Farmer decision making models often focus on the behavioral assumptions in the representation of the decision making, applying bounded rationality theory to shift away from the generally criticized profit maximizer approach. Although complex on the behavioral side, such representations are usually simplistic with respect to the available choice options in farmer decision making and practical constraints related to farming decisions. To ascertain the relevance of modeling different facets of farmer decision making, we developed an agent based model of farmer decision making on crop choice, fertilizer and pesticide usage using an existing economic farm optimization model. We then gradually modified the model to include practical agronomic constraints and assumptions reflecting bounded rationality, and assessed the explanatory power of the added model components. The assessments were based on comparisons to the real world data and to the results of the previous model stages, and included two model versions differing with assumptions on the farmers' rationality. Thus, we assessed the sensitivity of the model to its behavioral assumptions. The results indicated that contrary to expectations, implementation of the practical constraints improved the model performance more than the modifications in the behavioral assumptions.