I thought that it might be of interest to fellow CV enthusiasts if from time to time I might cite here some of the odd behavior that I've noted in some of the better known CVs in the decades that I have been associated with AAVSO. My objective here will be only to call attention to the behavior itself, not offer any explanation of why it has occurred.
The UGSS variable 2325+43 DX And has been a part of my regular observing program off and on since the late 1960's. Like so many variable stars it shows evidence of long-term changes in its lightcurve that can only be classed as rather odd.
During the 1960's and much of the 1970's when it was in its quiescent state DX And would range widely between magnitude 14.0 and 16.0, with most of the estimates close to the midpoint of the range. However, toward the end of the 70's the magnitude of DX And when in quiescence began to progressively and distinctly brighten.
Throughout the 1980's and 90's the star was far more often easily visible to modest aperture telescopes. Its average minimum brightness during this interval being widely reported to be in the mid 14's. At the same time there were virtually no reported estimates below 15.0 magnitude.
Then, as the end of the century drew near, the mean magnitude of DX And in quiescence began once more to decline. To my mind the outcome of this alteration was even more distinct and dramatic than the brightening phase that the star had experienced two decades earlier. In it new minimum state DX And had a mean magnitude typically in the early 15's, while its magnitude range at minimum had contracted noticeably as well. The observational scatter now never brought it above the late 14's, nor below magnitude 15.5 .
Try plotting a 20,000 day lightcurve for DX And using the LCG yourself and you'll easily see the odd changes this star exhibited in its minimum brightness over the past half century.
I would add that few individuals have observing careers long enough to actually detect such long-term oddities among variables they follow indepenently. Thus, the combined AAVSO lightcurve is a fine tribute to what the efforts of many contributors can reveal over long spans of time. It also points to why many of us find these stars fascinating to follow year in and year out.