WhatWouldHulkDo
Level 8 Valued Member
Physical attributes, movement patterns, skills - can you expand this thought a little more?
I wasn't making any attempt at defining which qualities to optimize, but I think a lot of other people here have listed out good ones. I'm more getting at the idea of balancing them against each other.
Forgive me, I'm going to go math nerd for a moment (I'm an engineer by trade):
Let's suppose you have two physical quantities A and B, normalized to have equivalent value (i.e. A=1 is just as "valuable" as B=1). You might think of a training program trying to optimize the total value: A+B. Problem with that is, in that case ending up with A=2,B=0 has the same "value" as A=1,B=1. To put it more concretely, that program would say that having an awesome deadlift and a piss-poor press is just as good as having a pretty good deadlift and press.
So, you need another term in what you're trying to optimize, to penalize the program if it lets your parameters diverge: A+B - |A-B|. Now, A=1,B=1 gets a total score of 2, but A=2,B=0 gets a score of 0. And, of course, A=2,B=2 is even better. I think that's what a GPP should do - look to optimize all parameters without letting them diverge from each other. Then an SPP would assign additional weight to certain parameters.
Question remains, how do you decide what A and B really are (and C and D, etc.). Pavel picked a set of parameters he liked for S&S, but that certainly doesn't mean it's the right set of parameters for everybody.
<end nerd mode>