I put one up, finally.
Will people understand it? Is it too cluttered? Should I stuff some pictures in there? I'm so familiar with this stuff I wouldn't be able to tell if this could teach anyone anything or not.
I figure there needs to be a page about it since there's a lot of rampant misinformation about how damage types are handled.
The edits look pretty good. I'm not sure where I would be putting the links?
The Great Guild Hall damage number gathering of February, 2013.
Charts from said gathering.
Damage versus Lost Souls at D24.
The chart from those experiments.
Lockdown defense thread.
Giant google document containing most of my research.
Skolver is a class armour. It has 'class' normal defence and 'base' piercing defence. Same goes for all weapon-oriented armour and specialist armour.
Jelly is a defensive armour set. It has 'base' normal defence and 'special' piercing defence. Same goes for all three defence sets.
Ironmight it a plate set. It has 'plate' normal defence and 'base' piercing defence. Same goes for Volcanic.
Ancient is an ancient set. It has 'heavy plate' normal defence.
Split-special armours, such as divine, has 'class' defence for both elemental and shadow.
Glass cannon armors, which only include Chaos, Black Kat and Mad Bomber, have 'base' defence for both normal and elemental/shadow.
Done. The text about Vanaduke's mask's defense is messed up, and I don't know how to fix it. Please advise or fix.
The last point on calculating damage: but...but...what about the shenanigans? Isn't the damage between 2x def and 1x def (or am I still misunderstanding all that)? And it's mixed damage which is not entirely explainable either.
You mean the normal+shadow example? Each half of the sword does 275, which is more than twice 75. So there are no shenanigans. I've edited the page, trying to explain this better, and tinkering with the numbers. Thanks for your feedback.
I see where I went wrong. I took 150 to be x. Thanks!
Actually, Kathrine-Dragon raises a good point. Which of these is the correct place for the threshold?
* damage = 2 * defense,
* damage = 2 * base-defense-for-that-level.
For example, for shadow against beasts at depth 28 is the threshold
* damage = 2 * 164,
* damage = 2 * 82?
Pretty sure it's the first one.
Edit: Oh, I see the problem.
Yeah, that shadow sword is going to need a base damage buff so it doesn't run into treshold shenanigans.
Alternatively, reduce base defence to 50?
It's fixed now. Kathrine-Dragon, sorry for dismissing your point before I fully understood it.
@Bopp
No hard feelings.
@Zeddy
See? Now we're doing it OOO style. Buffing things because they don't work otherwise.
I've been looking around for awhile now... and can't seem to find the exact point in which you are sourcing the "points" for knight defense.
I count only 21 black (tested) datapoints for the swords (Btw the merc FF entry looks like a typo). Not only that, but none of the DA datapoints (7) match the damage prediction equation (Yes, its less than double the armor - so they shouldnt be used. I'm just saying you're down one third of your "good" datapoints).
Is there other data I'm missing? I'm interested in finding out how you obtained the 125 numbers and so on.
Edit: Also. Would you prefer me making my own thread for my data or just add it in here?
With two defences accumulating and two damage bars working in some kind of mysterious way yet to be entirely figured out the logic off, an offset of 3-4 damage points is actually entirely withing the realm of acceptable error due to rounding errors and whatnot. You'll notice, crucially, that I haven't harvested the deathmark number for DA yet, and that would probably be why most of the numbers are off.
I've redone the extrapolated numbers in the lockdown sheets now. They were previous D25 effective + 70. But D25 effective was calculated as:
(4 * D24 effective + 1 * D28 effective) / 5
Now D25 raw is just (4 * D24 raw + 1 * D28 raw) / 5, which should be less layers of extrapolation. However, the D28 raw numbers are still extrapolated.
DA is still inaccurate, however. In fact, it's more inaccurate than before. I'll have to enter Lockdown and measure max damage manually. For now, I'll simply extrapolate the raw numbers from the defended numbers.
The actual defence numbers were mostly measured using Triglav/WRH's charge attack. It seems I didn't write the actual damage numbers down from those tests in the sheet. But they can be found here.
Parts of it could use fleshing out. I'll make some edits. Keep an eye on the page, in case I misinterpret something.
The article could benefit from some links to forum threads, Google spreadsheets, etc. where this research was carried out. Please add those.
Edit: Please give me examples of "base", "class", and "special" armors (or put them in yourself).