Talk:"The Old Man of the Mountain"/@comment-50.71.60.33-20180213041407/@comment-34208015-20180226002854

FGOHikaru, With all the respect, tools such as calculators are really time saving if coded right. It would be silly to not use such tools if they exist, the only problem is that I don't have. I already tested KH a few on my support list and while he has some really good NP gain on his (unique) Art card but 60% with just crit bonus is highly exagerated, I'm serious. Sure he can't fulfill his jauge faster than most servants in game if you have favorable draw (I never said the contrary) but with 2/15 chance to draw either his Art or Quick card the odd are not in his favor. If crit luck isn't on your side, KH needs at least 2 turns to refill his NP after the turn he fired it, meaning a total of 3 turns between each NP. JTR needs 1 to 2 turns (depending on overkill and Quick bonus) to reload and fire.

I didn't checked Stheno yet but on Carmilla, I'm sorry being a little bit vulgar (and don't take personally) but what kind of bullshit are you trying to make me eat? 0 to 100% Carmilla AQQ with nothing but criton Q? Considering her NP gain rate (2%) her number of hits on her Quick (barly 2) and Art (just 1) cards with nor passive skills or active skills on her personal skill set that increase NP gain rate, I can't see how see can be better than a Jack or even KH. I'm not saying that Carmilla is bad at NP gain but once again you're highly overestimating the NP gain rate likely because you're didn't see some factors while testing them. I have here a video of Carmilla and while she have NP gain, I'm not impressed (only 130% NP gain rate Up buff on AQQ brave chain with overkill on QQ Extra and crit on the last Q, JTR can do better). Click here to see the video

You're counting too much your experiments. I'm not saying that experiments are bad but both theorical calculations and pratical experiments should confort themselves. Typically while you say that Carmilla can gain 20 to 40% NP per Quick hit (meaning 40 to 80% per card) with no crit and overkill it's not credible: because the result of the estimation is aberrant (80% to 160% per Quick card with crit) compared to what you should expect from the theorical values meaning that something is off either in theorical parameters (hidden stats maybe? wrong formula) or that you messed during the experiment. And in this I think you clearly messed up.