-
-
- Registered: 2017-07-24
- Topics: 0
- Posts: 1
- Xerneis On 2018-02-04 20:43:02
-
You lost me there... That thing I've described is how it works right now. Which means your desc
ription of how damage is calculated (attacker's Atk - defender's Def) is not possible. On the other hand, I have no ob jections to the possibility that 30% decrease of Def/Res does actually mean 10% increase of damage or any other %, ba sed on level... I am aware that my testing wasn't nearly as thorough as it could have been. Which is exactly why there is this line at the end of the guide:
"If you, however, think I am wrong about something, please try and post a proof in your comment."
So, please, do just that. I acknowledged that you might be true, I've changed the respective part of the guide, to match the testing I've done. Now it's your turn to prove me wrong by digging deeper. ;)
Edit: Didn't realize I was on a separate account when I wrote this, this is the same as ShikiIga, sorry for any confusion.
Sorry if that was unclear, when I said "what you described" I meant your hypothetical "if you had higher defense than opponent's attack he would deal 0 (zero) damage", not the part where you described what actually happened. We are both agreed on what actually happens and that the formula is more complicated than attack-defense for the reasons you stated.
As far as evidence against the flat 10% boost, that's pretty easy to check. If the effect is anything more complicated than what you are proposing, anyone with different attack/*ion values would be unlikely to get the same value by chance. So I just checked my Suigetsu damage with and without tags.
Without tags, on the single strikes I got an average of 3.5k and with tags I got 4.2k. On the double strikes I got 2.0k without tags and 2.4k with tags. On the triple strikes I got 1.6k without and 1.8k with. The triple strikes could be consistent with a flat 10% boost but the single and double strikes look closer to a 20% difference.
A couple notes for future testing
- I wanted to use Suigetsu to be consistent with your tests, but he may not be the best choice due to the single, double, triple attack business. We basically have three subsets of data which means it'll take longer to get a statistically significant set, and it would be easy to miscount the number of strikes which would mess up our data.
-I'm less confident about the triple strike data since it often (always?) resulted in chases making it harder to catch the values. Do triple strikes resulting in chases needed to be tracked separately or is the damage the same?
-I noticed you gave a single value but I saw a lot of variation, did you give average values or did you only test each attack one?
This post was last edited by SeventyBilliont on 2018-02-05 05:55:08. -
Log in in order to Post. LOGIN | Register