Many decisions made in real-world situations involve a form of intuitive pattern recognition. One way to investigate training principles for developing this type of decision making utilizes implicit learning in an immersive environment, where training stimuli are generated by a finite-state algorithm. In the current study, we investigated the effects of manipulating training-sequence length and algorithmic complexity in an immersive implicit-learning paradigm. Results: training-sequence length interacted with algorithmic complexity such that performance was best when training-sequence length was long and the algorithm was simple, and when training-sequence length was short and the algorithm was complex. When training intuitive decision making, training-sequence length should be matched to algorithmic complexity.