Automated aids are often provided to assist human operators in multitasking, high-workload situations. Introducing automation can reduce task demands and improve human performance. If the automation is not perfectly reliable, operators have to strategically operate the system along with automated aids. As a part of an experiment investigating attention allocation and automation, 60 participants' subjective strategy reports were analyzed to understand how task properties (e.g., frequency, criticality) and automation reliability affect operators' strategy use. Given different levels of automation reliability, participants developed both automation-dependent and automation-independent strategies to accomplish the tasks. Some strategies (i.e., following a pattern) were significantly correlated with overall task performance. Top performers expressed preferences for different strategies compared to worst performers. This study provided insights into how operators naturally develop and use strategies when interacting with automation.