I realise I'm exposing my limited rugby education, but what is a SWAG algorithm?
You had to ask. It has taken me some time to get to the point where I sort of understand it enough to build it into the visual basic coding modules of the spreadsheet I use. Some of these I have just copied the relevant subroutines directly off a couple of geeky Excel VB forums without fully understanding how the actual boolean logic really works. Blind faith is useful sometimes. Here goes:
SWAG is an analytical technique that is endorsed by many of the worlds leading business schools, although it is not always acknowledged as such due to the degree of Intellectual Property that is associated with the first ever SWAG algorithm that belongs to Uppsala University Faculty of Business Management, and the fact that they all have a vested interest in promoting their own version of that technique/algorithm. Sometimes you may hear BOTE technique used in lieu of SWAG. The end result is often similar.
Essentially all the factors are fed into the analytical model, with various weighting factors applied, and then the SWAG algorithm uses a lot of mathematical probability theory considering both stochastic and deterministic approaches without ignoring the inherent contradictions in using a slightly mutually exclusive factorial methodology embraced in this dualistic approach. The algorithm also considers the impact of confounding variance and eliminates the possibility of any non-linear regression through a molecular consideration as opposed to a continuous discrete Newtonian spread.
The algorithm also runs the data through some fairly weird Mean Square Weighted Deviation (MSWD) modules before finalising the relative coefficient. The impact of random variables are considered as when the model is assembled, they are assigned a calculated potential distribution outcome factor derived from statistical inference and Sandyman's estimation theory using some really complicated vector spread distribution consideration that I really don't understand how it works. This occurs before any MSWD consideration. The final outcome is achieved through comparison of those coefficients driving which of the input data sets is the preferred solution.
I'm not sure that I have explained exactly how the SWAG algorithm works in layman's speak, but it is a fairly complicated concept. If the model is built properly with the right weighted input factors applied, while not always correct, over multiple scenarios, the SWAG algorithm will produce a statistically significant correct outcome.