In epistemology, it was generally accepted that knowledge is justified true belief. Call this definition of knowledge the JTB-account of knowledge. In 1963, Gettier raised two counterexamples to the JTB-account. Most contemporary epistemologists find his analysis correct (Weatherson, 2006). As a result, the major task for the analysis of knowledge is to modify the JTB account for the sake of preventing any Gettierian counterexamples. This task is known as the Gettier problem.

Nonetheless, we observe that Gettier's analysis relies on intuitive understanding of several crucial notions on the JTB account of knowledge. In this paper, we attempt to explicate the meaning of these notions by game theory. This allows us to provide a clearer analysis of whether or not the Gettierian cases are real threat to the JTB-account. Moreover, we attempt to show that the Gettierian cases fail to undermine the JTB-account of knowledge.

The presentation is organized as follows, we review the Gettier problem and present our primitive diagnosis of the problem. Then we present a game called the doxastic game to clarify truth and belief. In addition, we make use of the doxastic game to formulate a semantics of belief.Next we f review a popular account of knowledge called the defeasibility account of knowledge (e.g. (Lehrer and Paxson, 1969; Annis, 1973; Swain, 1974; Lehrer 2000; Stalnaker 2006). In this account of knowledge, there is a condition called the no-defeater justification condition. In order to clarify this condition, Lehrer (2000) invented the justification game. We argue against the defeasibility account of knowledge but make use of the justification games to explicate the notion of justification. We call our account of justification the social account.