New Job Traits & Abilities (12/7/10)

[Rena]

Thanks for figuring this out. I suspected there was some kind of d-stat, but didn't want to buy low level AGI gear and go mess with level 0 monsters. This would predict a Batallia Tiger AGI of 30, which seems about right for a 31-32 monster.

How sure are you of the cap? Tier 1 Scars NMs generally get off 1-3 TP moves before they die when I'm the only melee. I mean, I know I'm building towards 300/200 TP for the vast majority of the time, but more than once a monster's first WS has been at 25%.

Testing against level 20~24ish mobs. there was a clear enough pattern of -0.5% TP/AGI, capping at ~93; adding another 22 AGI produced no benefit in TP reduction.

[Byrthnoth]

Testing against level 20~24ish mobs. there was a clear enough pattern of -0.5% TP/AGI, capping at ~93; adding another 22 AGI produced no benefit in TP reduction.

(93-mobAGI+25)/2 = 50
100 = 93+25-mobAGI
mobAGI = 18

That seems a little low for 20-24 mobs. If we figure out Monster evasion and level, we should be able to solve for its AGI. Which monster set did you test on? I could try to sneak my 11PLD out and find the high to neutral evasion border.

[Rena]

Test was on bull dhalmels in buburimu, and I just realized that, myself... I got the 75 dAGI testing against level -1 mobs where it seemed to match up though, against bumblebees with 66 AGI I got 5.6/cast, and with 69 AGI I got 5.7/cast... most level -1 mob stats are in the 5~7 range.

[Byrthnoth]

Test was on bull dhalmels in buburimu, and I just realized that, myself... I got the 75 dAGI testing against level -1 mobs where it seemed to match up though, against bumblebees with 66 AGI I got 5.6/cast, and with 69 AGI I got 5.7/cast... most level -1 mob stats are in the 5~7 range.

I'm confused, did you say that backwards?

[Rena]

Yes, messed up those numbers, my bad. Anyway, I'm going to do some better, more controlled tests... original testing was flawed as I hadn't even intended to test this, I had just intended to try Reaving Wind out, and one thing led to another.

[Byrthnoth]

Heh, well whatever. =p We know it is linear and we've got m, now we just need b.

m*dAGI+b = Reduction

Your initial proposal was:
.5*dAGI+12.5 = Reduction

My guess in response is something more like:
.5*dAGI+10 = Reduction



Because we have no way to figure out important high level NM's AGI (where this will matter), one might argue it's not worth finding an entirely accurate formula. If you can find the accurate formula then we will have an easy and complete way to find the AGI of any charmable monster, which will be useful for Crit-hit builds post-Abyssea.

Also, it would be theoretically possible to find the AGI of uncharmable monsters this way too, really. If you accept that monsters use a TP move every time they get 100TP when they're <25%, you could count the number of Stone casts it takes at various AGIs (without getting hit) to trigger a TP move below 25% HP and calculate your approximate TP/hit from that.

[firefeng]

Pretty sure Tactical Parry grants 3 TP for a parry. Got numerous (and by numerous, I mean one every hundred mobs or so) 3-4 TP returns from parries. I'm assuming the 4 TP returns were from overflow from me being /SAM.

I'd test it without any Store TP on to be sure, but...no one cares, not even me.

[Rena]

I'm really not finding much that disagrees with (dAGI+25)/2 right now... assuming bumblebees don't have a ridiculously inflated AGI (most -1 mob stats are 5~7), then putting their AGI at 5 matches perfectly with the results I got now. You could bump it up to 7, 8 or so, and get (dAGI+22)/2, but even then, no way of getting this to make sense with the dhalmel tests, unless they just have very low AGI.

AGI - TP/cast
10 - 8.5
65 - 5.8
66 - 5.7
67 - 5.7
68 - 5.6
73 - 5.4
75 - 5.3
78 - 5.1
79 - 5.1
80 - 5.0
92 - 5.0

[Innuendo]

Nice with 5 stratgems gonna see if I can setup my own self skillchain MB on scholar for fun timing is gonna be interesting....

If you're doing it for shiggles rather than damage ouput you can just make the SC with tier 1 spells and then alacrity your (presumably tier 5) MB nuke.
Also, you can solo darkness if you have Retribution unlocked, hurray!

[Foldypaws]

TH+ procs from multiple thieves stack. I saw TH7 as my first proc when I only have TH4, after another THF proced TH6.

[Malacite]

^ Oh lawd

[Byrthnoth]

Did some Striking Flourish testing last night and this morning:
DNC/BST with 98 DEX and 99 STR (0% DA or TA) out in Sauro targeting Yagudo Heralds (horrible choice btw, forgot the ninja emote quest).

I'd build 300TP and find one, then melee it down to about 20-30%, turn, and wait for it to put up Utsusemi: Ichi. Then I'd turn around, let one attack round go through (-2 shadows), and do Striking Flourish -> Shark Bite. It did 600+ damage and gave me a 4TP return all three times I tried it.

Now! What the fuck? I was expecting 600+ damage (3.0 fTP) or 200+ damage (1.0 fTP), but not 4 TP.

So I build 300TP and find one, melee it down to 15%, wait 10 minutes for it to put up Utsu: Ichi. Then I turn around, take down two of its shadows, and do Shark Bite without Striking Flourish. It did 600+ damage and gave me a 4TP return. I guess any hit you land after the WS counts as the "first hit?"

After that, I got really lucky. Meleed up to 100TP, found an Evil Weapon, engaged a Tabar Beak in the distance. Striking Flourish -> Shark Bite (was just going to see what the max TP return was.) I'm single-wielding at this point, and have been since the first test. I luck out and miss the first hit. The weaponskill does ~400+ damage and returns 2 TP. Repeated it on a Goblin Pathfinder, 7TP return, ~700 damage.



Striking Flourish adds 1 fTP like any other DA, returns 1 TP. Also, doing tests with Utsusemi-casting mobs is misleading. <_<;

If there is something useful about this JA, it's beyond my abilities to figure it out. It doesn't add damage to the second hit. It doesn't seem to copy the fTP of the first hit. It has a 2 minute recast. Whatever the Enhancement on the AF3+2 body is, it had better be something really freaking serious (+6 attacks, for instance).

[Esvedium]

Did some Striking Flourish testing last night and this morning:

That's depressing. gogoworthlessabilities? =/

[Byrthnoth]

That's depressing. gogoworthlessabilities? =/

Yeah... works out like this:
* Odds of DA with Saber Dance, Aurore Doublet, and Brutal - 27% per hand
* Odds of DA with Striking Flourish - 100% on the main hand, 2 Finishing Moves consumed
It takes 1 step to get the 2 Finishing Moves that Striking Flourish consumes, so the whole thing takes 3 seconds to give +.73 Attacks on the round following Striking Flourish activation.

* 190 delay per hand, delay reduction cap (80%) = 76 delay per swing, or about 1.27 seconds per round.
* 1.27 swings per hand thanks to DA, so 2 swings per second on average.

So we sacrifice 3 seconds (and so 6 hits) to activate a JA and gain 0.73 hits. It's like a circle JA, but worse.

[Esvedium]

Yeah... works out like this:
* Odds of DA with Saber Dance, Aurore Doublet, and Brutal - 27% per hand
* Odds of DA with Striking Flourish - 100% on the main hand, 2 Finishing Moves consumed
It takes 1 step to get the 2 Finishing Moves that Striking Flourish consumes, so the whole thing takes 3 seconds to give +.73 Attacks on the round following Striking Flourish activation.

* 190 delay per hand, delay reduction cap (80%) = 76 delay per swing, or about 1.27 seconds per round.
* 1.27 swings per hand thanks to DA, so 2 swings per second on average.

So we sacrifice 3 seconds (and so 6 hits) to activate a JA and gain 0.73 hits. It's like a circle JA, but worse.

Law of diminishing returns. Gotcha. Guess I won't even macro it then.

[yunalaysca]

i think it's just SE trying to justify using Rudra's Storm on dnc

[Byrthnoth]

i think it's just SE trying to justify using Rudra's Storm on dnc

It would be, but I'm pretty sure it doesn't copy the fTP of the first hit. You'd be better equipped to test this with your Daka+1 than I am with Shark Bite though. Utsusemi-casting monsters don't work, so you'd basically have to eyeball the damage. This setup would be best:
1) 0% DA with capped fSTR, single-wielding a Daka +1/2
2) Find something that's very low level, but satisfies these conditions and can survive the first Rudra's hit.
3) unload spellcast, etc etc, get 100TP and 2 Finishing moves
4) Striking Flourish -> Rudra's Storm.
5) Check your TP return and damage, compare to the theoretical max damage with 9 and 5.5 fTP.

If you don't do twice as much damage with Striking Flourish -> Rudra's Storm as Rudra's Storm unstacked, then it doesn't mirror the fTP.

Again, I'm pretty convinced by the tests I did with Shark Bite, but this would be a much simpler test to do because it's a single-hit WS.


Amusing part is, Sneak Attack offers a much larger damage bonus than this would even if it did happen to mirror Rudra's fTP.

[Latheille]

Just some more testing with Bounty Shot:

*Consumes an arrow regardless of ability landing on the mob or not
*Bounty Shot will only show up on the log when it has no effect on the mob. The few times I got it to land (and they were not many...), I did not receive anything on the log. So I believe you will not receive anything in the log regarding to Bounty Shot if it has an effect on the target mob.
*Pairing it up with a THF did not seem to increase or decrease the likelihood of it having an effect but I would like do more tests on this.

I'll be trying to make a gear set made up of pure AGI to see if maybe I can up the chances of Bounty Shot landing on mobs. I figure if any stats were to influence Bounty Shot, AGI had the best chance of working. I'll be doing more testing on Bounty Shot over the next few days since what I've learned is very bare bones right now. It's possible you'll have a much greater success rate against normal mobs than against NMs.

[Mercilessturtle]

Didn't see this mentioned anywhere, pup got martial arts 4 somewhere between 86 and 90, I assume at 90.

[Toth]

Didn't see this mentioned anywhere, pup got martial arts 4 somewhere between 86 and 90, I assume at 90.

Awesome news :D

I was wondering why I was suddenly attacking faster.