One major problem with time series analysis and forecasting  is that the past is not necessarily a good predictor of the future… A more serious flaw is that sometimes things happen which have little or no CAUSE relationship but seem predictive. Despite this is it often possible to construct a convincing scenario for things such as skirt lengths or sunspots and the stock market. These are called spurious correlations.
Go to the spurious correlation site and produce one chart and short explanation of why (despite everything) one might convince another person it is true.
Feel free to have a little fun with this.
  • attachment