Studies of adaptation strive to understand the relationship between genotype, phenotype, and fitness. Because mutations are the ultimate source of adaptive evolution, the distribution of fitness effects of mutations (DFE) provides critical information about adaptive processes. Fisher’s Geometric Model of Adaptation (FGM) is a leading candidate for modeling adaptation because it captures the statistics of complex gene interactions within a simple framework of mutation and selection. Despite its success fitting experimental data, the model’s credibility is weakened because it does not explain all of the available data. FGM suffers an inherent ‘cost of reality’. It cannot simultaneously satisfy a realistic DFE and maintain a realistic rate of adaptation. My work addresses if improved biologically plausibility of the underlying mutational and selective regimes can alleviate the cost of reality. I test multiple modifications to FGM and ultimately find that restricted mutational pleiotropy successfully relieves the cost of reality when a small proportion of traits are maladapted. This result harmonizes with previous work that shows restricted pleiotropy to be beneficial when few traits are maladapted because it allows those traits to adapt without disrupting the entire system. My work improves the statistical fit of a popular model of adaptation and finds evidence for an empirically supported mutational regime.