An expanded case against synthetic-type control charts

During the last two decades, many new methods have appeared in statistical process monitoring with synthetic-type control charts being a prominent constituent. These charts became popular due to their simplicity and proclaimed excellent change point detection performance. Synthetic charts are nothing more than the application of well-known and long established runs rules. We show that better performance can be obtained by using exponentially weighted moving average (EWMA) charts. Expanding on some previous questioning articles, we critically reflect upon recently developed variants of synthetic-type charts in order to emphasize that there is no reason to apply this special class of control charts. This paper renews and extends criticism in order to respond to newly developed synthetic charts, called "revised" and "modified" by incorporation of further restrictions on the observations leading to an out-of-control signal. Furthermore, we demonstrate that "improved synthetic charts" (synthetic charts augmented with an outer Shewhart limit) perform weaker than EWMA-Shewhart chart combinations.