Magic Critical Hit Damage +XX%

[Asmoranomar]

I originally posted this in the ASA thread. At the time we (my BLM partner), wasn't sure what they did and we wanted to find out. Some people were skeptical, assuming it was +10% more crit damage, or 10 MAB to 11 MAB on crits (10% more MAB).

So for those that are wondering, we finally got off our butts and decided to crunch some numbers. The results of our findings show that "Magic Critical Hit Damage +XX%" gear adds an additional XX MAB during Crits. This does not include the +10 MAB you already get from a critical hit. If you have Magic Critical Hit Damage +10%, you will have 20 MAB (+10 for critical, +10 for gear) during crits.

We performed our tests on Wild Rabbits in Ronfaure, which have a base INT of 6. All tests were done on non-corresponding days/weather. No MB modifiers.

BLM was geared as follows:
HQ Staff
Wise Strap (Crit Rate+3%)
AMK Hat (INT+4 Crit Rate+10%)
Witch Sash (INT+5, Crit Rate+2%)
ASA Pants (INT+2, Crit Dmg +10%)

Phantom Tathlum (INT+2) was used as a control (pants).

Including his merits and gear, this gave him 92 INT and 42 MAB (LTNG) or 97 INT and 42 MAB (ICE).

Started out with LTNG II tests:
Non-Crit: 430
Crit w/o pants: 460
Crit w/ pants: 490

Burst II test:
Non-Crit: 1439
Crit w/o pants: 1541
Crit w/ pants: 1642

Freeze II test:
Non-Crit: 1455
Crit w/o pants: 1558
Crit w/ pants: 1660

During our testing, we would compare all our numbers against the known magic damage formulas posted on ffxi wiki. In all of our tests there was no discrepancies in the Non-Crit and Crit w/o pants.

After that, we just assumed an additional 10MAB on Crit w/ pants, and the numbers worked themselves out perfectly.

As for an explanation as to why "+10%". The only reason I can come up with, is that when MAB is applied to the damage formula, its done in XX/100. An additional 10 MAB would then be 10/100, or 10%.

[rog]

Doesn't change anything. Even with the pants (which you'd be an idiot to get crit+10 on), the only thing that's even slightly OK to use with mcrit, is witch sash, and that's only if you don't have sorc belt, since it's better.

Edit: actually wise will (situationally) beat bugard with the pants, but it still doesn't matter, because you lose far too much from your pants to make up for that extra ~1 damage average per nuke wise strap would give over bugard, not to mention potentially losing 1 macc (plus more from the pants).

[CDF]

I find it unnecessary to interpret posting of experimental results as an endorsement or recommendation of anything.

Thanks for checking.

[Chen]

I find it unnecessary to interpret posting of experimental results as an endorsement or recommendation of anything.

i agree, good testing. now if more of this stat pops up we can work out if it's any good easily

[Masamune]

You are on good way of testing (i used similar methodology for pDIF caps) but you need at least one another batch of numbers with same methodology but different INT/MaB combo(and same McritDMG+X% obviously).

Saying this because when testing pDIF caps, i used at first only one weapon rank (ie one INT/MaB combo here) to discover that wasnot saying me that another weapon would yield expected result matching with formula. So used 5 weapons of different ranks to be sure, but in your case i think only one another INT/MaB combo should be enough to prove or disprove your initial asumption (since you are lucky to not get any discrepancies like for pDIF tests).

EDIT: actually you need only another MaB test: your analogy with fSTR is a good idea i forgot it has a cap above which adding more STR does nothing, so you should get same numbers with 92 or 97 INT, but a different MaB value should be interesting to prove/disprove the +xx% "format" to Mcrit DMG.

[rog]

There's was no need to do any more than one spell actually. Magic damage doesn't vary at all, and there's a huge difference between 1 mab, and 10. There's no possible way the damage of them could possibly be the same.

It's not like we had no clue what it did, and needed to pinpoint the exact value. The only possibilities were 1 mab, and 10 mab, so two nukes (one crit, one not) is all that was needed.

[Byrthnoth]

Well, considering that you can get INT+2, MAB+4, MCrit Damage+10%, these pants would have made Magical Crit builds worthwhile and viable in the best DoT sets if they'd actually given 10% more damage during Critical Hits instead of +10MAB.

As it is, looking like an "eh."

Wise Strap would, on average, have to give a little more than .5 MAB on average to make it worthwhile for a well-geared BLM, which means it would have to give about 20 MAB during MCrit. Bingo.

Witch's Sash would have to give a little more than .5 MAB to make it worthwhile. Unfortunately, it'd have to give about 25-30 MAB during MCrits to be better than Sorc belt. Sadface.

KdE hat (INT+4, MAB+2 vs. INT+4, MCrit+10) would need 20 MAB during MCrits to be equivalent. Check!


So you could use MCrit pieces and sacrifice regularity for (maybe) a slightly increased DoT, or you could just stick to a normal build. Personally, I wouldn't touch MCrit because solo BLMs live and die by their damage regularity.

[Guppy]

When considering the amount of +M.Crit available, Augment gear is another source beyond the handful of pieces we have available, although it's expensive, time consuming, and all-around unreliable. I've heard of augments up to +5 M.Crit on Weskits, plus small amounts on other pieces (including many pieces far too expensive to spam for augments).

We also have "Increases magic critical hit damage" as a trait on the new Ebon/Ebur/Furia pieces (Coat, Cuffs, Sabots). While M.Crit is still a weak trait at present, the stacked increases could be substantial. At what point do we start considering dedicated M.Crit builds?

If that happens, getting a Witch Sash is going to be un-pleasant. Campaign's been going downhill despite the introduction of massively overpowered Campaign-only bonuses on gear. As a result, there's pretty much a fixed pool of them on my server, augmented only by the occasional server transfer or character hack/restore.

[Byrthnoth]

Mmm... Well, like critical hit rate, the percentage increase in your DoT from MCrit will actually decrease the more of it you have. So stacking MCrit isn't a good idea unless each piece individually trumps other options for that slot.

There have been +7 MCrit Weskits reported, but body is the only slot MCrit has appeared in to my knowledge. At current rates, a +7 MCrit Weskit would be worth less than +3 INT even with the pants.

[Masamune]

There's was no need to do any more than one spell actually. Magic damage doesn't vary at all, and there's a huge difference between 1 mab, and 10. There's no possible way the damage of them could possibly be the same.

It's not like we had no clue what it did, and needed to pinpoint the exact value. The only possibilities were 1 mab, and 10 mab, so two nukes (one crit, one not) is all that was needed.

It's not what i meant, you misunderstood i think.
If i understood correctly what OP wants to prove the following asumption:

"Mcrit dmg + xx%" translates into the magic damage equation as:
- +xx MaB (flat value)
or
- +xx% base Mdmg (+ have to prove before or after MaB factored in)

or like Byrthnoth understood also same as me:

...if they'd actually given 10% more damage during Critical Hits instead of +10MAB.

so still need another test with a different MaB value (same INT and same McritDMG+%) to compare to the numbers already posted, so can choose which asumption is the correct one(or another if both wrong?).

[Byrthnoth]

Well, nuking a 1 INT monster with those stats gives exactly 430 damage for a non-crit, 460 damage for MAB+10, and 490 damage for MAB+20 using Thunder II. Any Damage+10% would have given a much higher value. Because MAB is multiplicative, it doesn't matter what order the bonuses are applied in (except for flooring), so if there was a damage+10% in there we'd expect it to be about 10% larger than the normal magical crit.

For Burst II and Freeze II, something might have been different. It looks like he had 5 less INT here for some reason. However, 92 INT would have given his Freeze II numbers exactly with 42/52/62 MAB as they propose, and 87 INT would have given his Burst II numbers exactly with 42/52/62 MAB. It's possible he forgot to wear Witch's Sash for the AMII tests, or took it off to make his INT the same for Aquilo's and then forgot to put it back on for Burst II.

Either way, it's pretty convincingly MAB+20 during MCrit with the Damage+10% pants.

[Asmoranomar]

You are on good way of testing (i used similar methodology for pDIF caps) but you need at least one another batch of numbers with same methodology but different INT/MaB combo(and same McritDMG+X% obviously).

cINT = 128
tINT = 6 (Wild Rabbit)
M = 2
V = 710 (Freeze II)
StaffMod = 1.15
MAB = 56 (non) / 66 (crit) / 76 (crit+)

Tested and values were confirmed at 1711 (non), 1821 (crit), 1930 (crit+).

Well, nuking a 1 INT monster with those stats gives exactly...

We performed our tests on Wild Rabbits in Ronfaure, which have a base INT of 6.

For Burst II and Freeze II, something might have been different.

Including his merits and gear, this gave him 92 INT and 42 MAB (LTNG) or 97 INT and 42 MAB (ICE).

Some of us were wondering if it was possible to Partial Crit. We thought it would be interesting if MCritRate gear would indirectly reduce resists. I didn't recall seeing any testing done on this so we did some forbidden testing nuking ice elementals with ice in sky - complete with wearing the wrong staff and obi-day setup. Damage was very consistent (all nukes 1/16 resisted) and determined that it is possible to get a Partial Crit.

[Byrthnoth]

Ah, sorry. Got confused because my base damage table in my magic damage .xls is old and still had Thunder II's base damage as 173.

[Not Kuno]

Sorc belt is better then Witches now? I like my Witches sash and it's magic acc/mcrit even if it is one less INT.

[Byrthnoth]

Sorc belt is better then Witches now? I like my Witches sash and it's magic acc/mcrit even if it is one less INT.

Going from a pure DoT standpoint, unless you really need the MAcc from Witch's Sash, +1 INT is going to do better than it.

If you take:
1 INT ~=> .5 MAB for a well-geared BLM on T4s.
1 INT ~=< .5 MAB for a well-geared BLM on AMII.

Then you take a 2% chance at a +10 MAB Crit:
10*.02 = .2MAB
So that doesn't cut it for T4s or AMII. Double that using these pants (.4 MAB now) and it's closer, but Sorc belt will probably still win. Parsed, Witch's Sash would be worse than Sorc belt as far as your average nuke damage goes vs. pretty much everything short of Wyrms.

[Pendulum]

Wasn't sorcerer's belt always considered better than witch sash from the point of pure damage?

[Spekkio]

Wasn't sorcerer's belt always considered better than witch sash from the point of pure damage?

yes, given that we still wear an int+1 strap instead of a wise strap, assuming the target doesn't have a high resist rate.

[Aemora]

I've had Sorc.Belt for a while now... And even when I check the odd Blm here or there with a Witch's Sash, I feel a little gimp even though I know that the Belt does more Damage, the Sash just looks better.

[Riki]

Sorcerer's belt is not the definition of gimp.
Stacking Magic Crit. gear is.

[Virtuous Saint]

cINT = 128
tINT = 6 (Wild Rabbit)
M = 2
V = 710 (Freeze II)
StaffMod = 1.15
MAB = 56 (non) / 66 (crit) / 76 (crit+)

Tested and values were confirmed at 1711 (non), 1821 (crit), 1930 (crit+).




Some of us were wondering if it was possible to Partial Crit. We thought it would be interesting if MCritRate gear would indirectly reduce resists. I didn't recall seeing any testing done on this so we did some forbidden testing nuking ice elementals with ice in sky - complete with wearing the wrong staff and obi-day setup. Damage was very consistent (all nukes 1/16 resisted) and determined that it is possible to get a Partial Crit.

Pretty confident 1/16 is impossible, they likely take -MDT to ice, then you also are 100% resisted.