Blockchain networks depend on incentive mechanisms to ensure that rational miners/validators follow the prescribed consensus protocol. However, strategic behavior such as selfish mining, validator collusion, and reward manipulation can degrade security, decentralization, and fairness. This paper proposes a game-theoretic framework to model and analyze incentive compatibility in blockchain systems. We formalize the interaction among participants as non-cooperative and coalition games, where agents choose between honest participation, selfish deviation, or collusive coordination. A mathematical payoff model is developed incorporating reward distribution, operational costs, detection probabilities, and penalty/slashing functions. Nash equilibrium conditions are derived to characterize when honest participation becomes a stable equilibrium. We further introduce a fairness-aware reward shaping mechanism that reduces the profitability of selfish strategies while discouraging collusion through coalition-dependent penalties. The framework provides explicit equilibrium constraints and design insights for secure and fair incentive mechanisms in both Proof-of-Work and Proof-of-Stake blockchain networks.