Designing fair and efficient blockchain reward mechanisms requires going beyond raw execution time to account for behavioral variability. We present a simulation framework for evaluating BCRPs using entropy as a systems-level indicator of reward fairness and stability. Three strategies are assessed on simulated miner profiles \(\:(n=100)\) with log-normal execution times, Laplace-distributed noise, and tercile-based complexity classes: a classical execution-time baseline, “Mining \(\:2.0\) ” (penalizing miner noise and task complexity), and “Adaptive \(\:2.0\) ” (Mining \(\:2.0\) with exponential time decay). Reward distributions are summarized via KDE and ECDF and scored using Shannon, Rényi \(\:(\alpha\:=2)\) , Tsallis \(\:(q=2)\) , and normalized Shannon entropies computed on discretized rewards ( \(\:20\) bins). An interactive Shiny application accompanies the method for reproducible exploration without programming. Across simulations, Adaptive \(\:2.0\) yields the most behavior-sensitive and equitable allocations, achieving the lowest entropy on all four metrics. Quantitatively, relative to the Traditional baseline, Adaptive \(\:2.0\) reduces entropy by \(\:37.5\%\) (Shannon: \(\:2.684\to\:1.678\) ), \(\:37.1\%\) (Rényi- \(\:2\) : \(\:2.218\to\:1.396\) ), \(\:21.0\%\) (Tsallis- \(\:2\) : \(\:0.785\to\:0.620\) ), and \(\:14.6\%\) (Normalized: \(\:0.847\to\:0.723\) ); Mining \(\:2.0\) achieves intermediate improvements of \(\:29.8\%,\:31.7\%,\:17.2\%,\) and \(\:13.9\%,\) respectively. These results provide an evidence-based, deployable framework for evaluating reward fairness in decentralized systems.
