Well, I had some extra time to burn so I shoved this into a statistics program and did some quick analysis. Most of what I came up with we already have noticed by playing, but there are several places along the exp curve where the needed exp suddenly spikes. The points are at around level 60, 70, 80 and I guess it would be fairly safe to say that 90 would be another point. At these points, the exp curve seem to start using another equation for its determination all together. From 160 it's all approximately a simple quadratic equation, and from 60+ it's some power of 10.
Here is the output (which includes a prediction equation) in case anyone's interested:
Quote:
Originally Posted by Level 160
Polynomial Regression Analysis: EXP4NEXT versus LEVEL
The regression equation is
EXP4NEXT = 33646  8852 LEVEL + 630.2 LEVEL**2
S = 28413.8 RSq = 99.4% RSq(adj) = 99.4%

Note: Rsq is a measure of how well the prediction equation 'fits' the data given, the closer it is to 100%, the better, generally speaking, and with that in mind...
Quote:
Originally Posted by Level 6070
Polynomial Regression Analysis: C4 versus C3
The regression equation is
log10(C4) = 3.365 + 0.05807 C3  0.000061 C3**2
S = 0.000610440 RSq = 100.0% RSq(adj) = 100.0%

Quote:
Originally Posted by Level 7080
Polynomial Regression Analysis: C6 versus C5
The regression equation is
log10(C6) =  5.669 + 0.1873 C5  0.000037 C5**2
S = 0.000345030 RSq = 100.0% RSq(adj) = 100.0%

Quote:
Originally Posted by Level 8090
Polynomial Regression Analysis: C8 versus C7
The regression equation is
log10(C8) =  5.473 + 0.1827 C7  0.000010 C7**2
S = 0.000146737 RSq = 100.0% RSq(adj) = 100.0%

...Yea. But even though we have a nearperfect match here, I would hope that these are not the actual equations that allm uses...or someone has to fly over to South Korea and search their HQ for contraband liquor.
I may do more through analysis than just a simple line fit later on, but I'm too burned out from finals to do much of anything right now.
Edit: ...o. Looking at 7090 there doesn't seem to be much of a difference... so:
Quote:
Originally Posted by Level 7090
Polynomial Regression Analysis: C6 versus C5
The regression equation is
log10(C6) =  5.669 + 0.1874 C5  0.000038 C5**2
S = 0.000344360 RSq = 100.0% RSq(adj) = 100.0%
