Random Facts Thread: Magic

[CDF]

Looks like there are other values as well, such as max -20%/-10% (Boreal Coeurl, Nightmare Hippogryph) and -30%/-15% (Moonfang Warrior in Campaign) along with the observed -60%/-30% and -40%/-20%

It was also concluded back then that there are two windows of resistance after the triggering damage. Example given was for the situation where the max cut is -60% but not fully realized, with the first 2 seconds -40% (instead of the max -60%), next 2 seconds -10% (instead of the max -30%), then after 4 seconds no resistance.

[Argisto]

I did some quick testing a couple weeks ago on the Bounding Belinda UNM. I used D. Shantotto as a nuke wall partner since I found their AI reliably predictable by almost always opening with Blizzard V when starting this encounter. Each comparison between casts below was performed at the very beginning of new NM spawn to eliminate the possibility of incurring any unintended magic resistance.

As an Mlvl 25 RUN/SCH in some MAB gear I cast Blizzard II and let D. Shantotto cast Blizzard a while after to get a baseline.
RUN Blizzard II: 2237
Shantotto Blizzard V: 3151

Then I made a few casts within about a 1 second window before and after Shantotto's Blizzard V to make sure we had reached the damage cap to trigger the full amount of Accumulative Magic Resistance.

RUN Blizzard first cast: 2237, Shantotto Blizzard second cast: 1890
Shantotto Blizzard first cast: 3151, RUN Blizzard second cast: 1342

As can be seen, whoever landed second received a -40% damage cut despite there being a 914 point difference in damage between the two initial casts. This indicates that -40% seems to be the maximum accumulative magic resistance and the damage cap to reach max resistance is somewhere below 2237 for this particular NM.

Next, I went to confirm the duration of the second tier of resistance by casting after Shantotto at different intervals. I was just using a stopwatch and /echo timestamps to time this part over a few attempts at each interval so there may be some variance in actual timing due to server/input lag.

Shantotto first cast: 3151
RUN cast at 2 sec: 1789
RUN cast at 3 sec: 1789
RUN cast at 4 sec: 1789
RUN cast at 5 sec: 2237

Casts in the 2-4 second range only received a -20% penalty confirming the second tier is half the potency of the initial max resistance.

Lastly, I used a 3 Gelus rune Rayke on Bounding Belinda and cast within 1 second after Shantotto within the Rayke window.

Shantotto first cast: 3151
RUN second cast: 1342

Spoiler: show

As was found in the FFXI Math video, Rayke had absolutely no effect on the damage penalty incurred by Accumulative Magic Resistance. It really does seem like the testing referenced here didn't take into account/didn't know about the timing of the reduction of the damage penalty and mistakenly attributed it to Rayke instead.

[CDF]

Thanks for checking. I didn't even know Rayke was already addressed the video as I just skimmed the contents

[Byrthnoth]

Cardinal Chant enhancement works . . . how?

GEO with > 200 INT, 13 MDMG from gifts, 42 MAB from gifts, 20 MAB from job points, and 24 MAB from /RDM casting Stone on Darksday
Stone base damage is 160 + 13 MDMG = 173
MAB = 1.86

Stone from the West (MBB) - 321 damage = FLOOR(1.86*173)
Stone from the East (MAB) - 344 damage = FLOOR(1.99*173) - +13 MAB from cardinal chant on non-ra spells, as expected from SE's info post.
Stone from the West (MBB) with Geomancy Galero +3 - 321 damage = FLOOR(1.86*173) as expected
Stone from the East (MAB) with Geomancy Galero +3 - 370 damage = FLOOR(2.14*173) - +28 MAB is +13 MAB +15 MAB from Cardinal Chant +4, so I guess it isn't a straight +100%


With 220 INT:
Stonera from the West (MBB) - 954 damage / 1.86 = 513 base damage
Stonera from the East (MAB) - 1041 damage / 508 = 2.03 MAB = 17 MAB from Cardinal Chant

With 221 INT:
Stonera from the West (MBB) - 954 damage / 1.86 = 513 base damage
Stonera from the East (MAB) - 1128 damage / 513 = 2.20 MAB = 34 MAB from Cardinal Chant, so this one does double appropriately


I lazily tested MCrit (north) and did get it to proc for its amazing +10 MAB a few times, but the rate is genuinely very low even with Galero +3. I would believe the 11% from SE. Weirdly, I was about 50/50 getting +3 MAB (first tick of West) when trying to line stuff up to the north of me and I didn't think I was doing a terrible job.


I should confirm the potency of Cardinal Chant +4 for MBB, because that's clearly the most useful direction. Seems likely that it's by far the best GEO nuking hat if your enemy happens to be west and you're MBing, though.

[Byrthnoth]

I started testing again with fire because it was windsday and I didn't want to deal with day penalties and Stone. Fire has 188 base damage for me and I still have +86MAB base.

I extracted the X/Y position of monsters and my position relative to them (nuking with Geomancy Galero +3 on), then got these ranges:


A couple of conclusions:
1. MAB goes up by 3 or 4 per step and there are 9 discrete steps
2. a/h is monotonically related to the true metric they're using, but the fact that the ranges are nonlinear means I probably didn't guess the right one. Something more like log(a/b) would be more appropriate and symmetric, but the step size is still off and I doubt they used log. I've forgotten too much trig to guess what the right transform is here.

In this case, "a" is the distance the monster is east of me, "b" is the distance the monster is north of me (I threw away the 3 mcrits I got), and "h" is the length of the hypotenuse.

Edit: arc cosine is the function I was looking for. The easiest model is pi/16 slices of a circle with the slices aligned evenly around 0, pi/2, pi, 3pi/2. (so they fall in the middle of a slice).

[Byrthnoth]

For MBB, on Level 134 Apex Bats with +200 MAB, 250 MDMG, 358 INT, and 0 MBB:

Fira III
* Base: 4149
* South MB: 11823
* West MB without Galero: 15132 (+28 MBB, as expected from SE's graphic)
* West MB with Galero +3: 17496 (+48 MBB, so +20 from Galero +3)
* West MB with Collimated and Galero +3: 20334 (+72 MBB, so +48*1.5 as expected from previous testing)

Fire
* Base: 1266
* South MB: 3606
* West MB without Galero: 4398 (MBB+22, as expected from SE's graphic)
* West MB with Galero +3: 5046 (MBB+40, so +18 from Galero +3)
* West MB with Galero +3 and Collimated: 5769 (MBB+60, so +40*1.5 as expected)


So yeah, not quite doubling potency and it does affect -ras and normal nukes differently. My guess is that it's some hand-rolled special lookup table for Cardinal Chant +1/2/3/4.

[Byrthnoth]

I tested the extremes (that my mule can reach) and Cardinal Chat doesn't appear to have a cap. I reached +122 MBB with 40 MBB 1, 10 MBB II, and 72 Collimated Cardinal Chant MBB. I should be able to get higher with better gear (+6% MBB II?) and /BLM (+7% trait), but this is the limit of what's practical for me right now anyway. No functional cap.