Just did a damage check on winmillion in tier 2 (landing the slash of the charge attack is incredibly difficult, from experience)
Anyways, tried following the standard on which damage numbers we put on the wiki, but it did more damage on the first depth of each strata than the second with some attacks, and less with others. Kinda weird. No damage bonuses on either of those floors both times. Recorded damage table follows (Forgive how terrible it looks)
S3 ____________ D9 ____ D10 ___ D11 ___ D12
Hit 1/2 _______ 48 ____ 46 ____ 50 ____ 46
Hit 3 _________ 62 ____ 59 ____ 65 ____ 61
Disk __________ 23 ____ 24 ____ 25 ____ 24
Charge ________ 92 ____ 90 ____ 100 ___ 95
Charge Disk ___ 41 ____ 42 ____ 45 ____ 42
D11 numbers off; sword had UV undead Low, D11 was graveyard, so disregard that.
S4 ____________ D14 ___ D15 ___ D16 ___ D17
Hit 1/2 _______ 47 ____ 47 ____ 48 ____ 49
Hit 3 _________ 61 ____ 60 ____ 61 ____ 62
Disk __________ 26 ____ 25 ____ 27 ____ 28
Charge ________ 99 ____ 100 ___ ? _____ ?
Charge Disk ___ 44 ____ 42 ____ 44 ____ 46
Lining up those numbers was hard too. Forum doesn't understand tabbed spaces. Or multiple spaces, which would've at least let me do it with whitespace instead of underlines. No D16/D17 charge attack data, mainly because I wasn't able to hit anything with the charge attack while I was testing on those depths (Forgot until last room on D16, D17 was RT)
All I can think of is that in both stratas, the first floor was an arena. Thoughts?
AFAIK, arenas do not make a difference during damage testing. (Danger rooms do.)
IMO, if we disregard the values in D11, the numbers doesn't really look that bad.
Have experienced some weird damage numbers during my testing too. See:
http://wiki.spiralknights.com/Talk:Voltech_Alchemer (D2 Versus Fiends & Slimes)
http://wiki.spiralknights.com/Talk:Shadowtech_Alchemer (D2 Versus Constructs & Beasts)
Believe that this could be due to an rounding off error when using weapons "below" their star rating.
Think the best thing to do is to leave a record of the raw values on winmillion's discussion page.
This way, other testers can go thru & double check the discrepancies when they have time.