outliers and discontinuities.
- When you generate plots, be they cycles or trends, of historic
information there is the 'overall pattern' from which generic rules are
often derived. But there are almost always outliers which fall well
outside the regular pattern of the bulk of the data and do not fit
whatever 'rule' is created to explain/predict the bulk of the
information (Yes, I know this is elementart data analysis - I'm getting
somewhere). There are also discontinuities in which some trend or
cycle is going along and then, for some either ascribed or unknown
reason, procedes with a new/different trend or cycle.
What outliers or discontinuities have YOU noticed which, in the long
run or with knowing hindsight, are interesting/relevant and, in the
spirit of this group, what if that happenstance DIDN'T occur?