis 2000 larger-ball rule increased average rally exchanges via slower ball speed (Djokic et al., 2019). (3) Blau’s bureaucracy case: individual productivity monitoring triggered competitive defection (dirty tricks) in one section but cooperative norms in another; cooperation yielded higher section-level productivity; interpretable as prisoner’s dilemma moderated by norms/supervision. - Advantage over competing grand theories: comparable generality but clearer pathways to qualitative description, mechanism-based explanation, and formal modeling; integrates symbolic and causal dimensions and instrumental and normative action.

The theory addresses the charge that grand theories lack empirical traction by centering analysis on social games, whose elements are observable and reconstructable. By specifying players, goals, rules, representations, resources, objects, space/time, outcomes, and context, it provides a concrete heuristic for qualitative coding and model building, enabling translation into middle-range propositions and testable mechanisms. It reconciles key dichotomies—real vs constructed social reality, norm-following vs rational action, meaning vs causality—by treating mechanisms as interlinked symbolic-causal processes within games. The illustrative cases (sports rule changes, organizational monitoring) show how rule and object parameters produce measurable outputs and unintended consequences, demonstrating reconstructive and statistical explanatory strategies. The approach also enhances comparability across disparate domains (e.g., organizations vs disaster evacuations) by re-describing them as social games, facilitating cumulative knowledge and integration with formal and agent-based models that reveal deep structures (e.g., prisoner’s dilemma) and counterfactual dependencies.

The paper outlines a grand theory—the theory of social games—that matches the generality of systems, practice, discourse, and structuration theories while offering a more direct bridge to empirical research. It codifies a comprehensive game schema and two heuristics (descriptive-interpretive, explanatory) alongside formal/agent-based modeling to guide qualitative reconstruction, mechanism testing, and formal analysis. The theory integrates and reframes insights from multiple traditions (analytical sociology’s mechanisms, Searle’s constitutive rules, Luhmann’s levels, Weberian closure) by emphasizing symbolic-causal game mechanisms and multi-level nesting of games. Future work should deepen typologies of games (interaction, group, milieu, market, subsystems), analyze nesting/coupling, elaborate the heuristics (including trust and power), and expand empirical applications to demonstrate utility across qualitative, quantitative, and modeling paradigms.

The article is an outline rather than a full exposition. It only sketches game types and interlinkages (nesting/coupling), does not present the descriptive and explanatory heuristics in full, and does not develop treatments of trust and power. While formal game theory and agent-based modeling are established, the proposed descriptive and explanatory heuristics must still prove their usefulness empirically. Not all aspects of the social world are games; many elements (e.g., rules, players, goals) are not games themselves, and the life-world comprises complex couplings of multiple games.