The Predictive Model and the Goat Trail

I've been thinking that there must be some way to create a predictive model for what I do, but I haven't decided what the relevant set of indices or indicators should be, and whether they should be leading or trailing indicators, or a combination, or two separate sets. Obvious candidates are temperature, atmospheric pressure, probably humidity, at least one macroeconomic index (the Russell 2000 or the Tao?), at least one microeconomic index (outstanding loan balances?), diurnal medication patterns, diet, rapid eye movement (but that would require some expensive medical equipment, which would skew the outstanding loan balance), blood pressure. I'll require a sizable Excel spreadsheet to chart all of these, and I'll have to determine some common scale that will allow me to overlay the indices on a grid to look for coincidence, confluence, divergence. Having hedged the project with caveats, I still really want to be able to predict where I'm going.

Some people just beat a path, like the apartment dweller whose mountains of newspapers and magazines collapsed on him. The sociologist who studies this phenomenon called them goat trails--narrow paths for scrabbling from one room to another. I think of my recent dreams as something like goat trails. I can't see much to my right or left, and my footing is uncertain, but they always lead somewhere, even if it is only to the bathroom.