Changes in stimuli proximity have been shown to affect the performance of brain computer interfaces (BCI) based on steady-state visual evoked potentials (SSVEP). Specifically, closely placed visual stimuli compete for neural representations, which is called the effect of competing stimuli. Recently, steady-state motion visual evoked potential (SSMVEP) has been proposed to alleviate some of the inherent limitations of SSVEP. In this study, the SSVEP and SSMVEP paradigms were systematically compared under three different inter-stimulus distances to modulate the effect of competing stimuli. Offline analysis was performed to study the steady-state response characteristics, strength of the responses and overall BCI decoding performance. Thirteen healthy subjects participated in the experiment; two types of visual stimulus with seven classes were presented: a flickering stimulus for SSVEP and a radial contraction-expansion checkerboard for the SSMVEP. The canonical correlation analysis (CCA) was used to study the signal characteristics and the offline decoding performance using only three EEG channels, O1, O2 and Oz. The results demonstrated that SSMVEP was not only less sensitive to competing stimuli, but also consistently outperformed SSVEP in their presence. Further, the SSMVEP response reached its steady-state faster than SSVEP. The signal characteristics analysis revealed that the SSVEP performances were better than SSMVEP in the lowest frequency tested (<9 Hz), and SSMVEP performance was significantly better in the highest frequencies investigated (>13 Hz). The findings in this study indicate that SSMVEP is likely a more practical BCI paradigm than the classic SSVEP for many real-world applications.