Difference between revisions of "User talk:WinterRose/Racial Gains"
WinterRose (talk | contribs) m (moved Talk:Racial Gains to User talk:WinterRose/Racial Gains: Experiment failed, sigh.) |
|||
Line 17: | Line 17: | ||
-- [[User:WinterRose|WinterRose]] 12:10, 18 January 2012 (CST) | -- [[User:WinterRose|WinterRose]] 12:10, 18 January 2012 (CST) | ||
+ | |||
+ | |||
+ | --[[User:Mekrith|Mekrith]] 17:29, 19 January 2012 (EST) This information is not useless; but it clearly needs to be brought in line with things which are well known from lord 1 data. Some other things are clearly outliers (hlf, hel) based on our knowledge of how their lord 1 stats stand up. | ||
+ | |||
+ | I think some conclusions can be made with how con/int operate in respect to gains from the wide variation seen in giant and such. I'm not sure what those conclusions are, but I'll think about it longer and try to cook up a model which predicts that behaviour, if I'm super bored. |
Latest revision as of 17:29, 19 January 2012
The eternal HOGathon of January 2012 seemed like a nice time to do this project. I took the 20 creatables, plus what remorted heroes or lords I had that could be descended or rebirthed into mages, and ran them all to lvl 50. The results are in there, but keep in mind that except for half-orcs, elves, drow, and humans which were ran 2 times (due to suspicious results), all the others were ran only once.
Sample size of 1 is not enough for anything, but deviation can be observed: due/dwf hp are for example 516/499, and drw/elf mana is 398/408, so we can assume all these numbers are accurate to the average within 10 points, be it HP or Mana. Of course 10 points on a total of 200 or 300 gains is quite a lot, but the whole experiment is just a guideline, and never intended to represent some very accurate data.
If you have some further insight, please add it in, though I do not expect anyone has atheist 999 mag data...
Enjoy!
-- WinterRose 15:46, 17 January 2012 (CST)
I rerun the orc 2 more times so the total three runs gave: 556/343, 573/305 and 564/329.
It's interesting that the mana can vary so greatly, but anyway the mystery of half-org gains is solved - Hor probably get -20% same as Orc.
-- WinterRose 12:10, 18 January 2012 (CST)
--Mekrith 17:29, 19 January 2012 (EST) This information is not useless; but it clearly needs to be brought in line with things which are well known from lord 1 data. Some other things are clearly outliers (hlf, hel) based on our knowledge of how their lord 1 stats stand up.
I think some conclusions can be made with how con/int operate in respect to gains from the wide variation seen in giant and such. I'm not sure what those conclusions are, but I'll think about it longer and try to cook up a model which predicts that behaviour, if I'm super bored.