A Matriz de Jogos Estratégicos (MJE) como uma nova ferramenta para gestão estratégica via teoria dos jogos

Sistemas & Gestão 01/2009;
Source: DOAJ


Novas utilizações de alguns conceitos da teoria dos jogos para formulação de estratégias cooperativas e competitivas em gestão empresarial são aqui elaboradas e apresentadas, tendo como base a Matriz de Jogos Estratégicos (MJE). Aplicam-se os conceitos e resultados da MJE na análise e formulação de estratégias empresariais, como instrumento de apoio à gestão estratégica em complexas estruturas empresariais–hierárquicas ou não – tanto competitivas como cooperativas. É proposta uma nova tipologia para jogos estratégicos empresariais, derivada da análise e interpretação da MJE: Quatro jogos estratégicos clássicos da teoria dos jogos – Nash, Minimax, e Pareto, como jogos não-hierárquicos, e Stackelberg, como jogo hierárquico – são interpretados e aplicados a situações de conflito de interesses competitivas e cooperativas; dois novos jogos estratégicos hierárquicos, aplicáveis a situações de casos-limite, também derivados da análise da MJE – Dominante-Marginal e Paternalista-Solidário–, são também descritos e aplicados. Uma nova metodologia para análise e formulação de estratégias competitivas e cooperativas para gestão estratégica em complexas estruturas hierárquicas, construída com base na MJE, é apresentada e aplicada a jogos hierárquicos em três níveis. Os conceitos de jogo-de-cena estratégico e de dinâmica de posicionamento estratégico são também introduzidos e ilustrados.

Download full-text


Available from: Eliezer Arantes Da Costa, Sep 24, 2014
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The objective of the current research is the development of a new conceptual model aiming to represent, in an integrated manner, the many situations of conflict of interests as a basis for analysis and design of hierarchical multiagent systems control and for the improvement of the methodology for betterment of the managers’ skills to deal with the strategic management of such conflicts. The investigation method used was a comparative analysis of the unique characteristics of classical games from Game Theory – Nash, Stackelberg, Pareto, and Minimax – searching, among them, their commonalities and differentiations. This investigation identified two distinct dimensions that enabled the conception and construction of a matrix to represent, in a integrated form, those four games mentioned above. The resulting conceptual model provides a comprehensive analytical scheme, inspired in the theory of games, and is used to explain, describe, interpret and forecast behaviors of autonomous agents involved in situations of conflict of interests and, in some cases, to prescribe the more adequate decisions. The Strategic Games Matrix (SGM) proposed and used in this study establishes a conceptual reference framework mapping six different types of games. In it, the assumptions for classic game models, among others for limit-cases, are used in an integrated and complementary manner. The SGM deals with both competitive and cooperative games, as well as balanced and unbalanced ones, taking into consideration both the players´ competitive postures and the power-ratio assumed by each one. The SGM contemplates in an innovative way the treatment of multiple simultaneous strategic sub-games among the agents involved. The application of the SGM concepts to complex systems – hierarchical or not –, with multiple autonomous intelligent interactive agents, provides a methodology of utility for analysis and design of their control strategies. An important part of this study is the exploratory experiments with pedagogical purpose. Such business games, played in a computer, indicate that the participants increase their perception to understand the various games to play, and their ability to act at each one of them. This use of the SGM leads each participant to analyze conflict of interests’ situations and to improve its strategic decisions: Through an interactive dynamic process of trial and error he/she ends up learning how to make better decisions taking into consideration the likely decisions of the other agents involved as well as her/his evaluation of the consequences of their choices.
    Full-text · Thesis · Apr 2008