Death Blossom

[Thorshammer]

Advanced Version Update 4.22.09

Ahlah has been removed from the forums for seven days. We apologize for any inconvenience.

poor Ahlah :D

[Raineer]

accuracy floors and accuracy caps are the same (5%/95%)

Kaeko, I've heard this thrown around. Is this the assumption that magic will full resist a floor 5% of the time, or just resist floor 5% of the time.

The reason I mention this (for example) would be sleeps in Dynamis. There is no way I see full resists 5% of the time (let's ignore the recent patch). After sleeping in hundreds of Dynamis I would be confident to say a full resist from a non-paladin is less than 1%. Early wakers might be in the 5% range, and in that respect I could see the statistic being true.

[Kaeko]

Kaeko, I've heard this thrown around. Is this the assumption that magic will full resist a floor 5% of the time, or just resist floor 5% of the time.

The reason I mention this (for example) would be sleeps in Dynamis. There is no way I see full resists 5% of the time (let's ignore the recent patch). After sleeping in hundreds of Dynamis I would be confident to say a full resist from a non-paladin is less than 1%. Early wakers might be in the 5% range, and in that respect I could see the statistic being true.

I was mainly speaking in terms of elemental magic damage, but I agree that enfeebles are more accurate than 95% on a lot of things, especially with elemental seal. I'm really just throwing around theories at this point with very little to back myself up, but the best way I can reason it:

Enfeeble resists work like elemental damage resists (an assumption I talked about in an earlier post), where if you have a probability of landing something successfully of p, the chance to resist various tiers (1/2, 1/4, etc.) is...

0 resist p
1/2 resist p(1-p)
1/4 resist p(1-p)^2 ------ Anything here or below is just an outright full sleep resist
1/8 resist (1-p)^3

When you sleep something, p is capped at 95%, so there is a 5% chance the 1/2 resist will kick in. You won't notice this 1/2 resist kicked in until the sleep wears off early though so for all you know it landed fine at this point. But after the 1/2 resist, the game checks again for another resist. If you fail this check again (basically failed twice in a row), you get a 1/4 resist. Of course for sleeps, no such tier exists and at this point you just get outright resisted.

So based on this THEORY, the chance that you would cast sleep at capped accuracy and get at least 1/2 resisted is indeed 5%. But the chance of you getting outright resisted is 0.05^2, or 0.0025 (0.25%), which I think fits in well with eyeballing. It says you'll get randomly resisted fully once every 400 attempts with capped accuracy on a non-sleep resist trait mob.

Again, I'll stress this is just the best theory I can come up with that happens to fit with trusted data.

EDITED for Math (ty Teorem)

[Raineer]

I agree with all that.

[Teorem]

The idea was right, but the expressions aren't accurate. The elemental resist tiers were:

Given the probability of no resistance, p,

No resistance: p
1/2 resistance: p(1-p)
1/4 resistance: p(1-p)^2
1/8 resistance: (1-p)^3

There were no "etc." terms because the model in the tests only assumed up to a 1/8 resistance tier, and all of the probabilities have to add up to 1. They (p(1-p)^n and (1-p)^n) are close when the value of p is high, but definitely not as p decreases.

[arus2001]

I don't have a K.Club, but I could poke some Colibri with Enaero at varying levels of Enhancing Magic to see if we can maybe peg something. Would the targets be Lessers or the plain ol' Colibri people EXP on 60-64ish?

[Kaeko]

The idea was right, but the expressions aren't accurate. The elemental resist tiers were:

Given the probability of no resistance, p,

No resistance: p
1/2 resistance: p(1-p)
1/4 resistance: p(1-p)^2
1/8 resistance: (1-p)^3

There were no "etc." terms because the model in the tests only assumed up to a 1/8 resistance tier, and all of the probabilities have to add up to 1. They (p(1-p)^n and (1-p)^n) are close when the value of p is high, but definitely not as p decreases.

Thanks for the correction, you're absolutely right. I edited the previous posts to remove the bad math.

[Spider-Dan]

Some similarities between the 2 are 1) they both have 'level corrections' that penalize fighting mobs higher level than you, 2) inflection point at 50%, 3) accuracy floors and accuracy caps are the same (5%/95%), and 4) 1 skill = 1 acc in magic, 1 skill = 1 acc in melee until after 200 skill, then (0.9).

I thought the melee accuracy floor was 20% (which corresponds to the evasion cap of 80%)?

Also, while you addressed MACC vs. MEVA, you didn't really answer my question about MACC vs. -elemental resistance. Are these two presumed to be inversely proportional?

[pchan]

sleep only has 1/2 resist partial resist possiblity so it's :

non resist : p (=1:30 sleepII)
1/2 resist : p(1-p) (45 sec)
full resist : (1-p)^2 ("resist !")

[Ultimagi]

Wow this thread really took off over night.

I got Enthunder II and will try to test it out in combination with Death Blossom sometime soon.

For people who question the usefulness of applying 1 Death Blossom at the start of a fight: Troubadour Elegy lasts 7 minutes 12 seconds.

[Seraph]

Wow this thread really took off over night.

I got Enthunder II and will try to test it out in combination with Death Blossom sometime soon.

For people who question the usefulness of applying 1 Death Blossom at the start of a fight: Troubadour Elegy lasts 7 minutes 12 seconds.

Was this a single sample, or did it hit that duration multiple times?

[Ultimagi]

It is common knowledge. The base duration is 3 minutes, then Horn +1 adds 36 seconds. Troubadour doubles that up to 7 minutes 12 seconds.

Edit: Carnwenhan can add 1 minute 12 seconds more.

[The Blackrose]

am I the only one that finds it funny the ban avatar has a word, that if used in us normal folk posts, results in a temp ban?

That's the ban avatar for a default vB install. It looks like the forum update got rid of the custom avatar set up for it, too.

Wow this thread really took off over night.

That sort of happens when you finally get around to kicking out the threadshitter.

[Seraph]

It is common knowledge. The base duration is 3 minutes, then Horn +1 adds 36 seconds. Troubadour doubles that up to 7 minutes 12 seconds.

Edit: Carnwenhan can add 1 minute 12 seconds more.

I was more curious if there were still 'partial resists' even when using Death Blossom.

[Fng]

If people decide that DB is required for fights. These are the only things that will work:

SAM giving RDM TP
RDM melee (lol 0s)
RDM/SAM (what's the duration of DB additional effect? i dont know)
RDM/whatever w/ sleep pots (aka doing nothing else)

Otherwise as soon as the RDM drops ally and takes off blockaid to get into a tp group... you will get spammed with cures from jackass competition.

[Ultimagi]

Since everything in the game, without any sort of Magic Evasion or Resistance down, is either weak or immune to Elegy, I can't say I've ever noticed a partial resist on it.

[Spekkio]

if resist sleep is applied after ES, why do you never see full sleep resists on ouryu. is he not a pld and as such would thus inherit resist sleep?

[Kaeko]

if resist sleep is applied after ES, why do you never see full sleep resists on ouryu. is he not a pld and as such would thus inherit resist sleep?

I believe this is just misinformation popularized by oldWiki, which attempts to classify EVERY NM as one of the 20 jobs. Just because a mob uses a 2hr or spells from a particular job doesn't necessarily imply that SE programmed it to have everything that job has.

With beastmen it's pretty decisive that some have particular jobs (I mean, it's called Kindred Paladin) and would inherit job traits, but I don't see why a mob must have every trait a true PLD has just because he uses Invincible... If that were the case then AV's got an awful lot of job traits.

I'm not proclaiming that oldWiki is necessarily wrong - there's not enough evidence to show that Ouryu is not somehow designated a PLD by SE's programming; but to answer your question, I don't think Ouryu has resist sleep traits.

[Ultimagi]

If that were the case then AV's got an awful lot of job traits.

Absolute Virtue inherits the Job Traits of the last job's ability he used.

[Gergall]

Otherwise as soon as the RDM drops ally and takes off blockaid to get into a tp group... you will get spammed with cures from jackass competition.

More like the "jackass competition" would kill all of the TP mobs.

But if there isn't enough competition around to kill everything, just have the TP group CFH on the mobs so they never need to invite the RDM.

Alternative solution: the RDM leaves blockaid on the whole time, and creates the TP group rather than join it.