Auspice +ACC testing (July2009)

[RyaWHM]

So, in my insomnia, I worked up the courage/patience to test for ~2hrs.
The results are really nice/encouraging IMO. However, I need help with
a few things, and there are some really smart people here, hopefully
we can conclude some things!

NOTE: I accept that I am probably more willing than I am qualified to
be doing these tests; please let me know if I messed something up, so that
I can re-test (yuck) if necessary. Thanks~

Setup:
• WHM75/MNK37 (no dagger skill) using Trainee's Needle
• ACC should be 33 (DEX=60 and no ACC gear other than Swift Belt)
• Afflatus Misery kept up the whole time; did not wear off either
• Auspice reapplied every 2:15~2:30 (in case of interrupt) kept up 100%
• Auspice maintained even between battles (so, really truly 100%)
• One "control" before beginning the test (so, no misery/auspice)
• Subject= Robber Crab @ Kuftal

----------

Robber Crab #0: Lv.61 (control)
*no misery/auspice
98 hits
386 misses
20.25% accuracy
??? hp*
??? duration*
60 exp
(*switched to club and killed to verify level; should be plenty
when we know I'll have minimum hit-rate anyway; back to
dagger and ready to begin testing)

----------

Robber Crab #1: Lv.61
212 hits
416 misses
33.76% accuracy
~3355hp
25m37s duration
60 exp

----------

Robber Crab #2: Lv.63
240 hits
561 misses
29.96% accuracy
~3512hp
33m33s duration (ugh, zzzzz)
66 exp

----------

Robber Crab #3: Lv.60
209 hits
320 misses
39.51% accuracy
~3268hp
21m10s duration
56 exp

----------

Observations:
• acc gradually rose during Robber Crab #1, from ~20% steadily rising to its overall ~33% (reached about half way through the fight) and didn't fluctuate much after that
• acc did not rise/fluctuate very much on #2 and #3, it felt like I kept whatever ACC bonus I'd built up (although can't be sure) -- just the normal brief fluctuation at start of fight
• with this in mind, the acc bonus did seem to hit a cap, which I guess isn't surprising?

----------

Questions:
• was I indeed below the minimum ACC% before Auspice? (I assumed so)
• how much "natural" ACC (skill/gear; non-auspice) would I need to reach the ~20% point, where I break the minimum?
• how much "natural" ACC would I need to otherwise reach the hit-rates I experienced (in the ~34% range)?
(and, ultimately, the real question...)
• can we actually determine a maximum +ACC value that I reached?

----------

Thank you!
Thanks in advance, as I'm sure several nice, mathy people will come
hopping out of bushes and wow me with numbery fun.

[Kriz]

I approve of this topic.

Good job taking the initiative. Only thing I could suggest is getting those accuracy - rings if you really are worried you aren't at the acc lower cap, which with 0 skill I'm pretty sure you are!

[rog]

You can't use the standard /check testing?

[RyaWHM]

You can't use the standard /check testing?

Sadly no, it seems like we can't -- both Izzy and I have tried that method (and pretty sure others too), but were unable to cause a change in the "/check" readings.

---> see my attempt
---> see Izzy's attempt

...but hopefully this should be more conclusive than those previous attempts were. Fingers crossed!

[shaddix]

I went from 66 to ~75 or so on lv75 steelshells, 2 crabs for control and 2-3 with auspice+misery

[arus2001]

Is the +ACC always on, or only apply after a miss like a ghetto Zanshin? The latter was what I was guessing about the spell info, but SE's botched those tags before.

Edit: I bring this up because it might not be accounting a true yield in the accuracy bonus since if it only kicks in after a miss, you couldn't count consecutive hits toward your total accuracy. And if I'm not mistaken, hit rate is basically capped at 20%. Some of the above numbers were in the ballpark of 50/50 hit/miss, but the last one had way more hits relative to misses. Plus it just seems crazy SE would give a spell 40-50 ACC while the -EVA on Gravity is relatively meager, WHMs have a higher Club skill than RDM's sword/dagger, and Composure only panned out to be like +15 ACC.

[Valyana]

Sounds like the +Acc is cumulative, which is very nice.

I don't know crab AGI rank, so I'm going to guess C. From that a L60 crab would have 40 AGI, L61-63 43 AGI. So

L60: 142+(10*4.8 )=190, +40/2 = 210 eva, -15*4= 150 effective eva
L61: 142+(11*4.8 )+43/2-14*4=159 effective eva
L62: 142+(12*4.8 )+43/2-13*4=168 effective eva
L63: 142+(13*4.8 )+43/2-12*4=177 effective eva

#1: 33.76 = 75 + (acc-159)/2 -> acc = 76.5
#2: 29.96 = 75 + (acc-177)/2 -> acc = 86.9
#3: 39.51 = 75 + (acc-150)/2 -> acc = 79

So it seems the cap is about +45-50 acc. Which is very nice; the difference between B+ and A+ weapon is only 20 skill.

[chiyio]

I'm not sure if this was accounted for, from the looks of it it may not have been, however according to evilwiki, there's apparently some accuracy correction based on mob level. Has this been accounted for, or is it irrelevant?

* When a Player Character attacks a lower level monster, the attacker gains a 4 accuracy bonus per level difference.

Example:

* Lv75 character with 375 accuracy fighting a Lv70 monster with 365 Evasion

Level difference = 5
Accuracy adjustment = 4 acc/level * 5 level = 20 accuracy
New player acc = 395 → Player acc - mob evasion = 30
Hit rate = 75% + 30*0.5% = 90% hit rate

[Valyana]

Yes, that's what the -15*4 is in my post.

[RyaWHM]

Thank you Valyana!!

I prolly oughtta try some different [higher] mobs to see if we get the same +ACC estimate, but for now anyway, it looks very powerful and delicious!

[TummieGaruda]

Sounds like the +Acc is cumulative, which is very nice.

I don't know crab AGI rank, so I'm going to guess C. From that a L60 crab would have 40 AGI, L61-63 43 AGI. So

L60: 142+(10*4.8 )=190, +40/2 = 210 eva, -15*4= 150 effective eva
L61: 142+(11*4.8 )+43/2-14*4=159 effective eva
L62: 142+(12*4.8 )+43/2-13*4=168 effective eva
L63: 142+(13*4.8 )+43/2-12*4=177 effective eva

#1: 33.76 = 75 + (acc-159)/2 -> acc = 76.5
#2: 29.96 = 75 + (acc-177)/2 -> acc = 86.9
#3: 39.51 = 75 + (acc-150)/2 -> acc = 79

So it seems the cap is about +45-50 acc. Which is very nice; the difference between B+ and A+ weapon is only 20 skill.

Crab AGI rank is E

This is a work in progress, I have other unfinished sheets I am working on

Monster Stats

I am working on getting other mobs most fought against calculated and documented.

[Cream Soda]

Don't forget level correlation. Lv 75 would grant a 4 acc/lv bonus so on the lv 61 crabs, your 33 acc + (14 lvs x 4 acc, aka 56 acc) = 89

[Lokim]

Just figured I'd throw this out there: Auspice.. Testing? - Killing Ifrit

These tests seem to confirm that the effect is cumulative.

[shaddix]

I would like to know if it works more like you miss, then the next 20s you have acc bonus, which then wears(or maybe tapers). So then stacking, but then leveling off depending on swing speed.

When I tested mine, I got lower than claimed returns of a whm with 95% acc on birds with meat and full haste setup, but I wasn't /nin and I was tanking, so also casting a lot of shit.

So either that 95% with meat is bullshit, or it works somehow what I'm saying, or I just didn't fight the crabs long enough. But I didn't see it go much higher than 75% from 66% on my tests vs an EM.

[RyaWHM]

Series #2
*small sample sizes, so not that valuable

Spoiler: show

Orcish Dragonbranders @ Jugner [S]
approximately 315 accuracy
(no food)

----------

Orcish Dragonbrander #0: Lv.77 (control)
*no misery/auspice
98 hits
33 misses
74.81% accuracy
~5595 hp
6m39s duration
160 exp
(*after kill, turned on misery+auspice; ready to begin testing)

----------

Orcish Dragonbrander #1: Lv.76
79 hits
13 misses
85.87% accuracy
~5454hp
5m06s duration
130 exp

----------

Orcish Dragonbrander #2: Lv.77
89 hits
12 misses
88.11% accuracy
~5600hp
5m17s duration
160 exp

----------

...and I DC'd during the 3rd one

Anyway, I wanna test some more of these orcs, because they don't
give a very big sample size per fight ... if I do, I'll just edit I guess.

----------
EDIT: Series #3
Orcish Dragonbranders @ Jugner [S]
approximately 315 accuracy
(no food)
----------

Orcish Dragonbrander #3 - #7: Lv.76 (all the same level)
*Misery + Auspice
378 hits
87 misses
81.29% accuracy
130 exp (each)

----------

Orcish Dragonbrander #8 - #12: Lv.76 (all the same level)
*no misery/auspice
402 hits
118 misses
77.31% accuracy
130 exp (each)
(*went back for more NON-AUSPICE testing)

----------

[Kroot]

a good way to get large sample size for accuracy is to hit goldfish for 0 dmg. you can afk if you are fine with setting up a script to auto cast

[arus2001]

a good way to get large sample size for accuracy is to hit goldfish for 0 dmg. you can afk if you are fine with setting up a script to auto cast

Won't work because Auspice and Misery will have the Enlight going.

[RyaWHM]

a good way to get large sample size for accuracy is to hit goldfish for 0 dmg.

I can stab for zero, but enlight would anger the fishies.
EDIT: I was too slow!!

Also wanted to comment on...

I got lower than claimed returns of a whm with 95% acc on birds with meat and full haste setup

Me too -- I'm really skeptical about that one (I have heard that claim also).

If the +ACC fades fast, then maybe he was using a Kraken Club to get way more misses (even at high/capped acc) in a short amount of time?

I'll see if I can convince some friends to give me a DD's slot in a SHORT (casual) bird PT. That way I can swing lots and lots, get a large sample on mobs we know everything about, and only hafta stop for Auspice.

(yeah right, keep dreaming Rya)

[Tsuko_Asura]

I can explain the above. The white mages who 'tested' it are idiots.

[RyaWHM]

Bump time!
I finally got that merit party test, and this time I have a K-club!

----------

@Bird camp
*(not my parser here, just copy/pasting what was emailed to me)
• WHM75/NIN37 (279 skill) using Perdu/Kraken
• ACC should be 375 (321 before food; bream sushi NQ)
• Haste/Auspice/Misery kept up the whole time
• COR & BRD did NOT provide ACC just march/minuet/drk/cor
• Subject= Greater Colibri & Wivre @ Bhaflau

----------

Roughly 80 min of killage
3573 swings
2924 hits
649 misses
81.84% accuracy
(and just for flavor:
60,443 melee damage
72,953 WS damage)

----------

So basically, 375 acc gave me 81.84% hit.

How much of an acc boost did I get? (Please assist.)

*I dunno how much more difficult Wivres make the math on this, but
we didn't exactly seek them out, just the normal "MUST CHAIN!" kills.
Maybe 1/10 or 1/12 mobs was a wivre? Something like that?
(I don't merit regularly enough anymore, to know "the norm" for that;
we were making roughly 30k/hr if that helps)