Optimal Dual Wield(Paging Armando to the RDM forums)

[Armando]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

For all intents and purposes consider DEX only as a form of Accuracy unless you intend to make a crit build, which is another subject entirely. That entails breaking certain DEX-AGI tiers so that the DEX you add actually has a meaningful effect. Otherwise its effect on crit rate is practically non-existent. And that's even more situational than Mensur Epee vs Enhancing Sword since it becomes an all-or-nothing deal that depends entirely on what you're fighting and its level.

Can't comment on Sea Gorgets, can't use them yet.

As for..."I guess my only concern is as a RDM melee stats come at a premium, and I need to figure out how much STR is worth giving up for the sake of Acc/Atk."...well, it's pretty easy to figure it out. If you're soloing then you can easily get into the fSTR ranges in which every 4 STR is guaranteed to mean +1 fSTR (as opposed to needing 5 or 6 STR points.) 10 Acc means 5% more hit rate, and losing 1 fSTR from Warwolf Belt is definetely less than a 5% loss in total DMG regardless of your sword choice. Attack vs STR takes a bit more work to figure out but in general when your PDIF isn't high, Attack is better.Not gonna crunch the numbers right now but for T mobs Tarasque Mitts will likely help more.

[Callisto]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

Many moons and Old Styles ago I was actually intelligent, but it seems that this has in time worn off, because I've lost myself trying to do some math on this. For example:

WSC = floor( floor((A x A%) + (B x B%)) x α )

What is the 'floor' they speak of? I've seen it as 'integer' too, I'm not sure what I'm missing but I don't know what that means LoL!

Anyways, I'm trying to calculate what my current weaponskill damage is like vs. Greater Colibri, and then figure it if I dropped 11 STR(in favor of 10 atk and 10 acc).

I understand the WSD = (D + fSTR + WSC) * ftP) * PDIF equation, but I seem to be missing a couple of pieces when figuring it out.

Using a level 81 Greater Colibri as my target(I'm going to go ahead and use Nagamaki's #'s here, because he's most likely smarter than I am.

DEF 322
EVA 336
VIT 67

I'll assume that my STR is still 90, and my Atk as RDM/DRK is 381(8 + 252(sword skill) +45(str/2) +44(gear) + 32(atk bonus I+II). This Atk seems high to me actually, I'm at work so I can't check my actual atk, but I don't think it's that high as RDM/DRK...am I doing something wrong on the equation here?

If I understand correctly the things I can plug in so far are:

D = 40(Enhancing Sword)
fSTR = Lost here, I figured out my dSTR would be 23 and my weapon rank would be a 4, but can't quite figure out how to turn these into fSTR.
WSC = floor?((90 * .3) * .83 = floor(22) (22.41 rounded down--wtf is floor?)
fTP = 1.0 OMFG SOMETHING EASY--THANK YOU!!
PDIF = .916min 1.056max

This is all I could figure out on my own, and I probably made a mess of it. If you could help me figure out what my current WSD is, and what kind of effect losing 8-11 STR would have, I'd appreciate it.

[Armando]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

Floor basically means to drop all decimals (as you can see, it's an operation that's used quite a bit in all the equations.) Basically, both modifiers get added with their decimals still intact, then the sum gets its decimals dropped. Then you multiply that by alpha/level correction factor/whatever you want to call it and drop the decimals from that product.

By the way don't add Attack Bonuses. When they say Attack Bonus II is +22 Attack, that's assuming Attack Bonus I is completely overwritten. If you wanted to think of them separately then it'd be +10 from AtkBonus1 and another +12 for AtkBonus2. So your Attack would be 371. I believe it's correct.

For fSTR, you'd go to VZX's FFXI Doc and look at the fSTR2 table. Look up the fSTR2 value for your STR difference and weapon Rank, and halve that value (fSTR = fSTR2/2, fSTR2 is used for Ranged Attacks.) In your case it'd be floor(13/2) = 6.

Don't forget to take crits into account if you're going to be looking at Vorpal Blade.

[Callisto]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

Ahh thanks so much for clearing those up! I've been wondering how the math end of the game worked for some time now, you've been a huge help with my understanding of this stuff.

I've been a melee RDM pretty much forever, and I got flak for it for a long time until I decided to be good enough at it to shut people up. Especially against Greater Colibri, when if I'm not main healing RDM is flat out boring to play if I don't melee. I spent ages capping my Enhancing Magic and combat skills, got ridiculously lucky to get an ESword for a price that would make many RDMs seek out mob violence against me if they found out, and now I'm trying to focus on optimizing my gear so that I can not only hit the VT mobs, but do some respectable damage when I do. The only problem was I had no idea what kind of STR/Atk/Acc ratio I should have been looking for. You've saved me probably 800k+ in gil and hours of experimenting with gear by explaining this stuff to me, I really appreciate it.

[Armando]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

It's no problem.

By the way, you should run some numbers for Joyeuse vs Enhancing Sword single-wielded. As good as Enhancing Sword is, I don't think it can outweigh Joyeuse's 50% extra hits and 50% extra WS's. E-sword won compared to Mensur Epee when dual wielding because the extra Acc/Attack/Enspell damage was being added to both hands, whereas Mensur Epee's higher DMG/sec only applied to itself (i.e. half of your hits) not necessarily because Enhancing Sword on its own is all-around better than Mensur Epee. Plus, Joyeuse itself is stronger overall than Mensur when damage types aren't an issue.

[WishMaster3K]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

I use Joyeuse against Undead and I cast Banish as /PLD because I'm just soul-blindingly awesome.

But I'm biased, because this is the only sword that can even come close to having me feasibly keep up with melee. When my friend is on MNK he can get TP almost Twice as fast as me when I'm in my current setup.. And I barely keep up to his NIN...

I'd say that ultimately, Justice Sword + Joyeuse is the best combination. But I'm a TP whore.

[Callisto]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

Yeah I still toy with the two of them from time to time to compare. I play on 360 so I can't exactly parse things out, it's pretty much just done by eyeballing damage and roughly monitoring killspeed. I can't say the exact #'s, but honestly I can't find a noticeable difference in effectiveness between the two most of the time. It really does seem to even out due to the higher Enspell/WS damage/hit rate of the ESword, despite the slower swinging.

I currently use Joyeuse for farming Easy Prey or less, when I'm likely to be /THF and not have Vorpal, and generally have a good accuracy boost over the mob. For T+ I have yet to see the Joyeuse effect make up for the additional damage of the ESword effects.

Then there's also the aesthetic appeal of wearing a sword that used to sell for 16m when people check you. At any rate, it's certainly nice to have the option to choose between the two of them.

My largest issue before was a large misunderstanding of A) how weaponskill damage worked and B) what mob stats were like. I greatly overestimated what mob Def and Vit were like on xp-level enemies. I had assumed that Vorpal was doing low damage against Greater Colibri because I wasn't hitting hard enough, when in reality I was just likely missing 2-3 swings of Vorpal on a consistent basis(which would also explain the huge difference that Souleater made, thanks to the hidden acc boost from it.)

Running the #'s on my acc real quick, I was at about 304, a solid 30 under a Greater Colibri. If I make a couple of changes, toss on a life belt and even that ridiculously ugly OHat, I'm more in the 320 zone. I'd lose 5 STR from the belt, but with sole I'd be at pretty much the same, and I intend to upgrade to Chiv Chain and Amemet +1 soon, so it won't be too bad. My acc in the new set would come out to something closer to 370 with food, not quite the 80+% range but I'm sure my Vorpals would be noticeably more consistent.

I'm going to buy the life belt tonight and the tarasque mitts as soon as one comes on sale, then I can start testing things out. Then I just have to con some friends into coming out to Colibri for me so that I can test my new toys(I also want to test my Expunger/Sleep Arrow pulling on the Colibri while I'm there).

[Ryoii/Nonomii]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

I'm going off topic here, but does anyone know how to factor blind into accuracy calculations? Mostly I'm curious how to calculate my evasion rate against a set enemy when the RDM casts blind.

I'm fairly familiar with the formulas for accuracy and WS damage since I've done calculations for my ranged accuracy and Sidewinder damage. Some day I'd like to be able to trio Abraxas with a SAM/RNG, THF/NIN, and RDM/WHM party of friends. Mostly we would be using shadows and Seigan + Third Eye to avoid taking too much damage.

Thank you in advance ^_^

Ryoii - 73 SAM/RNG
Nonomii - 73 RDM/WHM

[Callisto]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

I have been unable to find a % decrease that Blind does(I'm assuming it's a % decrease with a certain cap of acc down) unfortunately. It's kind of one of those things you just cast because it helps, regardless of how much it does. Just an absolutely blind(no pun intended) guess, compared to other spells, I would assume that the effect varies depending on your INT - Mob's INT, with a cap that I'd guess be around 25-30%. Again this is completely baseless speculation drawing from how other enfeebles work, so I make no claim about being right.

[Ryoii/Nonomii]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

Does Blind really have a noticable effect on a RDM while soloing VT enemies?

When I use the player equations for hit-rate and reverse them for evasion-rate, it looks rather hopeless for a RDM to evade at greater than the minimum against VT enemies.

Example: Equipmentless Lv75 Hume RDM/WHM and a Generic Lv78 WAR/WAR Enemy

Assuming enemy ACC is similar to EVA
Enemy Accuracy: f(Lv, Job Accuracy Rank) + DEX/2 + Job Traits
Job Accuracy Rank (WAR, A?) = 147 + [(78-50)*4.9] = 284.2 -> 284

Assuming enemy DEX is similar to VIT
Enemy Dexterity: Race Ranking + Main Job Ranking + Support Job Ranking * Support Job Revision
Race Ranking (B?) = 4 + (78-1) * 0.45 = 36.45 -> 36
Main Job Ranking (WAR, B?) = 4 + (78-1) * 0.45 = 36.45 -> 36
Support Job Ranking (WAR, B?) = 4 + (78-1) * 0.45 = 36.45 -> 36

Converting Hit-Rate to Evasion-Rate
Hit-rate = 75% + [Defender Evasion - Attacker Accuracy + (Defender Level - Attacker Level) * 4] * 0.5%
Evasion-rate = 100% - Hit-Rate
Evasion-rate = 100% - 75% - [Defender Evasion - Attacker Accuracy + (Defender Level - Attacker Level) * 4] * 0.5%
Evasion-rate = 25% - [Defender Evasion - Attacker Accuracy + (Defender Level - Attacker Level) * 4] * 0.5%

Final Evasion-Rate Calculation
Equipmentless Lv75 Hume RDM/WHM Evasion = 210 (Base D Ranked RDM Evasion Skill)
Equipmentless Lv75 Hume RDM/WHM Agility = 59 (Based on FFXI Stat Predictor v2.0)
Generic Lv78 WAR/WAR Enemy Accuracy = 284
Generic Lv78 WAR/WAR Enemy Dexterity = 108

Evasion-rate = 25% - [f(210 + 59 * 0.5) - f(284 + 108 * 0.5) + ((75 - 78) * 4)] * 0.5%
Evasion-rate = 25% - [239 - 338 - 12] * 0.5%
Evasion-rate = 25% - 55%
Evasion-rate = -30%

Let's assume there is a lower limit of 5% on evasion similar to accuracy having an upper limit of 95%. Our RDM would need more than 70 more evasion to get over the lower limit. Do we really think the Blind spell can provide this much of a help? Are enemies much less accurate than I estimate? Does Blind just flat out increase evasion after the lower limit is applied?

Anyway these are just my thoughts. I look forward to all comments, in-game experiences, and criticisms.

Also if anyone has suggestions on how to test some of this theory, I would be willing to work on it with my accounts. I have one account with a Lv73 RDM/WHM and one with a Lv41 NIN/WAR with a bit of evasion gear. Would Lv45 Guardian Crawlers which are spawned with Rolanberry be a good enemy for some experiements? Is there anything wrong with using a NM?

[Callisto]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

Does Blind really have a noticable effect on a RDM while soloing VT enemies?

No parser or math needed on this one lol, the answer is No. As a RDM I will evade just as much like a brick against a VT mob regardless of Blind's effect, I can say that from practical experience, RDM's Eva skill is just too low to make up for it. Even as /NIN with a bit more eva added, I still almost never get 'missed' by a VT mob, it's just possible to tank them b/c of FC5 making my Utsu timer so low. However, as far as in game experience, I can say that NINs and THFs who have tanked things for me have said that Blind/Kura make a noticeable difference for their eva rates. This is where I came up with the 25%~ idea, because that would be enough of a swing to throw them into a noticealbly different hit rate tier.

As far as the actual math, there's honestly nothing I can figure out without a parser and a rough idea of how strong Blind really is. I would think you could get a rough idea of it by doing controlled tests in PvP with a parser, but I'm a console player so that's beyond my means.

[Armando]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

As a PLD, I only evade 5% on high T to very low VT enemies (this obviously excludes Flash.) Including Blind bumped it up to 7%. On anything higher than the lowest VTs it probably won't have any effect at all, and it certainly won't after I equip my DD gear.

Also, don't apply our formulas to mobs (well, don't use the same parameters.) They damage us definetely and I'd eat my fictional hat to wager that their hit rate formulas are different, or have different limits.

Blind is probably something small like -10 Accuracy. Possibly up to -20. But hey, it costs an absurdly small amount of MP to use.

[Vevau]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

Effect kicks in.

[...]

Paralyze proced once on that mob = You suck (You cast paralyze and only one action was paralyzed)

Thank you for clarifying that, I was almost certain :-)

But if Paralyze procs once on a mob that hits for 30+ it's good, 6mp saves 8mp (Cure I or Pollen). Now if thats a high-level (my highest is rdm52) exp mob I presume it would be hitting for 90+ so thats more like 24mp saved as a Cure II that someone didn't need to cast, so as a Red Mage, its great as it saves MP which is what we are really about

Ergo, Paralyze is a fantastic effect, especially in the form of Ice Spikes, every time it procs it saves HP, Shadows or Stoneskin, and as Ice Spikes does a little damage.

PS If this makes sense blame drink, it wasn't meant to

[Aeni]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

As a PLD, I only evade 5% on high T to very low VT enemies (this obviously excludes Flash.) Including Blind bumped it up to 7%. On anything higher than the lowest VTs it probably won't have any effect at all, and it certainly won't after I equip my DD gear.

Also, don't apply our formulas to mobs (well, don't use the same parameters.) They damage us definetely and I'd eat my fictional hat to wager that their hit rate formulas are different, or have different limits.

Blind is probably something small like -10 Accuracy. Possibly up to -20. But hey, it costs an absurdly small amount of MP to use.

The only thing about it is if it's duration is short, then you'll be spending a load of time keeping it up and thus prolong the fight (While your other buffs tend to also wear and you'll have to spend more time reapplying them - vicious cycle of extending those fights)

I, however, agree that it's a trivial cost (mp wise) to reapply Blind and to keep it up (casting time) ... this obviously gets better mileage when you have Utsusemi (Ni) to lessen the effect of having to spend so much time reapplying buffs/debuffs as well as, ironically, in very prolonged fights (reason being, %-wise, the return becomes more appreciative over a greater length of time where the damage nullified accumulates and becomes mp-conserved)



...



I have a question which may have already been addressed before, but I need to ask this anyway. Before I get to level 74 (When I have access to Utsusemi Ni - NIN 37) ... which is a better option for soloing on RDM?

RDM/NIN Dual Wielding (Not sure what combination of weapons)
-or-
RDM/XXX Sword (or En-spelled Dagger) + Shield?


(note: only subjob available that is leveled up to my 67 RDM are WHM, BLM and NIN ... WAR is at 19 and SAM at 22)

[WishMaster3K]

Re: Optimal Dual Wield(Paging Armando to the RDM forums)

RDM/NIN Single-Wielding Joyeuse.

Edit:
Ok, there's more.

To be honest, if I'm fighting something like the Sheep in the Mire, or for that matter, ANYTHING that is EM or lower (except for Demons... I don't eff with Demons unless I have /NIN), I have enough defensive abilities to be able to fight them on just Phalanx alone.

Genbu Shield, Jelly Ring and Cheviot Cape, coupled with Phalanx and Protect IV and if I feel like it, Bream Sushi or Tav Tacos.

That's my standard Def setup, and It works where I can fight DCs and EPs without Stoneskin, and Just Cure III when my HP gets into the Yellow. I find that Stoneskin takes too long to cast for my own personal tastes.


But for EMs or lower, I /PLD or /WAR, and either way works fine and gives me Offensive and Defensive traits in their own right, so it's really a matter of personal play style and tastes.

I know people who'll swear by /BLM when they solo, but I don't like having to sleep a mob when I want to reapply my entire line of buffs and/or Stoneskin (I learned that from Grendal, and that's how I soloed Crabs when I was in my 60s.)