Gaze-based interaction methods offer intuitive selection processes without manual input, but this can lead to accidental activations, often referred to as the Midas touch problem. Implementing a confirmation mechanism mitigates this, yet it demands extra physical and conscious effort from the user. Brain-computer interfaces (BCIs), especially passive BCIs that utilize anticipatory signals like Stimulus-Preceding Negativity (SPN) - triggered by the user's expectation of an imminent stimulus - provide a seamless and implicit confirmation method for selections. In a Virtual Reality setting, our study distinctively illustrates that SPN can effectively interpret the user's intention towards an object they are looking at. We enhance the scientific comprehension of how this works by tackling a pivotal variable - our findings show that SPN activation is more related to the user’s intent to choose an object rather than the feedback from the object itself. Additionally, we explore how the placement of familiar targets influences SPN, discovering that SPN might be elicited more rapidly as users become familiar with the locations of these targets, an important consideration for practical BCI applications.