There are coupious examples of spurious correlations possible. Check this website:
Are you talking about the old classic example: woman’s hemlines, and the stock market? It takes a little more than a plot to establish a hypothesis. There has to be a high correlation coefficient, and an established physical relationship. Flies and rotting meat have a fairly good correlation, but there is no physical relationship to indicate that flies cause meat to rot. That was an old wives tale that persisted for centuries. As my old statistic professor used to say when a high correlation shows up with no physical relationship apparent, start looking for a third, as yet, undefined variable. Many significant breakthroughs have arrived as a result of that approach. For the rotting meat that approach resulted in the discovery of bacteria. Many investigators have also spent years chasing phantoms with no results.
To be valid the hypothesis established from the correlation must be testable. The sample size must also be large enough to be statically significant for the population distribution being studied. To test the hypothesis embed in graph 222a above simply shut off the world's oil supply, and see what the GDP would be? That result can be concluded "a priori" without starving 7.4 billion people to death. The probability of a false positive is so close to zero that only an obsessive compulsive idiot would deny it.
There are several of those obsessive compulsive idiots who frequent this site. They offer an opportunity to study really disturbed minds. There are also several hypothesis that state when 5%, or more of a society becomes mentally deranged the society collapses. We are likely to be testing that hypotheis in the very near future.
