Mathematical Problems

First of all, if you have more physical strength/speed why does your cumulative stat go down? Shouldn't it go up? (higher net strength = more powerful)



If A = 4 and B = 4, then Z = 1.00
If A = 8 and B = 4, then Z = 2.00 (A goes up, Z goes up)
If A = 2 and B = 4, then Z = 0.50 (A goes down, Z goes down)
If A = 4 and B = 8, then Z = 0.50 (B goes up, Z goes down)
If A = 4 and B = 2, then Z = 2.00 (B goes down, Z goes up)

(((A^2 / B^2)+3)^2 / 5

eh.

I really don't know what you're looking for... like what Z would do exactly.

So... What's the problem with A/B=Z? A and B will always be positive because all you will do is at to them from their starting point of zero or one. So, Z will allways be postive and nonzero. All that's left to do is add functions to define a scaling rate, which you can just play around with on a graphing calculator application on your computer.

What if b=0. MIND BLOWN.

Seriously though, not sure what Sin wants here...?

use a root or logarithmic funtion.

Lagger09 wrote:

use a root or logarithmic funtion.

Yeah, this would work. Every time you chop wood you increase X by whatever ammount and Y=X^1/2 or 1/3 2/3 etc, just however you want it to scale.

both a logarithmic and root funtion have a decreasing slope but still positive for all of infinity contiuing toward a positive X.

K(a)^1/G(b)=z
K&G being constants

Then please, with your infinite wisdom, tell us what you are looking for with this amazing code! I'm sure once it is done, it will be able to train dogs, solve fractal geometry, and finally provide the question to the answer: 42!

And don't assume that we're retarded. I may be an English major, but, damn it! I can understand mathematics*

Spoiler:

You're talking about diminishing returns. Your Z would basically represent the derivative of your actual function that you need. The problem is that it varies with two quantities so it would be graphically represented by a surface, not a line.

Lets let F(A, B)=final "position" of your "power" derived from the statistics, in whatever the action or passive may use this sense of "power" for.

Let F_A=the rate of change of F with respect to the A attribute, at any given point on the surface

Let F_B=the rate of change of F with respect to the B attribute at any given point on the surface

As I understand it, this is what you require:

Quote :

As B approaches infinity and A remains constant, F_B approaches zero but never reaches zero.
As A approaches infinity and B remains constant, F_A diverges (also approaches infinity)
F_B must contain an instance of the A variable, such that F cannot be the addition of two functions g(A)+h(B)

One such function that serves your purposes would be F(A, B)=½ln(B)A², which would make F_B=½A²/(B) and F_A=ln(B)A

Sadly, you cannot use natural logarithms, or any kind of logarithms for that matter with any reasonable speed in WC3.

An alternative function would be F(A, B)=½B^½A² so F_A=AB^½ and F_B=¼A²/B^½

Note: Raising a number to the power of 0.5 is the equivalent of taking the square root of that number. The constant coefficients can be tweaked as you see fit, so can the degree of the A variable, as long as it is greater than 1.

Basically when you want to compute how powerful something will be based on those stats, use
½B^½A²
and then multiply or add that to whatever you want to use it for.

I hope this post was enlightening, and hopefully it met all your specifications. Let me know if you're unsatisfied, we can get more specific and complex if you'd like.