QUOTE(skillchip @ Feb 27 2017, 06:43)

I could try to make a less complicated version of my old optimizer spreadsheet?
having a way to calculate quality_mod which doesn't depend on PXP0 (even at the cost of losing a bit of precision) would be enough.
then, some proof check in my way to reverse-calculating the derived_base starting from a forged value.
if the forged_value re-calculated by derived_base coincides with the old one (which currently doesn't always do), then i could already have a public release to beta-test.
btw, in the sheet you sent me used roadshoe's shoes (since those were the datas i gave you), which are IW10. then you calculated quality_mod starting from item's level, rather than pxp0. is it intentional?