I completely looked past the Sword Skill, lolme. It'd be good for when he needs accuracy, otherwise Centaurus would win.
Agreed, but the good ole Suppa does have it's uses for the grand job of PLD, as situationally rare as they are.
Also for twilight belt, it seems all single handed and dual wielding jobs are on it, but no PLD, unless GS is considered too powerful to be given 7% haste by SE >.>
I could see how Suppa could be situationally useful. Just had never considered it until now since I'm not dual-wielding obviously. That'll work out well since I was planning on re-questing it since I leveled Ninja anyways.
Gotcha. I just wasn't sure about inside Abyssea since RR and buffs pushes your DEX so high anyways. Then again the weaponskill itself is a pure DEX modifier too. Thanks.Originally Posted by Nightfyre
No pictures, so I know I'm already doing it wrong, but I posted this in the What's Better thread and got no responses before it hit the second page. I'd really love some discussion and analysis from the local mathologists.
Typical thinking about X-hits and two-handed jobs values reducing the number of hits above almost all else, short of perhaps losing too much haste gear. At high levels of haste, the 2-second delay after JA/WS becomes a greater factor, which lessens the advantage of the lower hit builds. As I've inherited a WHM and a BRD from a couple of RL friends who have left the game and I'll be essentially guaranteed to have Haste and double Marches, and I'll completing the lv85 Masamune shortly after the servers come up, I've grown more curious about this scenario.The generally accepted TP build for a 5-hit Masamune SAM looks like:
Masa/Rose/.../W.Tath
Unkai+2/Unkai/Cent/Brutal
Unkai+2/Dusk+1/Rajas/Hoard
Tactical/Bullwhip/Unkai+2/Ace's
which gives 24% visible haste in gear, 21 TP/hit, and with Haste/Marches/Hasso is ~131 delay, or a swing every 2.2 seconds, yielding a full WS-to-100TP cycle of 10.8 seconds (2+2.2*4).
I'm not really sure what an ideal six-hit build would be, but I think it would look something like this, assuming identical WS builds (other than grip):
Masa/Sword/.../W.Tath
Zelus/Unkai/Cent/Brutal
Juogi+1/Dusk+1/Rajas/Strigoi
Atheling/Bullwhip/Unkai+2/Ace's
A six-hit TP build for Masamune SAM using a Sword Strap reduces delay to ~424, and allows a SAM to cap gear haste and keep quite a few of the better DD-related pieces.
Haste/Marches/Hasso yields ~121 delay, or a swing every 2 seconds, and the WS-to-100TP cycle is 12 seconds (2+2*5).
The 5-hit build gains 7 skill, and 1.2 seconds between WS, or an 11% gain in WS frequency.
The 6-hit build gains 10 STR, 3 DEX, 21 ATT, 8% Double Attack, 4% crit rate, ~1.38% Haste, and an extra melee strike along the way.
I'm not a strong mathologist, so I've run the sets through Motenten's DPS Calculator (default enemy of lv97, 420 def, 407 eva), and the 6-hit build has come out ahead, both with and without atma.
Thoughts? Comments? Discussion?
Edit: See post 969 for further discussion, specifically about the influence of atma.
And here I was, about to make stupid comments about the lack of Twilight Belt and Epona's Ring. What the fuck @ the job selections.
I may be remembering incorrectly, but isn't JA/WS delay initially 2s?
It's 2 seconds. 1 second of total inactivity, 1 second where you can use another JA/WS/whatever.
I personally only have goading and I use it on PLD (25% haste).
For TP ring inside abyssea I suggest Hoard ring.
For WS definitely Dex ring not str. In abyssea Perle Gloves might also be better even then n hands+1. Lithe Boots and Aurum are also better in abyssea.
Outside abyssea idk why you use dex earring when you don't in abyssea. You have more then enough dex in other slot to cap dDEX. Idk why you would want to switch to cerb +1 mantle too. 3%da and 3 att is a little better or at least equal to 1fSTR so its inventory space -1.
I'd like to see some commentary as well. I haven't run the numbers on SAM, but every time I run them on WAR I get better results with a 6-hit, also using Sword Strap. Granted I'm just running them on Motenten's spreadsheet (the extent of my interest in WAR is giving gear advice to Aza), but they're virtually identical (<0.5% difference) before considering set bonus procs (6-hit has more AF3 and more frequent melee hits) and Ukon aftermath (which 6-hit will again benefit more from).
Just some loose math, but just disregarding all the extra DA/attack gear you get from not needing as much STP, and the -3% delay on sword strap, you would need your 6th melee hit to do ~1/5th the damage of Ukko's to compensate and make it equal DoT. Checking a parse on 7 Orthrus, I averaged 3700 on Ukko's with a Maschu+2, so if we assumed a 90 Ukon, could bump it up to 4000. This means your extra melee hit would need to average 800. Going back to the parse, my crits were ~500 and my melee swings were ~150, with a 60% crit rate. This means Average swing damage should be about 360. With 40% (I think someone said aftermath on 90 weapons was 40/50/60?) proc rate on ODD, that's ~505 average damage. That's still 300 damage short.
Granted, that's assuming a lot of things, and oversimplifying even more things, almost all against the 6-hit. It may be comparable once you factor all that in, but I don't think it would quite beat out a 5-hit. Ukko's just plain does too much damage.
That's the argument that 6/5 = 120% increased WS frequency, therefore you need to do 20% of the WS damage in the extra round, right? That doesn't really address the situation I described above; the two second JA/WS delay skews the argument quite a bit the closer one gets to capped delay reduction.
In the example above, it was closer to 11%, which, with Fudo's lower average damage (compared to Ukko's) and the considerable improvement in TP gear in the six-hit set, seems more easily obtained.
I have no parses of my own (19 lanterns from Lv85 Masamune when the servers went down), so I'm pulling numbers from Motenten's spread sheet:
With an average Fudo damage of ~2700, it would require 297 damage for the extra hit to match total DPS. (2700 * .11 = 297). Average melee strike per the sheet is 376 without Bergressor, and 411 with, which seems to win by a pretty fair margin. I don't think the spread sheets take aftermath into account, either. Perhaps Motenten can clarify, if he reads this.
Edit: I realized I wasn't accounting for the average number of rounds to 100% TP, which is significantly different in Abyssea due to VV/Apocalypse. Here's what I'm getting now, with the new assumption for Abyssea specifically:
The following comparisons assume Abyssea, with full cruor buffs and VV/RR/Apoc atmas.
The additional Double/Triple Attack rate influences the situation heavily, reducing the difference between the two seemingly to the point where the WS frequency difference is almost negliglible.Spoiler: show
I'd love to have some input by anyone who can point out where I've gone off the rails, if that's the case. I'd love to root out any fallacious assumptions if they are there.
The arguement is slightly different on SAM since you can gain extra haste between the two builds, which changes the delay between WSes. WAR caps haste either way, you're just losing a few points of DA gear, and a fair bit of attack.
Also 131 * 4 = 524 delay to WS, 524/60 = 8.77 seconds; 121 * 5 = 605 delay, 605/60 = 10.08s. Add in 2s delay for WS to both and you're getting 10.77 and 12.07 per WS phase. That's a 12% gain in WS rate, not 3%.
I don't think the spreadsheet accurately quantifies the bonus gained from a 5hit, or at all, especially 5hit builds that depend on VV tics. Nor does it factor ODD aftermath. So, 5hit should have a sizable lead on any type of 6hit build.
As for the sam argument, i can't really say much for inside abyssea, because i could never think of a realistic time it'd be needed. Although, outside, it's still possible to cap gear haste and 5hit:
Masa/Rose/.../W.Tath
Zelus/Unkai/Unkai/Brutal
Unkai+2/Dusk+1/Rajas/Hoard
Tactical/Bullwhip/Unkai+2/Ace's
+ Carbonara
This would be realistic on lower defense mobs, when you're getting march/haste. If you're fighting higher defense mobs, you can swap back to the traditional build, and use RCB.
I'm pretty sure it does take into account the effect of VV's Regain on 5-hit builds. You can see this pretty easily by putting in a set that misses a 5-hit by a couple of Store TP, then toggling VV on/off and watching the average round number drop/rise by a full round.
The sole thing that a five-hit build has going for it is increased WS frequency. When the actual frequency is diminished significantly due low final weapon delay and the influence of the 2 second JA/WS delay, it doesn't seem like a stretch to think that it may not remain the best option, like it clearly is in lower haste situations.
Necro-edit - I missed this back in March:
That would be the case if one had no double/triple attack at all, but the situation described in the post assumes /WAR, Brutal, VV, and AoA; the increased WS frequency will be lower than 12%.
I was referring to the war spreadsheet: Although the # of rounds do change while toggling VV on/off - it's apparent it's not taking x-hit into consideration. Going from VV to GH, using a build that's dependent on 2 tics from VV, is considered better, which clearly would not be the case, as i'm sure 20% crit from GH - although helping melee damage, but hardly doing anything for WS damage, won't beat out the increased WS frequency from gaining 5hit.
The sam spreadsheet does seem to factor in 5hit, but i'm not familiar enough with it yet, so i'll leave it at that.
Edit: After looking at the sam version; i'm not completely convinced it's noticing the changes from going from a 6hit to a 5hit.
Using the default mob settings, turning atmas off, and using the same build i mentioned, which is definitely a true 5hit, i'm getting an increase in overall DPS going from carbonara - needed for 5hit - to yellow curry. So, basically, it's saying 1 str 20 attack, using the exact same build, but 6hit, is better than the 5hit with carbonara.