Elemental Seal Question

[Kirschy]

I jumped the gun there on shooting out 95% as a cap. I think the accuracy portion of magic spells elemental and enfeeble is probably higher based on empirical evidence such as ES Sleeping and running a tests on Lv 0 mobs. I know for certian it can't be 100%, as I've seen spells resist on lv 0 mobs /w capped skill at 75. I've seen everything from Carnage Elegy on a Gjalla BRD resist to Slow II /w merits on mobs outside Bastok. But yea it's gotta be higher than the 95% from melee accuracy which is easy to see when you miss 1 in 20 hits.

It would be hard ascertain if ES can possibly grant a 100% land rate. Again on empirical evidence, I've never seen an ES spell miss unless the mob had a resist trait, immunity, or resistance build up. (But thats not enough proof to say it can't happen.) This is different than Sneak Attack which always lands your next hit regardless of your accuracy or evasion of the mob you're fighting.

Like the example /w a Lv 20 BLM casting ES Sleep on Jorm, ES doesnt grant you the ability to automaticalli land a spell.

[Raineer]

I can certainly understand the 95% reasoning, since that number shows up elsewhere.

I just doubt it because, as was stated above on the Paralyze II test, there have been times I have seen many hundreds of Sleep II casts and not a resist. I sleep-nuked more things than I care to admit on RDM, and would rarely see a resist without ES, so unless we are saying there is an "inherent miss" in ES I don't believe there is a cap.

In regards to Kirschy's statement above, this may also mean Sleep has different characteristics than a bard song or some other magic. I surely have seen the SS's of mischarms or missed lullabys on level 0 mobs, but I never experienced one with sleep.

[Kaylia]

There was a thread about this a while ago, I'm sure someone could probably find it. But it just increases the accuracy. This accuracy just happens to cap at 95%. If you take a Lv 20 BLM up to Jormungand and try to ES Sleep, majority of the time it won't land. The extra accuracy in this case doesn't come close the 95% cap.

Really? I dont remember the blm not landing sleep?


If it is the case, I will make a guess and say Elemental seal give 100% accuracy, minus a level correction and special trait bonus. I cant think of any situation where my BLM got resisted with ES unless the monsters have an innate resistance (barspell, shell or innate resist trait). However, ES+nuke is very likely to get resisted on mobs with spell resistance (fire on tiamat and the like). On the other hand, ES always work on certain mobs, when normal sleep usually fail (try landing sleep on an ahkriman without ES).

I'm under the impression buff and damage spell resistance is calculated differently. 95-96% accuracy on damage spell seem about right, but enfeeble probably cap much higher than this. I will need to go test thunder 4 on bunny.


Pure speculation here:
- Damage spells cap at 95% acc, enfeeble at 100%
- Elemental seal ignores ~200 "elemental resist" (ie: barfire) at 75. Would explain why certain mobs with strong elemental affinity will resist (999-200=800 is still a ton) when other won't resist (400-200 = 200 -> sleepable).
- If the spells is unobstructed by elemental resist and/or trait, it will land.
- There is a level correction on spell accuracy or ES buff isn't as potent depending of your level

[shukuffxi]

I remember a similar conversation a while a go, someone ran up to Jorm THF75/BLM37 and ES > Sleep and it landed, there was a SS, but this was pretty long ago.

[Kaylia]

I remember a similar conversation a while a go, someone ran up to Jorm THF75/BLM37 and ES > Sleep and it landed, there was a SS, but this was pretty long ago.

Would be easy to test if there is a level correction doing this. Try a melee/75 without enfeebling skill vs a low level jobs. If the lv37 is resisted, the lv 75 melee should be resisted as well.

[Kirschy]

http://img228.imageshack.us/img228/5...0121820oi1.png
http://img440.imageshack.us/img440/2...0123214sk3.png

I honestly recall the thread from a while back as having the BLM resisted on their sleep, but I must be mistaken. I went 2/2 on ES Sleeps.

EDIT: Forgot to add. I was 43DNC/21BLM, 61 Enfeeble skill(No Merits).

[Vxsote]

Don't forget that unlike melee, there are partial resists for both nukes and debuffs. I could probably believe that the chance to land a spell fully unresisted is 95%, but chance of a full resist is a lot less. After all, in my long days of testing nukes on bunnies, I've seen my share of half resists, but I've never seen a 1/8 or 1/16.

[Kaylia]

Don't forget that unlike melee, there are partial resists for both nukes and debuffs. I could probably believe that the chance to land a spell fully unresisted is 95%, but chance of a full resist is a lot less. After all, in my long days of testing nukes on bunnies, I've seen my share of half resists, but I've never seen a 1/8 or 1/16.

Half resist on debuff shorten their duration, however, I'm fairly certain ES sleep is always full length

[FailureMidgard]

Don't forget that unlike melee, there are partial resists for both nukes and debuffs. I could probably believe that the chance to land a spell fully unresisted is 95%, but chance of a full resist is a lot less. After all, in my long days of testing nukes on bunnies, I've seen my share of half resists, but I've never seen a 1/8 or 1/16.

Half resist on debuff shorten their duration, however, I'm fairly certain ES sleep is always full length

Just finished a run of the Tonberry BC60 for torques, on 1 attempt with an ES sleepga2 at the start, 1 of the Tonberries broke before the other 3 mobs (was a half resist, only happened once out of ~10 runs with 4 targets per sleepga).

[Kaylia]

Don't forget that unlike melee, there are partial resists for both nukes and debuffs. I could probably believe that the chance to land a spell fully unresisted is 95%, but chance of a full resist is a lot less. After all, in my long days of testing nukes on bunnies, I've seen my share of half resists, but I've never seen a 1/8 or 1/16.

Half resist on debuff shorten their duration, however, I'm fairly certain ES sleep is always full length

Just finished a run of the Tonberry BC60 for torques, on 1 attempt with an ES sleepga2 at the start, 1 of the Tonberries broke before the other 3 mobs (was a half resist, only happened once out of ~10 runs with 4 targets per sleepga).

Did it have shell on?

[FailureMidgard]

Blm, Thf, Smn, Nin are the 4 mobs (I don't recall what element the smn pet was, but I'm 100% sure it wasn't light). I don't even recall which one woke up, I just recall seeing it wake up, and being surprised at the fact. Anyway, I don't see any way for those 4 jobs to have shell or any sort of resist sleep trait?

[Ashira]

Recently had a 50% resist on an ES Burst II in Limbus on a non-resistant mob. Shit like that's embarrassing.

[Kaylia]

Blm, Thf, Smn, Nin are the 4 mobs (I don't recall what element the smn pet was, but I'm 100% sure it wasn't light). I don't even recall which one woke up, I just recall seeing it wake up, and being surprised at the fact. Anyway, I don't see any way for those 4 jobs to have shell or any sort of resist sleep trait?

I would put my money on double light nerfing this sleep.

[Pendas]

The 95% Cap does exist for Weapon Accuracy and for Crafting, its impossible to reach a 100% hit/success rate with either no matter what level you are in correlation with the mob/synth. Magic Accuracy has just always been an enigma that it very difficult to test, so everyone has just agreed that if this cap exists on Weapon Accuracy and Crafting that it must be true for Magic also.

[Kaylia]

The 95% Cap does exist for Weapon Accuracy and for Crafting, its impossible to reach a 100% hit/success rate with either no matter what level you are in correlation with the mob/synth. Magic Accuracy has just always been an enigma that it very difficult to test, so everyone has just agreed that if this cap exists on Weapon Accuracy and Crafting that it must be true for Magic also.

Like I said a few post ago, there is definitively a difference between enfeebling cap and elemental spell cap. One caps around 95% (high level mobs vs lv 1 job or high lv vs lv 0 mobs), while the other is close to 99% (no cap probably)

[Arkanos]

Blm, Thf, Smn, Nin are the 4 mobs (I don't recall what element the smn pet was, but I'm 100% sure it wasn't light). I don't even recall which one woke up, I just recall seeing it wake up, and being surprised at the fact. Anyway, I don't see any way for those 4 jobs to have shell or any sort of resist sleep trait?

I would put my money on double light nerfing this sleep.

We recently had 1 out of the 9 or so gears in the first room of 5FN Zhayolm Remnants waking up early from ES Sleepga and raping our whm before he could /logout. It was thundersday.

[Spekkio]

i think the difference is what was mentioned above. with a melee hit or a crafting synth it's a binary decision. you hit or you didn't. meanwhile, your sleep2 can land for a fractional duration but we still classify that as a "hit" because we have no method to determine a partial resist from an unresisted sleep unless it wakes early. while 5% of our sleeps may have a curtailed duration, the number of flat out resists (what we're classifying as a miss) must first roll that 1/20 roll to miss then it must roll so low that the resist is a 100% resist. i don't know the math behind the different tiers of resists and their probabilities, but i'd imagine the odds of rolling a 100% failure on it are not good. i think there's 4 tiers for nukes (1/2, 1/4, 1/8, and 1/16) so let's assume that these all apply to enfeebles and add on a 5th tier, full resist. assume all 5 have equal odds (i base this on pure speculation to discern vague numbers. this could be totally off. remember that!) 1/20*1/5=1/100 odds that you get a full resist on sleep, which closer jives with our expectations.

[Raineer]

Don't forget that unlike melee, there are partial resists for both nukes and debuffs. I could probably believe that the chance to land a spell fully unresisted is 95%, but chance of a full resist is a lot less. After all, in my long days of testing nukes on bunnies, I've seen my share of half resists, but I've never seen a 1/8 or 1/16.

Half resist on debuff shorten their duration, however, I'm fairly certain ES sleep is always full length

Just finished a run of the Tonberry BC60 for torques, on 1 attempt with an ES sleepga2 at the start, 1 of the Tonberries broke before the other 3 mobs (was a half resist, only happened once out of ~10 runs with 4 targets per sleepga).

Sleep duration is determined by skill anyways, which might make it not *exactly* the same stat as ES (meaning, RDM+ES would probably guarantee 60 seconds out of it).

Still, pretty rare to wake early on a ES, I would agree.

[Vxsote]

Spekkio seems to understand what I'm getting at. I'm fairly certain that there are fewer divisions of partial resist for enfeebles compared to nukes, and I'm also certain that the rates of each resist "tier" are not uniform. But that's pretty much beside the point. Partial resists on debuffs happen more than most people notice because often the mob either doesn't live long enough to find out, people are too busy to notice an early wear-off, or whatever other reasons you can dream up. Any more detailed estimate of partial resist rate cap than has already been discussed here would certainly require a long, involved test. Perhaps it's time to sleep some trains.

Sleep duration is determined by skill anyways

Aside from the partial resist mechanic, and the fact that low skill will result in more resists (full and partial), this is just plain wrong. An unresisted sleep is 60 seconds, and an unresisted sleep II is 90 seconds.

[shieldsofwindurst]

i think the difference is what was mentioned above. with a melee hit or a crafting synth it's a binary decision. you hit or you didn't. meanwhile, your sleep2 can land for a fractional duration but we still classify that as a "hit" because we have no method to determine a partial resist from an unresisted sleep unless it wakes early. while 5% of our sleeps may have a curtailed duration, the number of flat out resists (what we're classifying as a miss) must first roll that 1/20 roll to miss then it must roll so low that the resist is a 100% resist. i don't know the math behind the different tiers of resists and their probabilities, but i'd imagine the odds of rolling a 100% failure on it are not good. i think there's 4 tiers for nukes (1/2, 1/4, 1/8, and 1/16) so let's assume that these all apply to enfeebles and add on a 5th tier, full resist. assume all 5 have equal odds (i base this on pure speculation to discern vague numbers. this could be totally off. remember that!) 1/20*1/5=1/100 odds that you get a full resist on sleep, which closer jives with our expectations.

Having been the primary sleeper in dynamis for over 2 years, this seems reasonable to me. I can also testify to the fact that sleep/ga 2 is far more accurate than sleep/ga, and that accuracy is directly tied to elemental skill (duh), as my effectiveness at sleeping dynamis mobs with sleep/ga increased significantly as I stacked more & more elemental skill.

What I'm actually more interested in is if you can combat innate resistance to sleep not by stacking skill gear, but by stacking macc gear. As anyone who's slept pld mobs in dynamis can tell you, there is a point of seriously diminishing returns on skill for blms at least for the cities; I could take off my 8 enfeebling merits & 20 points of skill gear and still probably sleep city mobs just as well, with the same sporadic resist by pld mobs.

What's always piqued my curiosity, however, is that while there's no question I'm far, far beyond said point of diminishing returns against the notoriously sleep-resistant puddings in manaburns, I seem to have just as much success (if not more!) when I mismacro in panic situations and end up casting sleep on puddings in macc gear instead of enfeebling gear. Do I have any evidence that this could be the case? No, but <wild speculation> let's hypothesize that macc is taken into account *after* the skill & resist trait check. Let's then assume (as people have in this thread) that ES directly increases macc instead of skill or w/e (as per the description), thus ES (and thus macc in general) circumvents such things as innate resistance. This would certainly line up with the fact that ES makes pld city mobs and puddings and whatever else with innate sleep resistance as easily sleepable as any other mob </wild speculation>. Of course as we all know, ES doesn't override accumulated resistance (though it is still far more effective than non-ES), thus macc wouldn't either, but again would still help.

In any case, story time-I have about 1 day every 6-12 months where I ES sleepga 2 a set of dynamis mobs and they all or mostly resist. It's a very rare, yet painful moment, but it does happen (if you try long enough). So it's clear that ES isn't 100% even for normal mobs.