This paper examines the factors that determine the rate of errors called in major league baseball, and changes in the error rate over time. We find that (1) the rate of errors depends on the quality of play and characteristics of the field, but that (2) these do not fully explain variations in the error rate. With regard to point 1, we find that the error rate is higher when the quality of fielding is suspect, i.e., the performance of an expansion team in its first year, or the fielding done by replacement players during World War II, and lower when playing conditions are better, e.g. on artificial turf and during night games. With regard to point 2, we find evidence that official scorers are biased toward the home team, but that the bias declined when the stakes involved increased, with the advent of arbitration and free agency.