The anchoring effect in psychology suggests that people may pay more attention to early information received. Cognitive anchoring effects can lead to unbalanced attention in decision-making, which is rarely considered in computational human performance and human-computer interaction (CHP & HCI). This paper proposed biased reasoning with unbalanced attention to derive autopilot commands in flight. First, we establish a fuzzy reasoning module to compute a binary indicator of attention toward early information. Second, we devise an unequally-spaced sampling window to generate cognition weights on information at different times. Finally, we apply the binary indicator and cognition weights to derive virtual heading commands in autopilot. Simulation results demonstrate that the proposed method can simulate the anchoring effect to reflect the unbalanced attention in decision-making under various autopilot scenarios. We found that individual differences in previous impressions affect attention and reasoning, a breakthrough in CHP & HCI. In summary, our approach represents a new route of biased cognitive reasoning due to unbalanced attention in decision-making.