SWTOR formula list

[Gorlough]

(01-27-2012 04:31 AM)Kor Wrote:  I'm working on a better way to measure both initial scaling rate (which is completely dependent on the divisor) and fall-off rate (which is dependent on both). Basically, the the smaller either the cap or divisor is, the "sharper" the curve (the more concave or "curvy" it appears to be). Unfortunately, it's not 1:1, I'm still working out the precise relation between the two (and how useful that actually is to us).

If I may assist:
It might be a good way to measure this by using a fixed (and taken out of 'real' stats) value to calculate the slope for.
As we have quite different curves for the formulae, I suggest we take 100 points for any given stat.

Why stat = 100?
Even the curves with the outmost exploding return per point invested, have their slope at 1 after 100 points invested. Additionally, the slope flattened out quite a bit already, giving more distinguishable values. Using a much lower value (i.e. 10) have the problem of having similar slopes for some stat curves.

Why not stat <= 100?
Well, using constants simplifies the calculation. And there's no reason to use a variable.

Ok, to get a practically useful result, I suggest to introduce this formula:

Code:

reward(stat) = delta(stat) / slope(100)

delta(stat) = StatRating(slope1) - 100

So what good is the value reward(stat)?
The lower the value for reward(stat) is, the more rewarding is it to put points into those stats up to the point where the slope hits 1.

Lets look at a simplified example (no real values):

Code:

reward(crit)  = delta(crit) / 2  = 500 / 2    = 250
reward(surge) = delta(surge) / 4 = 200 / 4    = 50
reward(aim)   = delta(aim) / 1.5 = 1500 / 1.5 = 1000

As you see, the numbers give you a good glimpse on how good a stat scales with points invested in regard of DR kicking in.