Selection and decision-making involve a series of complex cognitive processes, and time-frequency analysis of electroencephalogram (EEG) signals can help understand brain activity in different cognitive tasks. In this study, a decision-making cognitive task related to the rock-paper-scissors game was designed. By setting the selection probability of the computer (rock 66%, scissors 33%, paper 1%), the subjects are able to make safe or risky decisions. The complete and complex decision task was divided into three stages (planning stage, confirmation stage and feedback stage). During the entire experiment, the NeuroscanQuikCap EEG measurement system recorded data with 64 EEG signal channels in real time. By extracting the characteristics of the change of the average spectral power and the phase synchronization value of each EEG frequency band (δ, θ, α, and β), a comparative analysis of the EEG waves at different task stages was realized. The results show that in the planning stage, the subject’s decision-making activities that want to win or do not want to lose are related to the enhanced activity effects of the parietal lobe α wave and the forehead θ wave. The order in which the brain waves appear may indicate that the subjects will make different types of decision plans. In the confirmation stage, the α wave spectrum power suppression effect and the δ wave phase synchronization enhancement activity show that the subject will increase attention after confirming the choice. In the feedback stage, the θ and β wave power spectrum increase effect and the α wave power spectrum decrease effect are related to different feed back results. And the effects are most obvious in the feedback that the subject finds the game lost. This preliminary study shows that, combined with a complete cognitive task, time-frequency analysis of EEG signals may be an important tool for visualizing brain activity in response to different cognitive functions. It also contributes to the further development for objective cognitive assessment solutions.