Networks and local systems add delays to user actions in computer games, increasing the time between user input and rendering on the screen. Top-down studies using games have helped understand the impact of delays, but often do not generalize nor lend themselves to analytic modeling. Bottom-up studies isolating user input can better generalize and be used in models, but have yet to be applied to computer games. Our work builds a custom game for studying delay and the fundamental user input of selecting a moving target with a mouse. Analysis of data from a large user study shows target selection time is exponential with delay, and provides for an analytic model based on delay and the interaction between delay and target speed.