A teleoperation system consists of an operator interface (master) controlling a slave system in a remote virtual environment. When the slave’s environmental forces are fed back to the operator, this is called haptic force-feedback. An impedance-type master device ideally has a large range of stable achievable impedances, i.e. a small free space impedance, together with a large stable closed-loop impedance. In addition, the rendered force should be accurate: the resemblance between the rendered and desired force should be large. In haptic design, those properties are quantified as the Z-Width and the Transparency. However, the properties to achieve those results, often contradict. This thesis develops a method that relates component level choices the selected performance criteria. This allows direct design trade-offs, enabling the designer to optimize for desired haptic performance. It consists of three steps: first the Key Performance Indicators, KPI's, are determined: the Z-Width and the Transparency. Then the effect of individual physical properties on maximum stable rendering and the KPI's is evaluated. The final step is to determine how component level choices affect the physical parameters and thereby the KPI's, since the physical parameters are often interrelated in component level choices. The focus will be on the well-known Haptic Paddle Configuration, and tests will be performed on the Gemini haptic paddle device. This experimentally validated approach, in combination with the reduction in computational effort, enables designers of haptic devices to optimize for closed-loop haptic performance