Enmity testing - Still Going!

[Spider-Dan]

Can someone please explain how this picture is possible? Sequence started with TA+WS on NPC. (please forgive sh*tty mspaint compression)

http://venomweb.150m.com/ffxi/npc-accomplice.jpg

[Kaeko]

Accomplice works on NPCs?

I never really tested anything with Accomplice other than the range of effect.

[Max™]

Yeah, I do a lot of fighting against Elementals, and Jelly mobs, so I'm regularly thinking... "man, if I could only do a little bit more damage to these things for like 30 seconds, this would all go much better."

Rocs, Corses, Doomed, Elementals, Flans, Ghosts, Hecteyes, Leeches, Mimics, Skeletons, Slimes. Take your pick, all have at least one type of damage resistance. Also like Spekkio said, it's great for Jailer of Temperance, especially when he decides to stay stuck on whatever damage type you're lacking most at the moment.

Well, I was thinking "Limbus, and Nyzul, Limbus I'm utterly finished with unless my shell needs more bodies to throw at U/O because of lower than preferred attendance, and my luck at getting any sort of Assault static suuucks (0/Depressing amount) so yay for something I'd probably forget was there in between chances to use it."

Temperance is a good point, and I'll concede that it would have been nice to have the last time I did Temp, back when my schedule permitted me to make Sea runs.

Doomed Im pretty sure are weak to Slashing, or have a bunch of +Dmg%, Rocs don't resist enough for it to even occur to me, and I'd rather not melee Slimes, Flans, Mimics, Corses, or most of the higher level Ghosts I've bumped into. (yay, perdition'D)

So, yeah, compared to how awesome Angon/Feint are when they're up, and the amount of time I'd actually think "oh man, Tomahawk rocks!", I'd rather use the 8 merits on 30 HP.
/derail off.

[Zilong]

So the ares cuirass enmity was just a false alarm on 3% HP boost?

[Kaeko]

So the ares cuirass enmity was just a false alarm on 3% HP boost?

There is none (100% sure), or Ashi would have probably changed the title of the thread again =P

It does seem like Marduk Legs is the only one we've found with a hidden enmity stat. If anyone wants to test their Aegis, Assault Breastplate, Excal, or other funky item they are praying to have a hidden enmity stat, they are welcome, but I personally don't really think this is anything beyond 1 item screw up.

[Spider-Dan]

Accomplice works on NPCs?

I never really tested anything with Accomplice other than the range of effect.

Well, you can use it on NPCs. But apparently, it doesn't actually work on them...?

[Kirschy]

So the ares cuirass enmity was just a false alarm on 3% HP boost?

Yea I was dumb and wasn't paying attention. I almost made the same mistake when I used Diabolic Eye as after the JA wears off my HP increases. Nice thing about making that mistake is you know to watch out for it again =/.

[Ashira]

Monomi: Ichi is 80 CE. (Casting Cure 80 times sucks!). I didn't get a chance to test VE, last-minute late night salvage~

You need some access to Dispel and such, Kirsch, to make it a bunch easier on you.


Also, yeah, as to the curing mistake.. believe me, Kae and I made that one a few times. ;D

[Lordwafik]

Rocs don't resist enough for it to even occur to me

Rocs don't resist enough for it? -25% isn't enough of a slashing penalty? That's huge, that means even your GA you love so much takes a big damage reduction.

and I'd rather not melee Slimes, Flans, Mimics, Corses, or most of the higher level Ghosts I've bumped into. (yay, perdition'D)

Sometimes you don't have a choice, limbus/nyzul/salvage etc, unless all you do is sky and campaign, different story.

[Max™]

Nyzul got a little more interesting just recently, still doubt I'll have any luck getting a steady group for it, cursed with an Anti-Assault line in my code. Do Rocs take off 25%? Never really noticed it enough to care, did Simmy for LS bud little while ago, even with Colla spam, and TA+SATA WSing the Nin when I turned it, Simmy was facing me around half the fight, without voking... imagine if I'd had that extra 25% Damage for maybe half the time!

Limbus is dead to me, said that.

Enough silly derail, you don't like me, I don't care, we get it.

[Melari]

I'm sure I messed up somewhere along the way, since this seemed far too easy. So I'm posting my data here for others to pick apart. It's a little convoluted, and for that I apologize. My testing relates to CE and VE values for Cures that recover HP. The testing below is limited, and only covers 1) curing yourself, not others; 2) Cure 1, 2, and 3 -- Cure 4 is not tested below due to lack of resources. Not to say that either of those would vary results, just that they are not specifically tested.

Summary: 1) CE for HP cured to self is equal to 13/18 units (i.e. CE for X HP cured = floor(13X/18))
2) VE for HP cured to self is equal to 42/9 units (i.e. VE for X HP cured = floor(14X/3))

Player 1 and 2 were the both level 75 characters, with no positive or negative enmity merits, and wearing no gear with enmity bonuses of any type. I did not have a 3rd person available to act as puller, so the first part of every test involved me balancing the player's enmity to account for this fact. This was done as follows:

Player #1: Pull with Silence Total CE = 200 + 1 = 201
Player #2: Flash + Cure1 for 0 HP on self x 21 Total CE = 180 + 21 = 201 (HATE)

At this point both players have an equal CE level of 201, and it is from this base that all my testing began.

Test 1: Determine CE for a casting of Cure 2 for 90 HP healed on Self
Player #1: {Equalized Enmity as Above} Total CE = 201
Player #2: {Equalized Enmity as Above} Total CE = 201 (HATE)

Player #1: Casts Stun Total CE = 381 (HATE)
Player #2: Casts Cure 2 for 90 HP healed on Self Total CE = 201 + X

X being the CE value for 90HP cured. After these actions, player 1 has hate; therefor X is less than 180 CE. Player #2 then cast cure1 for 0 hp on self repeatedly to determine the difference in CE values between the two players. Hate shifted to Player #2 after 115 cures.

Player #2 Casts Cure1 for 0 HP healed on Self x 115 Total CE = 316 + X (HATE)

So X = 381 - 316 = 65 CE.

Test 2: Determine CE for a casting of Cure 3 for 180 HP healed on Self
Player #1: {Equalized Enmity as Above} Total CE = 201
Player #2: {Equalized Enmity as Above} Total CE = 201 (HATE)

Player #1: Casts Stun Total CE = 381 (HATE)
Player #2: Casts Cure 3 for 180 HP healed on Self Total CE = 201 + X

X being the CE value for 180HP cured. After these actions, player 1 has hate; therefor X is less than 180 CE. Player #2 then cast cure1 for 0 hp on self repeatedly to determine the difference in CE values between the two players. Hate shifted to Player #2 after 50 cures.

Player #2 Casts Cure1 for 0 HP healed on Self x 50 Total CE = 251 + X (HATE)

So X = 381 - 251 = 130 CE.

Examining both of these CE values, we see that 65CE/90HP = 130CE/180HP. So I hypothesize that the CE value for HP Curing is determined as 13CE/18HP.

Test 3a: Testing the above hypothesis
Player #1: {Equalized Enmity as Above} Total CE = 201
Player #2: {Equalized Enmity as Above} Total CE = 201 (HATE)

Player #1: Casts Stun Total CE = 381 (HATE)
Player #2: Casts Cure 3 for 180HP healed on Self Total CE = 331
Player #2: Casts Cure 1 for 30HP healed on Self x2 Total CE = 373 (331 + 21 + 21)

Player 1 still has hate at this point, and assuming my calculated CE value is correct, Player 2 should be able to establish hate by casting cure1 for 0 hp on self 8 times (381-373). And this is precisely what happened. Hypothesis confirmed.

Test 3b: Testing the above hypothesis
Player #1: {Equalized Enmity as Above} Total CE = 201
Player #2: {Equalized Enmity as Above} Total CE = 201 (HATE)

Player #1: Casts Sleep Total CE = 521 (HATE)
Player #1: Casts Sleep2 Total CE = 1001 (HATE)
Player #2: Casts Cure 3 for 180HP healed on Self x6 Total CE = 981

Player 1 still has hate at this point, and assuming my calculated CE value is correct, Player 2 should be able to establish hate by casting cure1 for 0 hp on self 20 times (1001-981). And this is precisely what happened. Hypothesis confirmed.


VE is of course a little harder to test accurately due to limited measuring devices (a stopwatch in my case). But I performed the following tests to determine an approximate value. The testing premise was to first match the players hate, then have player 1 establish a hate level that player 2 could easily match using the above CE values for HP curing.

Test 4a: VE of Cure 3 for 180 HP healed on Self
Player #1: {Equalized Enmity as Above} Total CE = 201
Player #2: {Equalized Enmity as Above} Total CE = 201 (HATE)

Player #1: Casts Sleep Total CE = 521 (HATE)
Player #1: Casts Cure1 for 0 HP on self x 6 Total CE = 527 (HATE)
Player #2: Casts Cure 3 for 180HP on Self Total CE = 331
Player #2: Casts Cure 2 for 90HP on Self Total CE = 396

At this point, the player's enmity levels is such that if player #2 now casts Cure3 for 180hp on self (130 CE), the total CE difference between the players will be 1 unit. Player 2 will have hate due to the unknown VE, but player 1 will soon have hate again after this VE decays. By timing how long this takes to occur and using the known VE decay formula, I can calculate the approximate VE value.

After Player 2 casted this second Cure3, he did indeed take over TE. Hate switched back to Player 1 after approximately 14 seconds (tested 10 times, same value every test). Using the known VE decay formula, I conclude that the VE for 180HP recovered is 60*14 = 840 VE.

Test 4b: VE of Cure 2 for 90 HP healed on Self
Player #1: {Equalized Enmity as Above} Total CE = 201
Player #2: {Equalized Enmity as Above} Total CE = 201 (HATE)

Player #1: Casts Sleep Total CE = 521 (HATE)
Player #1: Casts Cure1 for 0 HP on self x 6 Total CE = 527 (HATE)
Player #2: Casts Cure 3 for 180HP on Self x2 Total CE = 461

At this point, the player's enmity levels is such that if player #2 now casts Cure2 for 90hp on self (65 CE), the total CE difference between the players will be 1 unit. Player 2 will have hate due to the unknown VE, but player 1 will soon have hate again after this VE decays. By timing how long this takes to occur and using the known VE decay formula, I can calculate the approximate VE value.

After Player 2 casted cure2, he did indeed take over TE. Hate switched back to Player 1 after approximately 7 seconds (tested 10 times, same value every test). Using the known VE decay formula, I conclude that the VE for 90HP recovered is 60*7 = 420 VE.

Again, there is a pattern to these VE values. 840VE/180HP = 420VE/90HP. So I hypothesize that the VE value for HP healed is calculated as 14VE/3HP. It should be noted that I have not performed any follow-up testing for this VE value as of yet.


While the finding of such simple numbers to explain CE and VE excites me, I can't help but feel I've committed a gross error somewhere along the line. If it had been so easily calculated, it would have been done long ago by others. So, again, I'm posting my data here so that others can either confirm or, I think more likely, kindly point out my mistakes so that I can reformulate a means to test. Thank you in advance.

edit: changed VE formula to 14VE/3HP, failed to simplify my fraction before lol.

[ismarc]

While the finding of such simple numbers to explain CE and VE excites me, I can't help but feel I've committed a gross error somewhere along the line. If it had been so easily calculated, it would have been done long ago by others. So, again, I'm posting my data here so that others can either confirm or, I think more likely, kindly point out my mistakes so that I can reformulate a means to test. Thank you in advance.

One thing that you don't detail that, depending how it was done, could easily have tainted the testing...
What method was used for creating low-hps to cure? Was the testing started at low/medium hp, was gear swapping used, or did you allow the monster to do damage to cause damage down? I'm assuming you either started with lowered hp or gear swapped, but since it wasn't specified thought it worth mentioning considering allow the mob to do damage could alter the outcome.

[Melari]

No damage was taken by either player during any of the tests (used lvl0 mobs). Actually had to throw out one test set due to player1 getting crit hit for 1 damage lol, so it was excluded from testing.

All hp curing came from "cure swaps." i.e. Started at a lower hp value, added appropriate items to increase hp (no items were used that had any +/- enmity bonuses) by a set amount, then cast the spell. I only had enough equipment available to achieve a maximum of about 205 this way, which is why I did not test higher than cure3 for 180.

[Kaeko]

You test for CE is similar to our early tests for CE which showed the formula at lvl75 is...

CE = 0.7275 x HP Cured

13/18 is about 0.7222.

The VE numbers are new though which is great.

Also, to get the most accurate modifier, you want to use the largest cure possible.

[asnman]

Just out of curiosity did anyone try seeing if the Marduk body had any -enmity? maybe its just me but Notice how each of the pieces required to complete the set always had a stat that the final piece had? I guess it would make sense the legs had -enmity cause the level 35 piece has -enmity too. I also noticed the level35 body also had some -enmity, maybe its hidden also? Sorry if this was mentioned before, please don't hurt me

[Max™]

I mentioned that too, but it doesn't seem to have an exact correlation to the final stats.

[Kaeko]

Here is some info on cures...

Testing CE on Cure is done by having 2 players at equal CE. 1 player then cures for X amount (the higher the amount the better) and the other player uses CE producing moves unitl hate is evened out. Using /BRD is the easiest way in my opinion.

Testing VE requires that you first know the amount of HP that will be cured along with the amount of CE generated. For instance, if you know a cure will generate 300 CE, the player using that cure needs to be 301 CE less than the other. When the cure occurs, the curer will have 1 less CE in addition to the VE generated. The VE is then estimated by clocking the number of seconds using some recast on a JA - like Divine Seal and seeing how many seconds pass before hate switches back.

1) The amount of CE and VE is not, I repeat, NOT, based on the target's current or max HP, but the level of the target.

Simple test, cure a low HP job for 210 HP, then a high HP job for 210 HP both at 75. The amount of hate is the same. You can try this for situations where the target is not near max HP (near death) - the same will occur.

2) The amount of CE and VE generated do NOT vary if you cure for the same amount on Cure 1, 2, 3, or 4. (all same)

Cure a target for some set amount - say 30 HP and check CE or VE (different tests, but these have been outlined varioust times). Try using Cure 1, 2, 3, or 4. The same numbers will result regardless of the type of cure Tier you use.

3) The amount of CE and VE generated decreases as the target's level increases...

These were tested by just producing the highest possible cure on players of various levels... This is pretty painful to test on lower levels and it's kind of slow, so please forgive the slow progress on this. I use SCH/WHM for high end cures - Rapture Divine Seal Cure IV cures up to 1292 HP. Currently, these are the numbers I have...

@75: 1029 HP Cured gives 748 CE and roughly 4500 VE (75 seconds)... CE modifier ~0.7275 , VE modifier ~4.4

@70: 410 HP Cured gives 315 CE (VE untested)... CE modifier ~0.770

@40: 410 HP Cured gives 469 CE and roughly 2820 VE (46 seconds)... CE modifier ~1.1440 , VE modifier ~6.9

@5: 81 HP Cured gives 235 CE (VE untested)... CE modifier ~2.91

The goal is to find some type of pattern as to how CE and VE scale with level.

Probably the most important part is already figured out though... which is the Cure formulas at level 75. Assuming the target you are curing is 75 (which is 99% true in most endgame), the amount of hate generated is...

CE = 0.7275 x HP Cured
VE = 4.4 x HP Cured

Your typical PLD Cure IV is 390 HP, which generates 283 CE and ~1716 VE, which is essentially a Provoke with added CE. Extremely strong. You also have to weigh in the hate loss when getting hit if it's the PLD getting hit... but a PLD/NIN curing other players is very strong.

[Spekkio]

Just out of curiosity did anyone try seeing if the Marduk body had any -enmity? maybe its just me but Notice how each of the pieces required to complete the set always had a stat that the final piece had? I guess it would make sense the legs had -enmity cause the level 35 piece has -enmity too. I also noticed the level35 body also had some -enmity, maybe its hidden also? Sorry if this was mentioned before, please don't hurt me

back 2 pages, already covered. (might have to go back 3 or 4 for the full story.)

[Kaeko]

I'm working on the CE multiplier values for curing targets of various levels. I need some help from anyone familiar with curve fitting...

Basically the way I think cures work is like this...

[CE] = [CE Multiplier] x [HP Cured]
[VE] = [VE Multiplier] x [HP Cured]

The multiplier varies by the level of the target you cure 1-75. I have multipliers at various levels and am just trying to find the overall "curve" or equation that can define them all without actually testing at each level.

I am going to list the level of target, the HP cured, the resulting CE gained, then finally the range of multiplier it can be assuming CE is rounded down...

75: 1029 HP / 748 CE >> 0.7269-0.7279
70: 418 HP / 321 CE >> 0.7679-0.7703
52: 380 HP / 369 CE >> 0.9711-0.9737
40: 410 HP / 469 CE >> 1.1439-1.1463
37: 418 HP / 506 CE >> 1.2105-1.2129
30: 418 HP / 577 CE >> 1.3804-1.3828
15: 85 HP / 154 CE >> 1.8118-1.8235
5: 81 HP / 235 CE >> 2.9012-2.9136

I'm really looking for the best type of equation to characterize this set of data... inverse, logrithemic, etc.

[Freysi]

Didn't feel like making and hosting my own pic so found one that describes the relationship. If you plot the target level on the Y axis, and the CE/HP on the X axis you get something that looks like this:

Clicky

Which basically means that there is an inverse relationship between Target Level and CE per HP healed. Is that what you were going for? The information is pretty sparse, but it still is pretty clear that it creates this curve.