Precision here implies repetition of measurement to minimize deviation of the average result of the measurements from the expected value of physics quantity. Such deviation is normally caused by random (stochastic) errors. Nonetheless, this average value may still appreciably differ from the true one because of so called systematic errors. Such errors appear, for example, if you measure with a cheap/broken instrument, like a curved ruler. A reason for the systematic error could be failed calibration of an instrument as well as oblique angle of view when reading it. Apparently, there are two ways to discover the presence of systematic errors: either to measure a well-known quantity using the same instrument, or compare the readings with those acquired with a more accurate instrument. Such errors are usually either constant or proportional to the value of measured quantity.