They didn't even tell us how exactly resistance or defense reduce ninjutsu and taijutsu damage! Don't expect too much out of it.

Lol there is zero transparency in this game friend.

We should keep supporting their good way of doing things topping up.

Ikr I've been waiting for an answer to that question since the end of last year. I actually think they have no idea how it works either but when u think about it how is that question supposed to be brought to their attention when the only thing that really concerns them is how to rip off people with overpriced *, so I think they cant even being to think about stuff like that because it's completely irrelevant to what their goal as a company is. In many games you get the exact formula of how each stats interacts with one another but here it's left half-assed and completely all over the place.

Wether they know or not doesn't really matter. They won't give us that information either way.

I would like to say that we can infer the probability distributions from players experience, just by plotting the number of times an upgrade/refinement has succeded at a certain value as a function of the total number of maximum chances before success.

However, i do not believe (from personal experience) that the probability distribution remains constant at all levels. For example, upgrading anything from lvl 1 or 2 to 2 or 3 usually requires less than 5 tries out of 20 or 30. In other words for the first 2 levels a 20% filled bar will usually result in a successful upgrade/refine. However once you reach lvl 4 or 5 you usually need 80% or more of the bar for a successful upgrade/refine.

Although this only takes into account my experience when lvling up charms/tactics/refinement/battle armor and is not a very big sample, this behavior uis surprisingly consistent. Thus i believe that the probability distribution for the low levels peaks earlier than the distribution for the latter ones. Also this is supported by the fact that i know no one who needed 18+ tries to upgrade from lvl 1 to 2 however i have seen players who needed 65+/70 tries to upgrade at lvl 4.

Also i have noticed that its rare for an upgade/refinment to succeed at the middle values, its usually successful whe the first or last quarter of the bar is full. So i do not believe that the probability can be linear.

Those are my observations, again my sample is only me and what i have seen on the forums from time to time. It only serves to give a tendency and not the full behavior.

Ah, but it could be a consistent formula, say something like this

rand(0, (M-W)^2) == 0

Or, in words.

take your max wish credit, subtract your current wish credit from it, square this difference. Roll a number between 0 and the squared number, if it is 0, it is an upgrade.

At low bar, even if you are empty, M-W is small, so you can upgrade fairly quickly. Whereas for high lv even at 80% the difference is pretty large so it's hard.

Power of 2 might be a bit much, thou I don't think it's linear. Maybe it's ^1.5 and then round to a integer, but you got the idea.

So like i guess we can count the number of players using this forums on 1... i mean 2 hands. Or maybe that number on the top left saying Total 29.

Good point. But it also illustrates the fact that the only thing we can do would be to map out the distributions for each level of refine/upgrading.

It's impossible to infer from that the actual formula since we have way too many free parameters and possibly virtually any formula.

Even then our sample with the current forums would be ridiculously small. So we can only specify trends using our observations, such as upgrading/refining does not follow a linear distribution and it gets less likely to succeed with a given percentage of the wish/charm/armor bar full as you lvl up.