In this paper, we reevaluate leakage resilient cryptography from the viewpoint of game theory. We provide a general framework on leakage resilience in game theoretic setting and show that it is of each party's interest not to deviate from the protocol, i.e., their behavior is rational which assures them the highest incentive. We show that it is enough for a Public Key Encryption (PKE) scheme to be Computational Threat-Free Nash Equilibrium (CTFNE) for being leakage resilient. To be specific, CTFNE is not guaranteed if an internal state of a CCLA1 (Chosen Ciphertext with Leakage Attack)-secure Key Encapsulation Mechanism (KEM) of PKE is updated using a poor randomness with low-entropy. This work thus allows us for arguing about the strategic behavior of players in protocols where leakage is allowed.