Historical data play a crucial role in fields such as astronomy, providing accounts of past celestial phenomena that may not be predictable such as the appearance of novae, meteors and comets. Astronomers have made use of such information in their efforts to reconstruct past events; however, the reliability of these data and the willingness of astronomers to accept them have created difficulties. The expected comet of 1848 provides an example of the tendency of astronomers to utilize only the medieval data which fit their preexisting theories while discarding observations that challenged their assumptions. The resulting failure of the predicted comet to appear demonstrates the limitation of this selective use of historical data.