Reducing Magic Damage

[Genosync]

Eh, I went to edit my post deleted it and rewrote. To summarize what I had, basically Genosync you're somewhat incorrect. There are no tiers, only 'rate change' in your base damage where below equal INT to your target, it rises at a 1:1 ratio, then depending on the spell it rises at a 2:1 or 3:2 rate, then at another point rises at half the previous rate, then eventually caps. Adding to what aurik said, it's monotonic, continuous, and piece-wise linear, there are no tiers and it doesn't jump. The thing is, what's more, on anything practical you'll never hit the point where the rate lowers to half, let alone caps.

To say that it has any practical purpose on anything end-game related justified by what's written on the wiki is silly, since the wiki only concerns the damage of your spell figured after resists, not the chance of resisting itself. The argument people make for using INT instead of Elemental skill on things like Kirin is because they believe (rightly or wrongly) that it has an additional effect on resist rates. If you normally do say 700 damage on Kirin but you sacrifice 10% accuracy to do 800 unresisted damage, that's 10% of the time you're going to be doing 400 instead of 700 and it's up to you to determine what works out better of the course of the fight. It has nothing at all to do with people who say "I think target's INT affects resists" or "I think target's MND affects resists" which are silly discussions anyway because everyone has their stories about how they cast shock on stuff and eyeballed their resist rates or cast on DRK mobs as BRP is saying.

Each magic tier has an inherent mutliplier value M.
Each magic spell has a base value V.
Let dINT be (Caster's INT - Target's INT).

For dINT < 0: D = V + dINT (when dINT is a penalty, the tier mult. is always 1)
For dINT > 0, but less than some inflection point: D = V + (dINT * M)
For dINT > 0, but after some inflection point: D = V + (const + (dINT-const) * M / 2))
(above some critical value, adding INT becomes half as effective)
For dINT > 0, but after some cap: D = cap

M = 1.0 for all I/II single-target nukes
M = 1.0 for all -ga spells up to Blizzaga II
M = 1.5 for all III single-target nukes
M = 1.5 for all -ga spells Thundaga II and beyond
M = 2.0 for all IV and AM I/II spells.

that's what I meant by 'tier bonus', the multiplier bonus you get based on the tier of the nuke. why don't you read the article that I talked about a couple times in my post to understand the lingo.

And if you're too lazy to look it up yourself, http://wiki.ffxiclopedia.org/Calculating_Magic_Damage

[nivlakian]

I can assure you that Suiram is more than well-versed in the lingo on that page since he is one of the main editors/contributors to it.

[Suiram]

Yeah I think you just misunderstand what is written. Where "tier" is written it means for example Stone vs. Stone II vs. Stone III, etc. Basically when you have less INT than your base damage increases by 1 per INT, then it increases by more depending on the tier (again, by tier here we mean stone I, vs stone II etc.), then lowers again to one-half the previous rate, and at a later point caps. As I said in my previous post, you're never going to hit that point ever. Even with 100 more INT than your target you only see it in the lower end of the tier II (I hope by now you understand what we mean by "tier") spells. Here's a picture:

http://img360.imageshack.us/img360/9...sdamageam3.jpg

Now that we all can agree on the terminology you can see my point that it doesn't matter if you're getting 1 damage or 1.5 or 2 per INT, if you're sacrificing enough accuracy that your higher-damage unresisted nukes aren't going unresisted, you're not going to average more damage over the course of a fight unless you are getting additional accuracy from added INT.

[Terrance]

This is why BLMs and DRKs just so happened to resist spells more, BRP.

Dynamis BLM and DRK are very easy to land sleep on, RDM too. WHM on the other hand will resist occasionally. I don't nuke that often, but I don't notice BLM mobs resisting nukes any more than anything else.

Shell is not -magic, it does nothing to behemoth's meteor,

I strongly disagree, unless you can explain why unbuffed AV's Meteor easily does 1.7k-2k+ damage... And buffed it does 800-900 damage.

How are you turning 2k+ to 900 on NIN? Thats over 50%.

Anyway, I know I'm the only one standing after Chevalnik drops his death strike, and everyone that doesn't have -magic damage is taking the same number from the meteor regardless of if they died and didn't get reshelled or if they have shell 4/5 or if they're a blm and buffed themselves with shell 2. Merman items work, lamia items and shell don't.

[Valyana]

Anyway, I know I'm the only one standing after Chevalnik drops his death strike

Chlevnik's death move is an ability ("Chlevnik uses Meteor"), not a spell, and may work differently from the spell Meteor.

That said, Rukenshin's final Meteor screenshots show a lot of variation in damage:

http://rukenshin.livejournal.com/11067.html#cutid2
http://rukenshin.livejournal.com/13893.html#cutid1
http://rukenshin.livejournal.com/14661.html#cutid9

I don't know how final Meteor damage is calculated, but it doesn't seem that everyone is taking the same number.

[Sirlancelot]

So, would D.ring count as -10% magic dmg also? It just says -10% damage taken on it.
I've got Aegis, so I easily hit the -50% magic dmg cap, although I've never really tested to see if D.ring factors in to both -10% physical and -10% magical.
In other words, hit -50% magic damage with Aegis and capped Shellra V and get an additional -10% magic dmg for a total of -60%?
Or..would it still cap at -50%?

[Suiram]

Cap at 50%

[RKenshin]

How are you turning 2k+ to 900 on NIN? Thats over 50%.

I posted the buffs/gear I used, and the numbers I was hit for. I do have screens of taking the full damage but not sure where those are, although I'm pretty sure it goes without saying that the full damage meteors hit for about as hard as I said.

Anyway, I know I'm the only one standing after Chevalnik drops his death strike, and everyone that doesn't have -magic damage is taking the same number from the meteor regardless of if they died and didn't get reshelled or if they have shell 4/5 or if they're a blm and buffed themselves with shell 2. Merman items work, lamia items and shell don't.

The poster after you got me thinking, and I looked through all 8 or so LJ updates where I've posted about KS99 attempts... And through all those attempts and all the meteor damage screen shots (like 2-4 per entry), I've never seen the kind of consistancy that you claim to see. The damage variation is often rather big, and by big I mean like differences of up to 700 damage between players. I didn't see a single screen shot where even half the people hit were all within 100 damage.

I never realized Chev's final meteor is an ability (or something like that) though until just now, so yeah I'm not sure if that could have any kind of an affect.

[Kilik]

If dmg in ballista doesnt count, ignore this whole post.
http://www.fatesealers.com/images/GaloNukeS.PNG
total of 77+29 INT

galo thunder4 aegis only
706

Aegis with shell5
480x4(resist on earth day)
524x3(unresisted + with elemental seal)

Aegis with shell5 + merman ring (-2% more than earring and no change)
473(resist on earthday)
517(unresist)

Aegis with shell5 + merman ear
517x3(unresist)

Aegis with shell5 + merman ringx2 earx1 (-6% more than just ring, but no change comparing both resist numbers)
473(resist on earthday)

Aegis with shell5 + merman earring + lamia mantle+1 + lamian kaman
497x2(unresisted)

Aegis with shell5 + merman earring + 2Aqua Ring(10mnd)
517(unresisted)

Aegis with shell5 + merman earring + 2Aqua Ring(10mnd) and devotee mitts+1(6mnd,-3int)
521(unresisted)

Aegis with shell5 + merman earring + 2Snow Ring(10int)
504(unresisted)

Aegis with shell5 + merman earring + lamian kaman
513(unresisted)

adding 1MDB ended up being 4 less dmg
adding 1 INT ended up being 1.3 less dmg
adding MND did nothing

The difference between shell5 and no shell5 doesnt match up with the calculations posted before this though. However, everything else was consistent.

edit again: this is shell5 with only one merit in it btw.

[Mojo]

Hoping in here randomly-

I'm not 100% sure on this, but it seemed to me that that the damage you take from some 'magical weaponskills' or whatever you want to call them(Hurricane Wing, Great Sandstorm, etc.) could be reduced by stacking MND.

Also, from a few tests I was doing with a friend, we came up with the following for Magus' Roll.

NIN/WAR vs COR/WAR(No BLU in PT)

Base damage from a Ni spell was 85 damage. I was aiming for 100(I had 61+19 INT she had 58 INT) but it seems there is a magic damage penalty(15%?) in Brenner. Anyways, we got the following.

Roll  Damage

1     81
2     77
5     79
6     84
7     78
10    78
11    71

85*100/(100+MDB)=Damage

Thus for 2 and 11, you get the following.

85*100/(100+MDB)=77
85*100=77*(100+MDB)
8500 = 7700 + 77MDB
800 = 77MDB
10.4 = MDB

85*100/(100+MDB)=71
85*100=71*(100+MDB)
8500 = 7100 + 71MDB
1400 = 71MDB
19.7 = MDB

These were crappy tests due to low numbers, but it looks like Magus' Roll will give +10 MDB for a 2 and +20 MDB for an 11. I'll try it again later on RDM for more accurate data, and see how adding a BLU affects it.

[Suiram]

In response to the last two posts, if people take anything away from this thread, I hope it will be that the magic damage calculations for ballista/brenner are fundamentally different from the calculations between players and mobs. There are occasional instances where they coincide, for example the DoT damage and attribute penalty from elemental debuffs appear to behave the same in ballista as they normally do, but so many other things are changed--from resist penalties to the effects of damage mitigation spells and equipment--that any attempt to generalize from a PvP setting is unsound.

If you just think about it, it makes sense. They have routinely announced changes to PvP mechanics, from sleep resistances to multi-hit weaponskills, in an attempt to balance the environment without disrupting the overall balance of the game. While it might be interesting to know how these familiar effects function in PvP, it is still nonetheless a mistake to assume that this data would do any service to the knowledge of general, non-PvP game mechanics, except as a source of misunderstandings and falsehoods.

[Mojo]

While it is true that PvP damage calculations are different, I see no reason why some of them wouldn't be the same. I can say that Rampart functions the same in and out of PvP. Even if everything is changed, you can still take what is already known and use it as a source of comparison. I guess I'll have to try out nukes on someone with natural MDB in Brenner as well and see how it turns out.

[Suiram]

While it is true that PvP damage calculations are different, I see no reason why some of them wouldn't be the same. I can say that Rampart functions the same in and out of PvP. I guess I'll have to try out nukes on someone with natural MDB in Brenner as well and see how it turns out.

You miss my point. Sometimes they are the same, sometimes they are different, and there's no way of knowing which it will be. Since we know that they are so often different, you are better off assuming that they are, and test outside ballista if you want your tests to mean anything outside of ballista.

[Mojo]

I see your point entirely, I just think 'always testing out of PvP' is an unecessary precaution. If you take in to account the differences of PvP and compare what you test to what you already know, then you can still produce useful data.

For example, if Hyton: Ni was cast with 80 INT against someone with 58 INT and MDB I (+10 Magic Defense Bonus, RDM subjob) and the damage produced was the same as with a Magus' Roll of 2(77), then I see no reason why you couldn't conclude that Magus' Roll added 10 MDB.

[Suiram]

Even if you do it that way (which would prove difficult since you'd need to be able to achieve from trait and gear every conceivable value of MDB that your test may cover), you're still adding in an additional step of reasoning to extrapolate to general circumstances, which in itself still invites room for more error when you have to go from Magus Roll in Ballista to Known MDB in Ballista to Known MDB outside Ballista to get to Magus Roll outside Ballista.

Assuming the COR in your tests has WHM subjob available to cure the NIN, a NIN/WAR + COR/WHM could just as easily go to Aydeewa or somewhere else with higher-levelled worms and test with Stone IV/Stonega III/Quake, to get immediate, accurate results.

[Vail]

the numbers on magus roll look very pleasing since even at it's lowest it's comparable to barspell merits (but applies to all magic not single elements)

[nivlakian]

I see your point entirely, I just think 'always testing out of PvP' is an unecessary precaution. If you take in to account the differences of PvP and compare what you test to what you already know, then you can still produce useful data.

For example, if Hyton: Ni was cast with 80 INT against someone with 58 INT and MDB I (+10 Magic Defense Bonus, RDM subjob) and the damage produced was the same as with a Magus' Roll of 2(77), then I see no reason why you couldn't conclude that Magus' Roll added 10 MDB.

I believe the only thing you could conclude from this is that Magus' Roll added 10MDB in PvP. In the interest of balancing PvP, the designers can and have changed spells as they see fit. There is nothing to say that they may have left the MDB job traits alone, but decided that Magus Roll was too strong or too weak and thus adjusted it. It would be better to do a test like what Suiram suggested above in order to make sure it is accurate since I believe most people are more concerned with information that has been shown to apply to the normal game environment.

[Kylen]

According to its description, Magus Roll "enhances Magic Defense", so my guess is that it falls in the same category as Shell (and is subject to the 50% cap) instead of adding MDB.

And yeah, testing against some worms with higher-tier spells would probably be the best way to get some more accurate results.

[Valyana]

The most convenient worms I found in Aydeewa are at I-9 on map 3, on the way to Hydra.

Choco to the zone at Wajaom E-10 and follow the path inside until you see some worms on the right. Often two spawn close together, so BST sub is useful to keep one under control, but it shouldn't be hard to find one by itself if you wander up toward F-7.

[Elipse]

Sorry to raise an old thread.

Can someone point me in the direction of where the tests are on mnd and resists?


Some people use INT and the word "resist" in the same "test". INT lowers damage but doesn't seem to affect resist rates.

I want to define this because people seem confused by what "Resists" mean.

"Resisting magic damage causes the target to take 1/2, 1/4, 1/8, or 1/16 of normal damage" - Wiki


The only thing I can find here on mind is a ballista test with a guy putting on 10mnd and stating mnd doesn't affect resists. Has anyone done any serious testing on mnd in relation to producting resist rates?

At one point it was believed that barspells barely did anything. It was only when a group went all out and did +300 Fire resist did we see an "value" to it. So has anyone done any serious testing to see if MND is a variable either in increasing resist rates or possibly in affecting what "level" of resist you'll obtain if you do?

Anyhow I'd love to see where the testing is, if there is any.