i'VE BEEN THINKING AND THINKING ABOUT THIS
(bugger it. wrote that several times an pressed capslock by accident every time
)
Suppose Dessy's formula is accurate (very convincing so far)
And suppose that 1 enhancer per 1/10th level is also accurate
There are still a few posible explanations for the disparities
The first is purely mathematical:
I've just discovered a new texture called Exasperite Texture, which starts its SIB period at lev 26.55 (or maybe just a teeny bit below)
MA of course rounds that down to recommended level: 26.5 (yes , they really do. We found that much out)
Now, applying Dessy's formula to my discovery, I find it maxes at lev 31.705.
You, however, only know what's written on the texture , so you apply Dessy's formula, find it maxes at level 31.65, which you round down to 31.6.
Our 0.05 difference has turned into a 0.105 difference. So you now need two more enhancers than you think
Yep, even that extra little 0.005 disparity can be fatal, since the system can be very unforgiving of tiny shortfalls, eg I got about 95% once, just through applying a stack that was 1 texture short. ie 322 , not 323. Yes, it was a texture that I'd maxed)
There are also those rare, totally random fails , of course (Just because we
all compromise on the cast-iron 904 paints/textures
)
And there's human error.: miscalculation; putting the wrong stack in, etc.
Do you think that these factors, all taken together, could explain it?
And what can we do? We can always round up our own figures (apart from our level. Round that down instead) but we can't make MA do the same. I think that we'll just have to assume that MA rounded down, unless we happen to know otherise, and apply the formula to our adjusted "recommended level". Then round up the result. The resulting "maxed" figure would then always err on the "safe side" if it erred at all.
jay