This series of lectures describes the work that lies between the raw data taken by the detector elements and the physics variables used to study particular reactions. We start by defining some simple physics variables of interest, then describe the fitting process used to extract values from the observed patterns in typical detectors. This is followed by a discussion of the various problems of pattern recognition in tracking, calorimetry and particle identification detectors. The process of calibration and alignment is surveyed, with emphasis on getting "reasonable" results in the absence of formally complete information. Finally, the role of Monte Carlo simulation in understanding the quality of the obtained information is examined. Throughout, we emphasize how the use of "composite" observables is required due to what our instrumentation and reconstruction can achieve