the att & int stats for the defending hero in AeroCalc are completely dependent on other factors like how the war horn is used or if you give defending units that 40% buf or not etc etc.
the way I see it is that the defending hero will have the same stats for every round. That is the basis on which I work. This means that we must get sensible results for every wave with the same D.hero stats. This is what I call the linearity of the model and it's tuned by multiplications in our formulas. I think I managed to get it very close if not 100%. My reasoning is that I use war horn and corselet the exact way as Evony describes them and it establishes that linear result.
BUT... my patched formulas might give a kill rate that is twice too high. This is compensated by increasing D.hero INT. The thing is, that these changes do NOT affect the linearity of the model... they are merely like a volume control on a radio.
So, in the end, it is the ratio between D.hero ATT & INT that matters. The absolute value of each (as long as their ratio is maintained) is merely that volume control.But yes, I tried 5 or 6 different captures of hc14s and ATT=110 with INT=790 is pretty accurate. I will probably make INT=800 when I would give it to our members to make sure they send enough.
BTW, I am really putting some work into the formulas right now lol. By the end of today I will probably have rewritten 80%. An important change is that I take the leadership max-stack-size calculation to a single cell so that it can easily be changed. It also improves performance again.
Q: can you open the sheets I sent to you in Excel? Or do you have OpenOffice? I may be able to save in Excel format too, I never tried that yet...
cheers,
DeVerm.



Bookmarks