A computer system detects occurrence of an action exchange in which one of a player character and an opponent character performs a first action and another one of the player character and the opponent character performs a second action to mutually exchange the actions. Based on a first stiff time related to the first action, a second stiff time related to the second action, and a time difference between the first stiff time and the second stiff time, when a stiff time (hereinafter referred to as an “own stiff time”) related to an action performed by the player character is shorter than a stiff time related to an action performed by the opponent character, the computer system presents a player controlling the player character with a recommended candidate action recommended to be performed after an end of the own stiff time.