20/03/2014 - we were seeding in our dreams - only way of taming this colossal force - though our differences were usually met with a human sized hole

 Brave socio-economists can, when called for, raise or lower the value of one constant number so that they can prove that 3 > 4. You get results such as "the smarter he gets, the dumber he becomes" :

supposing a=b,
z refers to a potential relationship between a and b, and here z is interpreted as the sum of a and b when a and b are alternatively polarised magnitudes of the (same angle and origin (the implied relationship between a and b, smarter and dumber)). z = a-b = 0.

So the trouble with z is that it doesn't communicate much about its constituents. Also, it's not a very functional relationship. Since z introduced the idea, what if z represented the terms of the relationship (i.e. they exist in only in z), as well as value? if we treat z as encompassing the the terms of the polarity shift,
, -+z = |a|*|b|, it instead provides a sign independent magnitude.

Let's say we can't determine for sure the 'direction' or angle in relation to the unit sphere in which (-)'smarter' is pointing. Instead, we map all directions around an origin on the complex / higher dim plane. (If we had more information about surrounding systems, their interaction would be accounted for).
w will refer to a vector on the complex plane which will

//Possibility is the goal.

...perhaps even z = (z+iw)(z-iw). At z = 0, z = i cos(w) and z = -i sin (w).

So the trouble with z is that it doesn't communicate much about its constituents. Also, it's not a very functional relationship. z = z0 = (z+iw)(-z+iw) seems more fitting, since z refers to the implicated relationship between sign and attribute, and w holds the current magnitude of that relationship (?) of that relationship. (Programmers solution : set first iteration of z to 1, that way progression isn't biased in the pos/neg direction, unit vector of subsystem(?)). Note that

The suggestion of the compliment ( ;) is that z = a, z = -b, where z is a magnitude which is sign agnostic. (That is after all what you're doing in both cases: imagining some relationship between some (potentially ill-defined) attributes).

...This definition is actively limiting the set. You get 'z' or '-z' defined as the upper limit of the set, so that no number can be larger than 'z' or '-z' in the context of the implied logical boundaries of the system (the relationship between z and w) . Because you can only define one upper limit on the set, or determine that z=w, whichever limit chosen as the active limit will always be equal to or greater than the unchosen limit. In the situation described, z >= w or w >= z. There are a few ways to describe this phenomenon, though I shall probably stick to 'phase modulated waveform' to account for probability in time (?)
This means that the difference between the two is of significance from a probabilistic standpoint.