Thief, FFXI, and other unimportant matters

I've been working on finishing the previous entry, but Excel's being fussy with me.  If anyone's an excel master, I could use your help.  I can't seem to get the if functions I want working.  Other than that, there is still a lot to do to perfect the spreadsheet.  But once it's done I think I can use it as a base for other analyses.  Once it's complete, the most difficult thing might be having to weed down the results.  I want to compare all of the situational pieces for sa/ta gear in all of the possible situations, if necessary.  However, this would mean at least 250 results to input and look over for TA and almost 1,000 for SA.  If it wasn't for being busy busy at work, I'd have this done by now.

Just throwing this in as a placeholder for now, and I'll be editing it soon.

Recently good fortune has shined on me, and among a few other things I've obtained both a Speed Belt and some very much awaited N.Legs aka Hecatomb Subligar +1 (this piece should (sorta) never be NQ'ed!!).  It's quite obvious that N.Legs are great for EV/DE/SB/MS and Sa+Mercy, but what of Ta+Mercy?  I imagine they likewise dominate Enkidu's, but I'm as cautious as I am frugal, so I'll be double-checking the math.

It's also obvoius that speed belt is at least going to be used for TP, ni, and ichi (part 2).  But what both S.Belt and N.Legs have in common is that they have both positive and negativen facets associated with their use for solo SA/TA's.  Awhile ago--I'm going to guess 6-9 months ago, there was a very good thread on the KI thief forums regarding the merits of haste/DW/d.a. and the demerits of slow for SA/TA sets.  Even in 2009, there isn't even an inkling of self-inspired thought on this subject for most thieves.  They simply stack as much dex/agi (and sometimes attack) into their set (if they have one) and call it a day.  The major problem lies in haste stacking.  Frequently, even thieves in burn situations will find themselvs with one or more march, haste spell, and high levels of haste gear accessible.  As haste is an exponential function, the value of dmg.spiking stats such as dex/agi/atk/str begins to diminish as more available haste is stacked.  Furthermore, stacking slow has always given detrimental effects.  This is no exception for SA/TA.  However, it should also be noted that Slow =/= -Haste.  It's actually a seperate part of the delay calculation and is a decreasingly marginal function.  That is, the more slow you stack, the less of a problem supplemental units will cause.

Given that thieves find themselves often enough in high haste as well as low haste situations, and also in between, coupled with the wide variety of gear choices which have haste on them, dmg stats coupled with slow, and many other gear choices in between, it can be difficult to tell for sure which pieces to use.  Shit being situation as well, the perfect sa/ta set for a zerg fight (which really wouldnt matter manyway) will not be the ideal set for when you're soloing, nor will it be the best set in a march/minuet exp party.  So I'm going to try and figure out what the optimal sets are for these various situations.  Excel might help.

***

I'm not a slouch.  I remember this being particularly difficult to calculate.  Trying to figure out the optimal set for SA/TA is difficult, and the formula looks somewhat like this; Total Dmg =

Simple Form
(SA attack round) + [(haste/slow mod) * (1 TP-set attack round)]

More Detailed
[(SAmain*1)*100%crit + (SAoffhand*1.24 + SAmain*0.24)*x%crit] + [(Haste/Slow mod)*(TPmain*1.24 + TPoffhand*1.24)*(x%Crit*x%Acc)]

Even More Detailed
[(b.Dmg + fStrA + Dex)*(pDifA +1) + ((b.Dmg2 + fStrA)*1.24 + (b.Dmg + fStrA)*0.24)*(%pDifA + %pDifA+1)] +

[(1 / (1 - haste%)) * (1 + slow%)] * 
[{(b.Dmg + fStrB) + (b.Dmg2 + fStrB)}*1.24*(%pDifB + %pDifB+1)*acc%]


That is, in order to figure this out we need to find out how much the SA does plus the possibility of another melee round, modified by how strong that round is.  That is, /nin, the round of attacks you do when your SA strikes is modified by the equipment you have on at the time of the SA.  This is indicated in the formula by "A."  Then the next successive strike is represented by "B" in the formula, and we want to know how much dmg that will do.  We also, more importantly, want to know the probability of another attack going off.  The probability directly corresponds to the inverse of the delay you have on at the time of an SA.

In simpler terms, your SA does a lot of dmg, but the haste/slow you have on when you strike with SA determines the delay before your next melee swing, and this is just as important to the total damage of an "SA event" (even though I hate the way that that sounds) as the actual SA itself.

Unfortuantely, quite a lot of factors come into play here.  We need to know pDif (median) versus the mob, crit rate, +fStr, and accuracy.  And we need to know these values for both the SA set and the TP set.  Otherwise we can't mathematically calculate how much damage is occuring (on average) during an "SA event."  Also important is the total haste% and slow% during an SA, as well as the multi-proc chance (ie brutal earring, triple trait, triple attack merits, etc) for both the SA round and the TP round.  Knowing and calculating these values will bring us extremely close to finding out just how good an SA set is to a particular individual.

Unfortunately as well, this doesn't achieve perfection.  This formula doesn't account for the presence or probability of non-haste-related buffs (minuet, rogue's roll, food, and others), nor the probability of debuffs.  It also assumes we're only ever fighting one mob of a certain mob level.  That is, for this mathematical examination as others, we'll be using a level 82 G.Colibri.  This doesn't account for the level 81 G.Colibri, the probability of other pulls in the area (Wivres and Mamools).  If it could accurately predict such a thing, we'd have a good analysis for SA sets at "mid bird camp."  Still, it does not serve as a great sample for other mobs in the game.  It's only good for (fairly) accurately determining SA (and TA) set usefulness on level 82 G.Colibri.

If anyone read that, I apologize, for I know a lot of it may not come off being very clear.

means "giraffe" in Japanese.  I was surprised to find this out today.

  =  ?

epac

Edit: Also found out that the Japanese word for "dentist" and "loser" are one and the same.  No joke; the word is "heisha."

This is just an effort to tackle some things that I wanted to talk about a few months ago, so I'll start with an easy one.  It was back in August of last when einherjar had finally gotten beefed down, that I was looking for some outside-the-box strategies for my LS's runs.  Something that I had seen overlooked again and again was defense down stacking.  Defense down on mobs is great.  But stacking it is even better, and I don't think that many people realize why.  It'd be best if I explain by comparing it to some other buffs/debuff/property functions.  Let's look at.

Haste
-Enmity
SubtleBlow
DefenseDown
etc.

vs.

Attack
+Enmity
DoubleAttack
etc.

Defense down is a cumulative-growth function
The buffs/debuffs in the first list operate via a function which subtracts from a value x to give us a value y.  That is to say with haste, it subtracts from a value of delay(x) to give us a value of attack speed(y).  -Enmity subtracts from a TE(x) (total enmity; CE + VE) value to give us a value of "decreased hate pulling"(y).  And so on.  The buffs/debuffs on the second list are ones which add a value x and that's it.  +Enmity adds from TE(x) to give us a new TE, for example.  We could also think of the functions int he second list which give us values via subtracting values, but it would not be useful for our purposes as players so we might as well ignore that possibility (for example, we could think of double attack increases as a function which decreases how slow our rate of attacks are).

The above may have sounded a bit confusing.  If so that's my bad.  The important thing to remember is that: 
1. The things in the first list, being effects stemming from negative functions, have a cumulative rate of growth (increasing marginal rate of growth) as you add more and more of the stat, and are thus exponential functions.
2. The things in the second list, being effects stemming from positive functions, have a decreasing marginal rate of growth as you add more and more of the stat, and are thus, uhh, let's call them decreasing marginal functions.

Functions with decreasing marginal gains
For example, if you're a Mnk/Nin with no brutal earring, and you add brutal (+5% double attack), you gain a 5% boost in swing volume.  Previously they had 100/100 (1) hits in terms of multi-procs, and now they have 105/100 (1.05).  Divide the new number by the old number, and you get 1.05/1 or a 5% boost.

Meanwhile, if you were to add a brutal earring to a War with 5 d.a. merits, you'd get a ~4.35% boost in swing volume.  Previously they had 115/100 (1.15) hits in terms of multi procs (5 from merits, 10 from native traits), and now they have 120/100 (1.2).  Divide the new number by the old, and you get 1.2/1.15, or a ~4.35% boost.

Functions with increasing marginal gains
It's been well known for maybe 2 years now, I think, that stacking haste can lead to very good results.  Commonly we've applied this to zerg fights, stacking as much as possible to make normally difficult fights a complete joke.  So lets take haste for example.  Let's say you have 0 haste gear or effects active at a particular moment.  Then, you have the haste spell cast on you (15%).  This decreases your delay by 15%.  Your increased attack speed is 1/(new delay) divided by 1/(old delay) or, simpler, just (old delay)/(new delay).  In this case, that would be 100/(100 - 15), or ~17.6%.  So that 15% haste gives you a marginal 17.6% increased attack speed from 0->15% haste.  Now what happens if we added haste spell onto an already haste-stacked situation?  Let's say you had 25% haste in gear, 10% from a ja (hasso), 10% from another JA (capped haste samba), and 20% from 2 march songs.  Adding a 15% haste spell do that brings you from 65->80% haste, or the current haste cap.  Your increased marginal attack speed going from 65->80 (not from 0->80) would be (100 - 65)/(100 - 80), or (35 / 20), or 75% marginal increased attack speed.  That's right, that very same haste spell that gave +~17.6% attack speed in one situation can give +75% in another.

The point is that the other buffs in the first list are also exponential functions.  As you stack more and more haste, -enmity, subtle blow, or defense down, it has a cumulatively increasing (exponential) effect.  Which is pretty nice.


Types of Defense down
When I was planning out einherjar, I tried to think about what the maximum defense down value was that I could expect to get on mobs and bosses.  As with haste, defense down comes in a variety of flavors.  Haste comes in the 'flavors' of: equip haste, magic haste, and ja haste.  Defense down comes in flavors too, but they haven't really been named.  So far, here's what I've figured out.  Any effect which produces something in your chat log which reads "Additional effect: Defense Down" is of the same 'flavor' as all other effects which also do that.  That is, they don't stack together.  I'm guessing stronger effects overwrite lesser ones, but I really don't know.  If someone can prove this wrong, please let me know.  Other than that, the only other types seem to be Dia, Light Shot (only when it can enhance dia), and Box Step.  So:

Types
1. Additional Effect: Defense Down
2. Dia
3. Light Shot (enhancing dia only)
4. Box Step

Some LS's have used Blue Mages to do one of various spells which lower the defense of several mobs in their radius.  This is not a bad strategy.  I would encourage it for normal mobs, if you have a blue mage.  However, let's take a look at the strongest effect in this 'flavor'.

Type 1
The strongest of type 1 is Angon.  When fully merited angon lasts for 1:30.  It's recast is 3:00.  It gives a whopping -20% def down.  2 drgs is all that is necessary to keep full time angon on long duration fights (bosses).
Type 2
Next we have Dia 3.  Rdm is a complex job and is open ended in terms of the style that the player wants to take.  However, if a Rdm prides himself over being "ready for anything," this means he ought to have Dia 3.  I recommend two merits in it.  The longer duration helps w/ MP and micromanagement.  One merit is all that is necessary, but it is imperative for zergish fights that Dia 3 be on full time.  Dia 3's base duration is 30 seconds with one merit, and 30 seconds more with each additional merit.  It gives -15% defense down.
Type 3
A corsair's Light Shot adds a little bit of boost.  Corsairs are always great to have, and this is a little bit of extra utility that they can add.  Light shot will add 5% more defense down, but only if a dia effect is already on the mob.
Type 4
Dancer has been mocked by many, along with scholar, as a useless job at 75.  I'd have to disagree.  While I don't see either job being exceptional in a large amount of circumstances in endgame, I do see them both having niche uses.  Scholar is exceptional in low man dynamis since they can double as both a healer and a timed-nuker.  Dancer is exceptional for mobile zerging parties (little known concept) since they have access to haste samba, which stacks cumulatively/exponentially with other haste effects.  Box Step also stacks cumulatively, and that is an awesome thing.  The first box step drecreases defense down by 5%.  However, it can be stacked up to 5 times, and each additional step gives 2% each, totaling 13% def down.  The delay beween steps is 0:15, so it would take 1:00 after the first step to achieve this level of defense down.

The maximum possible Defense Down
-20% - Angon
-15% - Dia 3
-5% - Light Shot (on dia)
-13% - Box Step
-53% Defense Down is our grand total for maximum possible defense down.

Since Box Step takes a full minute to reach 13%, it isn't instantaneous.  Thus, the maximum defense down obtainable "instantaneously" is -45% (since the first box step gives 5% and not 13%).

-53% and -45% are large numbers.  In my opinion, defense down is stacking is underestimated and just purely not recognized as highly beneficial.  Most zergs that I've been apart of have a maximum of angon and dia 2 for -30%.  The increasing marginal gain in terms of attack/defense from adding dia 3, light shot, and box step 1 to that is 27.3%.  That is, if you just added those 3 effects, we'd marginally increase our dps by nearly a third.  If we change box step 1 in that figure to box step 5, that's a marginal increase of 48.9%.  We're doing about 50% more DPS.  This is huge.  In a sense, that's like taking an alliance with 15 DD's and adding 7.5 more DD's to the mix.  Of course, corsair's and dancers aren't terrific zerging DD's on their own, but the raw DPS increase could be enough to justify it.  If we said that the dancer and corsair did absolutely no damage (and didn't count samba or roll effects), that'd still be like going down to 13 DD's and increasing their dps by 50%, which sould be as if having 19.5 DD's.  For einherjar, an event that can cap at 36 members, it seemed quite obvious to me that if you were going to bring along a lot of members, having a dancer in there was a win-win situation.

Why care about defense down stacking
It's all about maximum DPS zerging and max DPS in general.  Here are a number of situations where I could see defense down stacking working nicely: 

  1.Swarm-of-mob style, larger (alliance+) events.  In dynamis and einherjar, for example, dia3, box step, and light shot on every single mob would make a huge difference.  With two drgs, most mobs should have defense down on them.  Going from 0% defense down, this effectively double melee DPS on mobs.
  2.Long zerg fights.  I can't see a cor, rdm, and dnc necessarily being practical w/ some of the easier, shorter, common zergs such as BahamutV2, Kirin, and KV.  I can however see it being very useful on einherjar bosses.  These bosses are unique in that they are something which is "zerged" which actually takes awhile to kill.  Normally we think of zerg fights as short, but in einherjar this is not so.  With just 2 drg, 1 rdm, 1 cor, and 1 dnc, we can have just about permanent -53% defense down on the boss, which is just huge.
  3.Absolute Virtue.  This may seem like a bit of a joke to some, but I'd been toying around with the idea of whether AV was still zergable from a math point.  When I say zergable, I think of "what is the maximum possible DPS we could hope to produce without souleater."  Now if we know AV's relative defense, we can actually say theoretically if he is zergable or not.  Most of the other variables are known (such as player DPS formulas and his HP.  We can rule out accuracy/evasion just based on the power of feint and 2hr madrigals).  I think that stacking defense down would be integral to such a fihgt.  However, it would require a bit of coordinative strategy.  What I mean is having the bards 2hr, dropping the bards, adding a drg, rdm, and thf.  Having them do their thing, and then adding a cor and a dnc.  I'd discuss this in more detail in an AV post . But there we have a possible -45%+ defense down on AV, which as we've seen earlier increases DPS quite a bit over a marginal angon (which is all that I think most successful souleater AV zergs had in them).


And that about does it, I think. 

And it's not mine :(



Feriel, a JP from Bahamut
BG Thread

A little about the RL

Those who know me in-game know that I haven't been on much for the past month.  The reason for this is that I have a new job, as of late March '09.  When I graduated from Uni almost a year ago, our US economy turned pretty shite.  I chose to procrastinate from serious life plans and was enjoying my time playing FF and chillaxin while I worked entry level jobs.  The first job of which was a waiting job.  I worked a bow-tie waiting staff job at a luxury dining hall in an old home for the well-to-do.  I really hated that job :(.  But everybody knows that waiting sucks horribly, so that is no surprise.

Then I worked at The Limited, a women's retail clothing store, as the freight guy.  I got to handle all of the hardware and receive, stock, and handle all of the clothing (yes, panties included).  And for those without inside knowledge, Limited mannequins come equipped with a certain type of gel-filled ass-pad underwear that we have to put onto them manually.  True story.

But while all of this was going on I was looking into getting an English teaching job in Japan.  This kind of thing (in Japan or overseas in general) is becoming a little bit of a trend for recent college grads, I think.  It's not only an opportunity to experience something new and test yourself, but also a great way to buy more time to postpone your life (don't we love to do that).

Although I can't talk about all of the more specific or private details about my job, I can still say a fair bit.  For example, that I'm employed by a company called Interac, and the board of education contracts my services through them.  The board of education in my case is that of Miyagi Prefecture.  I work and live in a little area near the ocean; a place called Tagajo city, which is about 200 miles NE of Tokyo, just north of Sendai.

Soooo.., first impressions
If I could sum up my impression of Japan so far in few words, what would I say, hmm..  It is very clean, and very safe.  Food is healthy and tasty, but they eat more carbs than one might think (noodles and rice).  Actually, sometimes it seems like noodles and rice is just about all they eat.  Public transportation is great, but the bullet train is expensive.  Prices are a bit high, but technology is often a step ahead of us...or to the side.  Think like, public toilet seats with seat warmers, but that start to play a tune when you sit on them.  The western writing system is everywhere, and western slogans and slangwords are common.  There are plenty of beautiful women.  People are healthy.  People are "overly polite" in formal or business matters, but in casual situations (which is much of the time), they behave very much like westerners.  People care about their external appearance and many young people seem to take fashion seriously.  Though when the Japanese are sick it is not uncommon to wear a surgical mask in public, which to me looks rather unfashionable.  The TV is full of garbage game shows, but I guess American TV isn't really any better.  Hip-hop is popular, both foreign and Japanese.  I still don't know what to make of Jpop.

As for my particular high school, it is not what I expected.  Firstly uniforms are not required.  I made two speeches to the student body and faculty in the gym on my first/second day, and I was surprised upon looking into the crowd at how "American" they looked to me.  Of course the school is Japanese and neither anglo nor multi-ethnic, but their behavior and dress made me feel like I was at home at an American HS.  A lot of the kids at my school really don't give a damn, or at least pretend not to.  They skip classes often and some try to sleep in class.  They spike, bleach, and color their hair (some totally frowed out, some with dreds).  They wear torn jeans, thug themselves out, and you also see the occasoinal trenchcoat.  Some girls wear lots of makeup and they seem to be very fond of high stockings, either black or with fancy patterns.  Not all students are like this, but a lot of the body is comprised as such.  And beneath such an exterior, many of them have also turned out to be very shy.  And most of them are really great kids and do try in class when they're tired of pretending that they really don't care (well, someof them aren't pretending).  I should say that is more true of my day students than my night students, though (I work roughly from ~11am to 7pm).

There are pretty much no Gai-jin (foreigners) around here, though in the rare event that I do come across one, there is a likely chance they also work for my company.  A lot of the gai-jin hang out with each other, and can become very close just because they're fellow gai-jin.  Some of us 'ALT (assistant language teacher) no eigo sensei (english teacher)' play as hard as we work.  It is common among the Japanese to book a spot at a bar or restaurant for a 'nomi hodae'(sp) (all-you-can-drink drinking party) for you and your friends or co-workers.  And we foreigners do the same (some of us; less me than some others).  And I have to say that I was surprised at how great of a time could genuinely be had at the 'karaoke box' (karaoke party room).

Many of us, especially new people like me, don't really have time for much play (like FFXI, for example), nor the money to do so (we don't get paid anything until the end of the first two months).  Adjusting and getting set up here is rough.  Learning some amount of both written and oral language is advisable, and that takes time.  The work is also not always easy.  For me it is very difficult, but that makes it very rewarding.  I have no previous teaching experience and know (knew) almost no Japanese upon arrival, but my high school expects me to make full lesson plans throughout the year with no real guidance (and the texts here are more like a condiment than an entree), and essentailly teach 45 minute to hour-and-a-half sessions often very much solo, using no Japanese preferably.  So I'm very much in control of what and how I teach at my school (was not expecting this).  If that sounds tough, I can assuredly affirm that it is, haha.

With what little play time I have..
So adapting to this job leaves me with little play time.  I am starting to log on FFXI for at least a short time almost every night though.  I find that I can afk between HNM windows and use the time to lesson plan and study Japanese.  In a week and a half of consistent (though not arduous) study, I can now read and write the 46 base Hiragana.  Oral communication is another ballgame though.


As for current FF goals, I have a few.  I would say that salvage gearing and just having a good time in my new HNM shell (as opposed to getting geared in it) are top priorities.  My most fundamental in-game goal currently, however, is actually upgrading Vajra.  Yes indeed; I've had my heart set on this for awhile.  I currently have about 54m in easily convertable cash.  So with current prices that is more than half way.  But gil isn't the problem.

Right now by "currently upgrading Vajra," I mean I "currently attend einherjar twice a week."  I have only about 30k ampoules.  This means that if I go as often as possible, I still have like 7-8 months of ampoules ahead of me.  I've just now gotten a regular einherjar shell, but they go at rather late hours so I might not always be able to make it.  Either way I am confident that by the time I have 100k ampoules, I can easily have set out to've fulfilled the other arduous requirements ahead of time.

Why would I want Vajra?  Just to upgrade it for completion's sake really, in the name of my Thief and Thief in general.  I can picture it having use in non-/drk zerging (and maybe /drk zerging above m.kris), but other than that I'm not sure it will be very impressive.  Or at least that's how it appears so far, though for all we know the SA/TA boost could be rather significant (beyond X's or Rogue'sArmlets+1).  Mandalic stab is also increasingly useful on opponents with increasingly higher defense, so Vajra might have niche use for such enemies (Ixion, albeit not melee'd, comes to mind).  And at least it can't poison things.

	Random Pix

.oO: Click to enlarge :Oo.


It's a bit self-absorbed to use myself as a model for how one should gear their thief.  So instead, I'll say that what is presented here is one way of gearing one's thief, and that I feel that the builds I have created have been very useful for my purposes thus far.  Likewise I welcome other thieves to try similar builds if they have not done so already.

I'll say a little about each of my builds; firstly about the gear that they contain and secondly about the purpose and uses behind each gear set.

Set Compositions
Weapons (main/sub): Mandau/X's mainly.  Other options: Blau Dolch, Thief's Knife, Ermine's Tail, Adder Jambiya, Tatami Shield, etc.

TP Build #1 (VT+, /nin): Fire Bomblet, Walahra Turban, Love Torque, Brutal Earring, Suppanomimi, MirkeWardecors/HomamCorazza, Homam Manopolas, Rajas Ring, Toreador's Ring, Forager'sMantle/Cuchulain'sMantle, Swift Belt, Homam Cosciales, Homam Gambieras
TP Build #2 (T-): Fire Bomblet, Walahra Turban, Love Torque, Brutal Earring, Suppanomimi, Rapparee Harness, Homam Manopolas, Rajas Ring, Ulthalam's Ring, Forager'sMantle, Swift Belt, Homam Cosciales, Homam Gambieras
TP Build #3 (TW or Feint): Bomb Core, Walahra Turban, LoveTorque/JusticeTorque, Brutal Earring, Suppanomimi, Rapparee Harness, Homam Manopolas, Rajas Ring, Ulthalam'sRing/TriumphRing+1, Forager'sMantle, Swift Belt, Homam Cosciales, Homam Gambieras
Subtle Blow DPS Build: Bomblet, Turban, L.Torque, Brutal/Suppa, DragonHarness, HomamHands, Rajas/Toreador's, Forager's/Cuchulain's, PeisteBelt+1, Enkidu'sSubligar, HomamFeet

WS, Mercy Stroke: BombCore, Heca Cap, Justice Torque, Brutal/Triumph+1, Heca Body +1, Heca Hands +1, Rajas/Triumph+1, Forager's, Warwolf, Enkidu's Sub, Heca Feet +1
WS, Evisceration/DE: FireBomblet, Heca Cap, Love Torque, Brutal/Hollow, Antares Harness, Heca Hands +1, Rajas/Adroit+1, Cuchulain's Mantle, Cuchulain's Belt, Oily Trousers, Heca Feet +1
WS, Shark Bite: BombCore, Heca Cap, Love Torque, Brutal/Hollow, DH, Heca Hands +1, Rajas/Adroit+1, Forager's, Cuchulain's Belt, Enkidu's/Oily, Heca Feet +1

Trick Attack: BombCore, Denali Bonnet, Love Torque, Brutal/Suppa, Denali Jacket, Rogue's Armlets +1, NimbleRing2x, Forager's/Charger, Scouter's, War Brais +1, Strider Boots
Sneak Attack: BombCore, Heca Cap, Love Torque, Brutal/Suppa, DH, Heca Hands +1, Rajas/Adroit+1, Forager's/Charger/Cuchulain's, Cuchulain's Belt, Enkidu's Sub, Strider Boots

Max Evasion (Alchemadian/Nefarious set =D): Ermine's Tail + another dagger, GoblinCracker/UngurBoomerang/LeoXbow+1, Optie Hat, Eva Torque, Musical/Novia, Antares Harness, DenaliHands/MasterMitts, Nimble2x, Boxer's, Scouter's, Raven Hose, DanceShoes+1
UtsuNi, UtsuIchi(Part 2), and alt Eva/Haste Hybrid: Turban, Eva Torque, Musical/Novia, Rapp, Homam Hands, Nimble2x, Boxer's, Swift, Homam Legs, Homam Feet
Mid-Range Enemy Tanking set: Denali Bonnet, Eva Torque, Brutal/Suppa, Antares Harness, Denali Hands, Rajas/Toreador's, Boxer's, Scouter's, Homam Legs, Homam Feet
High-level Enemy Tanking set: Optical Hat, Eva Torque, Musical/Novia, DH, Denali Hands, Rajas/BqRing, Boxer's, PeisteBelt+1, Enkidu'sSub, DanceShoes+1

Ranged accuracy: LeoXbow+1(Needs upgrade to Ziska), Optie Hat, PCC, Suppa/Hollow, Pahluwahn Kzahagand, Deadeye Gloves, Behemoth/Behemoth+1, Jaeger Mantle, Precise Belt, Dusk Trousers, Homam Feet
Ranged DPS: Same as above, only with Coffinmaker, Iron Bullets, TriumphEarring+1, AmemetMantle+1.  Other options are Gun Belt, Beater's Earring, and Trailer's Kukri.

/war /drk HNM'ing TP set: Mandau/Tatami, and same as normal TP set except with Hollow Earring, Charger Mantle, no Mirke Wardecors, and sometimes Homam Head.
/drk Zerging: (needs M.Kris), Mandau/Tatami, Turban, Love Torque, Brutal/Cassie, Rapp, Homam Hands, Rajas/BqRing, Gigant, Swift, Homam Legs/Feet.


Set Applications
Weapons (main/sub): 
Mandau: Not much to say about it.  If you're reading this, you already know something about it.  If you don't, there's a whole post on it lower in my LJ.
X's Knife: This was a pleasant surprise for me.  At first I was not terribly impressed.  I noticed a TP gain loss changing to this from blau.  Also, my first eviscerations were rather mediocre.  Playing Thf with Mandau is different than playing without.  Playing with Mandau/X's is a different style in itself.  If things go well I'll make an entire post about X's Knife and evisceration, as there's enough to say elsewhere.  To summarize, though, this is excellent if you find yourself hovering over 100% tp often.  It's also excellent if one has a very good evisceration build.  Otherwise it is nice, but only a mediocre increase.
Blau Dolch: Generally the best mainhand or offhand dagger.  Post-mandau I use it with X's or another combo when I don't want to poison things.  There are times where I don't get to do a whole lot of SA/TA and I use blau instead of X's.
Ermine's Tail: Best max evasion dagger.  Great for solo/tanking.
Adder Jambiya: One of several options to use for a evasion/tanking build, dual-wield or otherwise.  I'm about to craft this and use it over Azoth due to it's nice additional effect.
Thief's Knife: TH, yada yada
Tatami Shield: Handy single-wielding partner.  Superior to viking shield if you're using mercy stroke.
Vajra: Currently upgrading.  Hey, it doesn't poison things.  And it looks like something a power ranger might use.

TP Build #1 (VT+, /nin): This is essentially my "merit build" and my overall "main TP build."  I've found MarinaraPizza(+1) to be far too good to be using other alternatives except lolRedCurry in merits.  If I'm cheap or low on invent space I use CoeurlSub for some events, but not merits.  This is not true at GreaterColibri of course.  If I'm there, I'll use this build but I'll also use MeatMithkabobs or Crab Sushi.  Take a look at crab sushi if you don't know about it.  Consider it.  As a final note, I am still indecisive as to when or if I should TP in Cuchulain's Mantle, other than if I'm a skillchain partner and need the accuracy.
TP Build #2 (T-): This was my sushi build.  I rarely use sushi now, but if I do I might use this.  However, this set also works great for EM and T mobs.
TP Build #3 (TW or Feint): The goal of this set is to maximize on haste and other DPS stats while stressing as little accuracy as possible.  It's similar to the T- build slightly.
Subtle Blow DPS Build: This build stacks as much subtle blow as reasonably possible.  All other stats go into maximizing DPS.  I use this occasionally in solo, but more often in low man or on HNM.  I consider this a very important build for Thief, but one that few have created for themselves.

WS, Mercy Stroke: Same set for both SA/TA mercy.
WS, Evisceration/DE: I don't really use Dancing Edge anymore.  This is for unstacked Eviscerations.  I followed the advice of some recent threads, for example this one and this one, as well as things I read and heard in game and on Allah'.  The goal is to stack as much dex as resonably possible to reach or approach the crit cap.  Accuracy is the second most important stat.  Put enough into a build like this and you will see good numbers.
WS, Shark Bite: When there is a need for light, shark bite is there.  I try to stack as much dex and attack into this build as I can.  I don't have Mandalic Stab, otherwise I would make a build for that, but from what I hear it is similar to this build.

Trick Attack: The set for solo TA.  Movement speed and swing delay are taken into account for optimal performance, as opposed to raw spike damage.
Sneak Attack: The set for solo SA.  Movement speed and swing delay are taken into account for optimal performance, as opposed to raw spike damage.

Max Evasion (Alchemadian Set, tm): As much evasion as possible.  This set is used for the first part of ichi.  It is also used for the second part of ichi if getting shadows up is more critical at the time than shadow recast.  This is chiefly the "oh shit" set.  This is also the set to use when shadows are up but recast timers aren't looking too great.  This is seldom my main tanking build.
UtsuNi, UtsuIchi(Part 2), and alt Eva/Haste Hybrid: I mostly use this set for Utsusemi: Ni.  It is also the second part of Utsusemi: Ichi.  Occasionally I will use this set to tank in.  It has more haste but less evasion than my mid-range enemy set, so it is useful if the mid-range set is semi-evasive.  Generally if that is the case though, I'll be using a normal evasionless DD TP build anyway.
Mid-Range Enemy Tanking set: This is the set I ideally use for tanking most mobs in the DC->T range.  Occasionally it is OK for a VT, or there is occasionally an EP that attacks fast or accurate enough to where this is necessary.  This set hybridizes offensive and defensive stats.  One could achieve the same effect by interchanging between a DD and heavy/max evasion set, but this set does the job in a simpler fashion.
High-level Enemy Tanking set: This is the build I most enjoy tanking in.  This is for VT+ enemies.  There are few enemies that I could/would tank where this set does not have enough evasion without mambos, but those few enemies do exist.  In most cases though, this set is terrific for reducing an enemies TP moves while also evading attacks.  For difficult mobs, I also consider HP to be very important.  A mithra Thf/Nin in an evasion set has very little HP, (between 1k to 1.2k).  Difficult enemies that would call for this set can often do TP moves for over that amount of HP.  Therefore I often use BombQueenRing and Carbonara when using this set.

Ranged accuracy: This set is for stacking as much ranged accuracy as possible to land your bolts or whatever it may be.  Oftentimes during solo/low-man play this set can be counter productive in that it lacks evasion.  On tough enemies where you might have difficulty landing bolts, you might also have difficulty evading.  Therefore I sometimes hybridize this set with evasion pieces.
Ranged DPS: This is for ranged DPS, /rng or otherwise.

/war /drk HNM'ing TP set: The H in "HNM's" signifies that they are difficult.  Thus, different stats are sometimes required.  Excessive haste sometimes feeds too much TP.  You don't typically get the merit party buffs you're used to.  Your fStr, hit rate, and pDif levels are all lower.  So sometimes different equips are required.
/drk Zerging: The goal of this set is to stack as much haste as possible, and otherwise stack as much HP as reasonably possible.  And it favors attack and str over accuracy.  /drk zerging is also something that should often be done with outside cures in mind, as souleater decreases in potency as HP decreases.  Without enough cures, a /war zerging set would prove better.



And so that is a list of some of the nicer gear that I have accumulated as I have been playing, and most of the different combinations in which I have come to use said gear for my Thief.

From: www.playonline.com/ff11us/polnews/news15265.shtml "Jan. 22, 2009 03:00 [PST] From: FINAL FANTASY XI

Handling Misconduct (Jan. 22)

In November 2008, we discovered an issue that allowed players to create multiple items for certain treasures and rewards in areas such as Salvage and Assault by exploiting the game system.

The issue was fixed during emergency maintenance on November 26, 2008. However, we discovered that some players had already exploited the issue and used it to gain additional items before it was addressed. We then investigated more than a year's worth of logs throughout all areas.

As a result of the investigation, approximately 400 players were temporarily suspended based on the evidence gathered. Approximately 550 players who committed more serious misconduct had their accounts banned.

We are working hard to ensure that these kinds of issues do not occur again in FINAL FANTASY XI. However, if you discover any problems with the game system, please submit a GM call or fill out the feedback e-mail form to report the problems to us, instead of exploiting or spreading the issue.

We thank you for your cooperation in this matter."
-SE, January 22nd, 2009

I never expected the salvage dupe repercussions to be this harsh.  400 temp bans and 550 perma bans might not seem like a whole lot, but I've definitely felt this number on a personal level.  Perhaps it's just that Asura was one of the largest servers hit, or maybe that I knew a lot of people that just loved salvage too much.

The bans this time aren't "nameless" RMT.  These are long time, otherwise very legitimate (most of them), NA endgame players--some of the most loyal subscribers SE has.  Among the count of people banned, we so far have a confirmed 67 relics removed, 2 mythics (the only two to have been confirmed completed), and 10 endgame LS's currently left leaderless.

Not only is the mass ban of somewhat questionable ethic, but it is probably a poor financial decision on their part.  Not only did they permanently ban 550 of their most loyal customers, but they've caused a severe general upset.  I guess I can't say that they've lost my subscription, but I know many who are thinking seriously about quitting because they've just lost some of their best friends--just about the only thing keeping them here in this "dying" game.  And with those people leaving, I imagine I'd be more keen to the idea of leaving.  I probably wouldn't, but I can see others who would as part of the spiral.

Is it really so hard to temp ban people, SE?  I guess not, there are 400 temp bans.  It seems like everyone and their mother that I'm hearing that has been banned, though, is calling up SE and hearing that their LM-17 is a perma ban as opposed to temp (LM-11 is usually perma; LM-17 is temp, pending review for perma).  Many of these people, a great many, have never before had a ban, been jailed, or even had a GM call made on them before.  Some people that got LM-17'ed didn't even know they were part of a duping run, where others who duped rampantly got by with a warning letter or nothing at all.

There is almost no consistency here.  Sadly, like most of the other problems in this game, I think what's to blame regarding SE's actions is their lack of current financial support for FFXI.  Interviews these days talk more and more not only of PS2 limitations, but of fund limitations.  It's an old MMO, so that's understandable.  If SE never had the funds to have enough employees or investigative power to have ever found out about or acted on the bug in the first place, it's easy to see that they wouldn't have the funds for a more thorough investigation, to guarantee consistency.  Lack of funds are really the reason why we're getting really slow content updates now, and why job and content adjustments never come as promised, or come many months to years down the road.

This is just my own very subjective opinion, but slavage duping to me did not very much seem like any other exploit in the game.  Sure, it was an unfair advantage of game mechanics.  But it is nothing like other exploits and cheats in the game we are normally familiar with.  Firstly, there had never been any action taken on SE's part for this kind of behavior prior, so players really wouldn't know for sure what the consequences would be if they (1) duped regularly, (2) duped once or twice, (3) took place in the duping process, but got no gear, or (4) partook in a run involving duping but held no interest in duping specifically.  It is clear that the duping was a bug, and that the TOS is against the exploitation of bugs.  Still, some level of severity is attached to this.  Perma-banning is the ultimate retaliation, and should be reserved for an ultimate offense.  Similarly, as offenses scale down, repercussions should also scale down accordingly.

No, you can't compare this to duping gil.  There was no end to how much gil you could dupe, when it was dupeable.  You could (more easily) RMT the duped gil, or use it to control various aspects of the game markets.  Not the case with salvage gear.  And again ease of duping here is important.  It's not like people went into salvage and pressed a button and got everything they wanted in one run.  It should be a testament to how gay salvage really is, that many of the people have put so much effort into salvage regardless.  I can think offhand of one person who has played for many years, just completed his full set of skadi in the past month or two, and went to salvage 6-7 days a week during most weeks for one-and-a-half-years.  And that's with knowing about duping.  Yeah, salvage drops really are that bad.  Now compare that to the duping of gil which occurred in the very early stages of the game.  You could dupe whenever you wanted and as much as you wanted.  Quite a difference.  And people currently know the reprecussions for duping gil, if they could do it.  People know the reprecussions if they bot claim or fish bot, speed hack or pos hack.  People know these things and can expect a perma ban.  Even so, they may be subject to only a temp ban.  Maybe it's me, but I think less dramatic action should be taken to first-time bannings of new offenses.

I'm in a bit of a depressed mood, but there's still a bit of hope.  Out of the dozen or so people I know well that have been hit by the ban, only a few have confirmed perma ban.  Here's to hoping the rest are not.  And to the more than dozen people that I knew less well to not know their IRL info, I'll miss them if they don't come back.  Like a particular JP player I know, I know that if he's perma banned (which he probably is), I'll probably never be able to talk to him again.  Sucks not having had the chance to say bye.  And yet I still tuck my balls and play this game.  Thanks, SE.

Relevant Links/Threads
POL "Handling Misconduct" Notification
Blue Gartr: LM-17 Thread
Blue Gartr: List of Players Banned
Blue Gartr: Quotes from Original Dupe Thread
FFXI-AH: LM-17 Thread
Allakhazam: LM-17 Thread
KI, Asura: LM-17 Thread
KI, Asura: List of Players Banned

	I have all that stuff at the bottom to update, but I really am that lazy that I haven't written anything yet.  I do, however, just so happen to have a bunch of my own JoJ popsets though, so I figured if some friends and I are going to pop him for fun, I might as well know if the torque happens to be better than soil gorget, etc, for mercy stroke.  Normally when I do these equip comparisons I just write them down on wordpad, figure out what I want, and throw them in a folder.  But I figured if I did the number crunching in the LJ post box instead of wordpad, that I actually might post something.

So for this comparison, we're going to use one of the golden standards (for whatever reason) of mobs as our basic settings.  That is, the lvl82 Greater Colibri.  Sad that I know by memory that it's defense is 327 and its VIT 67.  I don't know it's evasion by heart, but as with other comparisons I've done, that's not really important here and is just extra work, as if you're at that camp you're usually sporting madrigal, which if you don't suck brings you basically to the acc cap.  Further, for mercy stroke the mobs evasion really doesn't matter very much anyway.

For the next part of the settings I'll use my character stats.  Lvl 75 mithra Thf/Nin,

, with a total of 125Str and 432 attack.  Let's say march/madrigal so no minuets.  And let's also say I'm too lazy to have brought mithkabobs or any other food for this test.  Let's also restrict the test to Sa+Mercy, as for Ta+Mercy it'd pretty much just be Justice Torque vs Soil Gorget.  But for Sa+Mercy we could also compare Love Torque and Kubira Bead Necklace as well.  We know that as fStr approaches its cap for whatever weapon class someone is using, the Str:fStr ratio becomes 4:1.  That'd put me at 14".5" for this example, or over the cap of 12 for Mandau and 10 for Blau.  pDif median here is going to be ([(432/327) - 0.35]*1.2 - 0.25), which is 0.914 for non-crit.  This is with no neck stats.  For multi-hit procs on WS, I have brutal for 5% double and native+merits for 10% triple.  If we pretend we know that double attack check goes before triple attack, what I get is that I probabilistically get an extra 24% swings on each hand.  This is from (1*1.05 + [0.95*1.2 - 0.95]).  So from an average probabilistic standpoint, my average number of hits per /nin mercy stroke is 2.48.  With these settings in mind, we can now do our comparison.

I can do this comparison on all hits (which would be more accurate), or for ease stick with a 1-hit comparison.  If it holds that Justice Torque is better for 1 hit Sa+Mercy than soil gorget or love torque, it'll be better for multihit.  But this is not the case when comparing to Kubira Bead Necklace.  However, Justice Torque would be superior for Ta+Mercy anyway; and if Justice Torque is superior to love and soil for Sa+Mercy, it'll obviously be better for Ta+Mercy as well.  I'll start iwth some 1-hit comparison and then switch over to the multihit.

1-Hit Sa+Mercy Stroke
Soil Gorget
1-hit Sa+Mercy = {(39bD + 12fStr + f[0.83 * f(0.6*125)]WSC)*3.1fTP + 103dex} * 1.914pDif * 1.25piercing
= ((113".25" * 3.1) + 103)*1.914 *1.25
= 1083dmg
To generalize this comparison, to make it more accurate for everyone and not just my own set, I can alternatively not truncate anything.  Skipping truncation gives me,
= 1086.4dmg

Justice Torque
Additional 5str will not affect my fStr here.  It will change pDif slightly, from a non-crit median of 0.914 to 0.924 with only 2".5" attack.
1-hit Sa+Mercy = {(39bD + 12fStr + f[0.83 * f(0.6*130)]WSC)*3.0fTP + 103dex} * 1.924pDif * 1.25piercing
= ((115".74" * 3.0) + 103)*1.924 *1.25
= 1077dmg
Skipping truncation steps to generalize this, I get,
= 1082.8dmg

So far it looks like Soil Gorget is better here for 1-hit Sa+Mercy.  If Str is lower, Justice Torque will become better.  Similarly if fStr is lower (ie you're fighting much harder stuff), Justice Torque will also become better.  Same if pDif is lower.  Given how close they are here for 1-hit, it's very likely that Justice Torque is better for anything HNM level.  This, again, is also 1-hit.  The more hits you add, the better Justice Torque becomes in comparison.

Now let's take a look at love torque and kubira bead necklace for 1hit SaMS, this time also skipping truncation to favor generalization for each example.

Love Torque
1-hit Sa+Mercy = {(39bD + 12fStr + f[0.83 * f(0.6*125)]WSC)*3.0fTP + 108dex} * 1.937pDif * 1.25piercing
= ((113".25" * 3.0) + 108)*1.937 *1.25
= 1084.1dmg

Kubira Bead Necklace
1-hit Sa+Mercy = {(39bD + 12fStr + f[0.83 * f(0.6*129)]WSC)*3.0fTP + 107dex} * 1.923pDif * 1.25piercing
= ((115".24" * 3.0) + 107)*1.923 *1.25
= 1088.2dmg

2.48-Hit Sa+Mercy Stroke
As mentioned earlier, /nin with traditional merits and WS gear (5/5 triple, brutal earring), the average mercy stroke is 2.48 hits.  While there are a number of offhands available, the most common used for DD'ing is still probably blau dolch, which is also what I have.

Soil Gorget
2.48-hit Sa+Mercy =
({(39bD + 12fStr + f[0.83 * f(0.6*125)]WSC)*3.1fTP + 103dex} *1hit *1.914pDif
+{(39bD + 12fStr + f[0.83 * f(0.6*125)]WSC)*1.0fTP} *0.24hit *0.914pDif
+{(33bD + 10fStr + f[0.83 * f(0.6*125)]WSC)*1.0fTP} *1.24hit *0.914pDif)
*1.25piercing
= 1.25*(869.1 + 24.8 + 119.3)
= 1266.5dmg

Justice Torque
2.48-hit Sa+Mercy =
({(39bD + 12fStr + f[0.83 * f(0.6*130)]WSC)*3.0fTP + 103dex} *1hit *1.924pDif
+{(39bD + 12fStr + f[0.83 * f(0.6*130)]WSC)*1.0fTP} *0.24hit *0.924pDif
+{(33bD + 10fStr + f[0.83 * f(0.6*130)]WSC)*1.0fTP} *1.24hit *0.924pDif)
*1.25piercing
= 1.25*(866.2 + 25.7 + 123.4)
= 1269.1dmg

Love Torque
2.48-hit Sa+Mercy =
({(39bD + 12fStr + f[0.83 * f(0.6*125)]WSC)*3.0fTP + 108dex} *1hit *1.937pDif
+{(39bD + 12fStr + f[0.83 * f(0.6*125)]WSC)*1.0fTP} *0.24hit *0.937pDif
+{(33bD + 10fStr + f[0.83 * f(0.6*125)]WSC)*1.0fTP} *1.24hit *0.937pDif)
*1.25piercing
= 1.25*(867.3 + 25.5 + 122.3)
= 1268.9dmg

Kubira Bead Necklace
2.48-hit Sa+Mercy =
({(39bD + 12fStr + f[0.83 * f(0.6*129)]WSC)*3.0fTP + 107dex} *1hit *1.923pDif
+{(39bD + 12fStr + f[0.83 * f(0.6*129)]WSC)*1.0fTP} *0.24hit *0.923pDif
+{(33bD + 10fStr + f[0.83 * f(0.6*129)]WSC)*1.0fTP} *1.24hit *0.923pDif)
*1.25piercing
= 1.25*(870.6 + 25.5 + 122.7)
= 1273.5dmg


The reason I'm so anal about my comparisons is because it turns out that, even though SE probably doesn't know what they're doing when they're designing this gear, the numbers from one situational piece to another always turn out to be so close.  It seems to always come down to situationality in the end.  Given that, I kinda feel bad that I didn't include accuracy as part of this comparison, but I got other things to do anyway, yo.  So here's what this comparison probably boils down to: 


#2 for Sa+Mercy by this comparison (greater colibri march/madrigal no food), and #1 for Ta+Mercy.  Should be the best mercy neck for HNM's.


#1 for Sa+mercy by this comparison, and should be #2 for Ta+Mercy.


Lingers somewhere in between the above two necks and soil gorget for Sa+Mercy.  Likely inferior to all for Ta+Mercy.  When you're very high on pDif , gorget should be best, but I'm not sure if there's a point where love torque would ever be the best option.  #3 by this comparison for Sa+Mercy, and #4 for Ta+Mercy.


#4 by this comparison for Sa+Mercy, and #3 for Ta+Mercy.  However, this is probably going to be the best mercy neck at some inflection point of pDif, when pDif is high enough.  If not on merit level mobs, this might be the best option for very weak mobs, like too weak to even match.  But then again, the first hit of mercy becomes less and less important as pDif rises.  Either way, it doesn't matter.


Also, ding, Justice Torque later in the day after getting half way done with posting this /(^_^)/

I bought some popsets and some friends and I went out and 5-manned it for fun.  Went 1/1.

Rather than letting this LJ rot completely, I'm going to try and start adding a lot of what I've been meaning to for a long time.  Over the course of the past half-year and more, there have been a number of topics that I wanted to post or talk about, either on here or on various thief forums.  People can be lazy though and I'm no exception.  I think the main reason that I don't post much though is that I don't like to do so unless I feel that what I've written is near as perfect as possible.  So I'll start to write a huge WoT fraught with insight, questions, and mathematical breakdowns but end up letting it sit on wordpad somewhere.  Most of the topics below are subjects which I've already written extensively on or have actually tried out in game.

Anyway, here are the options that I'd like to start writing about, either because I feel that they're important topics for furthering the game, or subjects that are worth looking into much more than they have yet been.  If I manage to post on even some of them I guess I'll start striking through them until I'm done.

The smaller, italicized are entries that I have now already posted since this entry had been made, or are otherwise which I find no longer relevant or important.

PvE TP feeding
How much tp/time do melee DD's typically feed to mobs?  How do the various classes stack up against one another as far as TpFed/time is concerned?  When is subtle blow worthwhile, how much does it help an individual, and what are available options?  Does mob TP move behavior always reflect the amount of TP being fed, or is it conditional; that is, when a mob gains x amount of tp/time, how does it relate to y amount of TpMoves/time--under what conditions are x and y unequal, why, and what can we gain from this knowledge?

Thf/Drk Zerging
When if at all is it practical?  Does the individual need their own personal cures or "PL"?  What are the ideal weapons and builds to use for it?  How does thf/drk stack up against other zerging dd's?

Dancing Edge vs Evisceration
Does there come a point where having a certain amount of gear or buffs allow EV's damage to excede DE's?  What are some good equipment and party setups to support evisceration?  How good is X's knife really and does it really require much to make EV excede DE if you own one?

Thief Tanking, etc
When is thief tanking actually feasible or indeed desirable--vs what mobs, and with what party and equipment setups?  Are there any tank party setups that would take thief tanking beyond previous limits, and how effective could such setups be?  What is the ideal combination of evasion, HP, subtle blow, DPS, and other stats for thf tanking?  How many macros could one use to jumble these stat combinations into effective setups, and when and how would they be used?  What are some personal tanking strategies (which don't include gear or macros) that are effective?

Vajra
How good can we expect it to be?

Theorycraft: Relic/Mythic Thf/Drg
As with many DD classes, dragoon is a good sub at very early levels.  With the exception of elite endgame Sam/Drg setups within elite party setups, dragoon sub is not very common past the early levels.  Could it be possible that thieves, specifically relic and mythic thieves, could also put this sub to use in the endgame?  What are some advantages to the sub and what does the math predict as far as overall situational DPS?

Feint Testing
Not that it matters much, but will we ever know the precise potency value of feint?  Is it a specific -% to the mobs evasion, or a straight chunk of -evasion?

Unanswered TH Questions
When and how is TH value registered on a mob?  That is, is it ok to strike the mob with TH3 or 4 but let the mob die with only TH2?  Will TH3 / 4 still remain?  Are there some items or types of items where the corresponding droprates are unaffected by TH?  Will we ever figure out if day, moon, or facing east during windsday while killing the mob with a wind staff will increase drops?

Mobile Zerging
Most FFXI players are by now familiar with zerging.  But what about mobile zerging?  My friends and I sometimes joke around about our "mobile zerging unit."  Is having a plethora of permanently buffed 70-80% haste DD's really all that?  Would this be an effective way of doing damage in situations like einherjar, dynamis, bcnm's, merit parties, or anywhere?  What kinds of equipment and party setups are required?

AV Zerging
Can we mathematically deduce a setup that could possibly zerg AV at the rate that KC Drks took him down in September '08, without souleater?  What would such a setup look like and could it be pulled off?

Gardening for Gil
How did one man fund a relic weapon through virtually nothing but gardening, and is he sane?

Stacking Defense-down
What is the maximum amount of -% defense that players can inflict on a mob, and is it worth it to tailor party setups and strategies towards achieving this number?

Marinara Pizza vs Meat vs Sushi
Could pizza be as epic in-game as it is in teh real?

Went out to test the mods for Pyrrhic Kleos today, the mythic Dnc WS.  Here's what me and my cohort Leonstrife got.

Test Conditions
*jambiya, base damage 33
*fStr cap of 11 ; new base damage of 44
*leonstrife's base attributes: str65, dex77, vit57, agi76, int59, mnd51, chr59, as a mithra 75Dnc/Thf
*West Ronafure mobs: Wild Rabbit, Wild Hare, Tunnel Worm, Orcish Fodder.
*All WS were done at ~100 to 105 TP, with Sneak Attack.
*capped pDif multiplier range w/ crit, assumed to be roughly 2.6 to 3, based on public data. (edit: I know the spread is larger than this, and I know the max is slightly higher than this.  I just don't know how high.)

Field Data
WS's, naked, 301, 312, 332, 293, 321, 327, 337, 290
WS's, STR etudes, +25, 362 ; +26, 352; +25, 360; +26, 363; +25, 351; +26, 311;
WS's, DEX etudes, +25, 320; +26, 355; +26, 362; +25, 360; +26, 319; +26, 364; +26, 327; +27, 316; +26, 356;
WS's, CHR, +25, 290; we know that agi, int, mnd, vit cannot mod either.

Checking Data for tolerable Accuracy
3/2.6 is 1.153, so we want a split similar to that with our high/low ws's to ensure accuracy.  We have:

naked, no etude high/low = 337/290 = 1.162
dex etudes high/low = 364/316 = 1.152
str etudes high/low = 363/311 = 1.167

Skillchain Properties
Distortion/Scission ; With Wild Flourish, Scission
(from http://wiki.ffxiclopedia.org/wiki/Pyrrhic_Kleos)

Thanks Fred :)

WSC Guesswork
dmg = 332 = pdif(3) *remainingformula,
110.6 = ftp*(wsc + 44),
110.6/ftp = wsc +44,
110.6/ftp = 44 + f(f(dex + str)*0.83)
pDif cancels out.  Actual numbers used below.

ftp cancels out here as well: via naked/ftp / dex/ftp
dexHigh = [44 +f(f((77 + 26)*%)*0.83)]
/nakedHigh   = /[44 + f(f(77*%)*0.83)]

DEX
0.2?
364/337 = (44 + f(f(103*0.2)*0.83) / (44 + f(f(77*0.2)*0.83)
1.080 = 60 / 56
1.080 = 1.071 = close
0.25?
1.080 = 64 / 59 = 1.085 = close
0.3?
1.080 = 68 / 63
1.080 = 1.079 = very close
0.35?
1.080 = 73 / 65 = 1.123 = meh
0.4?
1.080 = 78 / 68 = 1.147 = eww
0.5?
1.080 = 86 / 75 = 1.147 = eww

STR
0.15
1.077 = 54 / 51 = 1.059 = meh
0.2?
363/337 = (44 + f(f(91*0.2)*0.83) / (44 + f(f(65*0.2)*0.83)
1.077 = 58 / 54
1.077 = 1.074 = close
0.25?
1.077 = 62 / 57 = 1.088 = closeish
0.3?
1.077 = 66 / 59 = 1.119 = meh
0.35?
1.077 = 69 / 62 = 1.113 = eww
0.4?
1.077 = 73 / 65 = 1.123 = eww
0.5?
1.077 = 81 / 70 = 1.157 = eww

WSC Modifiers appear to be STR20% and DEX30%.

Solving for fTP
Our +Str high/low had the largest spread, so those high/low most accurately represent the pDif crit cap/floor.  Let's assume the pDif crit floor is 2.6 and the cap is the spread times that, which would be 2.6*1.167 = 3.034.  Let's run with that.  Let's assume that we've found the Str and Dex WSC's, 20%STR and 30%DEX.

363 = 3.034 * fTP * (44 + f(0.83* f(91*0.2 + 77*0.3)))
119.64 = fTP * (44 + 34)
119.64 = fTP * 78
fTP = 119.64 / 78 = 1.534

fTP appears to be 1.5 @100tp.

This is all just copypasta'd from notes I took in wordpad, so some of it is a bit sloppy.

Nothing breathtaking here, but I've been grinding through some anime the past few days, namely Welcome to the NHK, when this completely random FFXI reference came up.  Was humorous to me, because I was actually in the middle of using the free trials portrayed in the video to make gil for myself.
I like how the kid describes RMT so admirably and innocently.  Good anime btw; I'd recommend it to anyone who would bother watching more than just shounen.

I'd also like to send out propz to my friend Athena for this sketch rendition of my char.

...although she's a cheater at iSketch... >,<;;

Como se dice, update?!
So I haven't updated in ages.  I've been playing a lot, but have been less active on forums and such.  The most important update is, of course,

Mandau Obtained!
Upgraded from start to finish in a little less than 9 months.  It was some of the most grueling work I've ever done, both in game and IRL.  I'd like to shout out thanks to Asuraka and Vistani, for covering the 3k byne loan.  Thanks as well to my good friends Quintessence and Kenjiri.  If it wasn't for their inspiration and encouragement, I wouldn't have been able to develop the specific methods I used to almost entirely self-fund the weapon in such a short period of time.  Thanks to Sige and Zombiejonathan for helping out when it was needed, and a great big thanks to the Provenance linkshell, and Derivation dynamis linkshell, for allowing me the opportunity to sponsor runs.  I've had it unlocked for about 3 months now; I'm just uber lazy with updates.


Me and my alt, Tepic.

I'd like to offer something fresh and unique to say about Mandau.  Something that hasn't been said or done before.  But it's not as if an unlocked relic dagger is something new.  It's currently the second most popular relic, with almost 200 or more owners.  But I guess I'll give my own two-cents about it, even if most/all of this has been said before.

Unlike a fraction of relics out there, Mandau doesn't generally break the game by offering utility which can greatly affect strategies and outcomes.  G'horn and Aegis do that, and Apocalypse pretty much used to do that before the March '08 haste cap adjustment.  But mandau is essentially like the other weapon relics in that it's primary utility is from much enhanced DPS.  The foremost two marginally useful relics that come to mind would be Mandau and Amanomurakumo, simply because they offer WS's that are clearly more damaging than alternatives.  The other weapon relic WS's are either mediocre, or for situational use.

Some Advantages of Mercy Stroke Mercy Stroke is what makes Mandau uber-tasty.  A strong one-hit WS means consistency and high damage when stacked with Sneak Attack or an assassin-trait Trick Attack.  All of the Relic WS (~save knuckles) are strong 1-hitters.  Being that SA/TA are designed to maximize the power of the first hit of a WS, any strong one hit WS will do nicely when stacked with them.  Given that so much about Thf DPS is centered around SA/TA, a thief's Mercy Stroke is able to take full advantage of its WS potential.  Here are just a couple epeen examples,


SA/TA excel on targets with high defense and/or evasion.  Stacked SA/TA one hitters guarantee consistency no matter how high the mobs evasion.  Meanwhile, a forced crit offers a straight 1.0 addition to pDIF, making the marginal increase of a crit more potent as atk/def gets lower.

Hide is fun, but I can't seem to break ~2600 at birds yet :(

Mercy Stroke is a very powerful DPS tool at Jailer of Love.  Not only is JoL's evasion and defense high, making SA/TA useful, but JoL also has a common Yovra TP move called Fluorescence, which is a boost effect that greatly raises the base damage of the next attack.  So for a 1-hitter like Mercy, this guarantees a tremendously powerful WS if you can absorb the effect using the aura steal trait.  The above 2.5k is actually mediocre, compared to what others have done.

Relic WS's and SA/TA with /Thf sub
Just to exemplify what I mean about relic WS's being nice when stacked with SA/TA, its probably worthwhile to show this screenshot from when I was messing with bard,

This was brd/thf with a full sata.  So you don't have to be Thf main to produce high SA/Sata relic WS's.

And then, you get general high damage from MS, not simply in virtue of its being a one hit WS, but in virtue of it leaving room for plenty of multi-hit procs.



The first of these was a Trick Attack MS, and the second two were Sneak Attack MS's, all 5-6 hit.  Mercy Stroke with nin sub will be 2 hits instead of one, and the great thing about that is also that either of these hands can proc a multi.  With full triple merits and brutal earring, thats a 10% triple and 5% double rate on both hands.  Whereas Dancing Edge and Evisceration hits the multihit WS cap of 8 with only a single triple proc, WS's like Mandalic Stab, Shark Bite, and Mercy Stroke can take full advantage of potential multi-procs.  While a full 6 hit mercy is rare, doubles and triples still proc very often, turning an already strong WS into an even more powerful one.

On the other hand, if the SA/TA don't land, and especially if the mainhand misses, you end up with low numbers.  A more extreme example,

This is especially aggravating in situations where, say, a mob puts up a blink the moment that you WS.

Skillchain Utility
Lastly and importantly, Mercy makes for a nice skillchain closer.

I'm sure I've had darkness match on higher damage WS before, but it usually doesn't show up in the log, and closing a skillchain with it is oftentimes overkill.  But there is a lot of utility in being able to do high damage darkness SC's.  With a reliable partner, you can burn through mobs very fast in merit parties.  I've most commonly skillchained with warriors, samurais, and dark knights.  Skillchaining offers a bit of utility when you want to either zerg something down really fast, plant a lot of hate on yourself or another, or simply do a lot more damage.  There are a lot of situations where you can take advantage of skillchains.  Mercy is also a very nice alternative to evisceration, when required to do darkness for magic bursts rather than in virtue of the skillchain's own damage.

Skillchain Properties
There are two things about melee DPS game mechanics that I haven't taken the time to fully figure out yet, fStr and skillchain properties.  By that, I mean how you determine the skillchain effect that occurs when you combine multiple weapon skills, based on specific properties of each WS.  For general purposes, Mercy Stroke is in the bottom left corner of the level 3 skillchain chart.  But combining relic WS's together can get a little tricky.  While, for example, Mercy <-> Mercy (yes, the WS can make darkness with itself =D) and Mercy <-> Catastrophe make darkness in either order, Mercy Stroke can open darkness with Blade: Metsu but cannot close it.  Actually looking at more recent charts, Mercy -> Metsu should make fragmentation, but it doesn't.  This isn't the only exception.  There are at least a few more.  For more info, I would check the following BG thread, Gunslinger's Renkei Chart, for some nice discussion on relic skillchain properties.  A link to his improved version of Aden's skillchain chart is also in the thread.
Edited November 18th, 2008

3x Damage Proc
Is tasty.  The proc rate is low, but as you play you notice it often.  I've heard quotes of the proc rate being 5%, which sounds like a nice even number and seems about right based on experience, but I haven't went about and tested that number or seen any tests on it.  Here's some examples,
3x Normal Hit

What's interesting about this is that 125 is not evenly divisible by 3.  A sign of how the melee damage may or may not be calculated.  It can't be the case that the 3x-proc simply takes what was going to appear in the log, and multiplies it by 3.  Otherwise we wouldn't end up with 125 in this example.
3x Critical Hit

3x SA/TA


Aftermath: Additional Effect, Improved Critical Hit Rate
I've been meaning to play around with some robber crabs and find out the exact percentage of the boost.  Some have speculated that it's 5%, the assumption being drawn as a parallel to other relic aftermaths (can't remember offhand the relic WS's with which the potency %'s of the aftermaths are known).  I can tell you that when I'm meriting, I do notice a higher crit rate than I did before.  And with mandau, one does less free SA/TA in order to stack more WS's, so the increased rate for normal melee crits is actually higher than what's shown on the parse, since parsers don't/can't distinguish between normal crits and SA/TA crits.  The difference I'm seeing is roughly (based on memory and not loading up actual parsers) 11-14% pre-mandau, to 15-17% post.  As for duration, that was much easier to test since during the recent Fall '08 mythic weapon update/patch, an aftermath icon actually comes up after you WS (looks like a little pink spark).  So what I have so far: 

• Potency: Speculated to be ~+5%
• Duration: *20 seconds @100TP;   *20 seconds per every 100TP.

Edited November 18th, 2008

OmgPoison
The aggravating thing about the Mandau is its additional effect.  It's a testament to how nice the relic is though, that so many people have upgraded it despite this.  But just to clear a few important things up: 

• Additional Effect: Poison, is not a 100% proc; the proc rate isn't very high.  Maybe 5%.
• The poison does not last for more than a fleeting moment on HNMs (as if you'd need to sleep them).
• Mercy Stroke does not have a 100% additional poison proc attached to it.  In fact, I don't even think that additional effects on select weapons can ever proc on weapon skills.
• The poison does pose a problem for certain, dangerous, non-HNM level mobs.  And it can pose a problem depending on what a thief is trying to solo / low-man.

When the poison procs, you'll see this message.  This is standard to all weapons that have "Additional Effect: X" listed in its help text.  If you don't see this message, the mob has not been poisoned.  This will also not show on WS's, meaning that Mercy Stroke alone cannot poison mobs.

Fortunately, Mandau poison will very infrequently cause any significant problems as long as the weapon is not used when it shouldn't be used, and retarded mistakes aren't made.  Essentially, don't attack mobs that aren't appropriate targets.  IE random mobs in exp, non-targeted mobs in dynamis, etc etc.  It also doesn't help to poison a Zeni target mob that you're trying to take pictures of (my bad on that ; ;).  On top of knowing which mobs to attack and not attack, which should be common sense, one should always be prepared to simply unequip the dagger for appropriate fights.  A macro can do, perhaps via a macro set for the specific situation in question (dynamis, limbus, etc).  A common example are mobs that you sleep upon them 2houring, usually monk mobs.  Come prepared with an alternate dagger, and when you reach that mob, switch daggers, DPS, and switch back after you've burned your TP and the mob is dead.

The same applies to solo.  If you're fighting something that is difficult, they may often times resist sleep bolts.  But in the even that they do not resist sleep bolts and you wish to sleep them, simply opt out of using the Mandau unless you're confident that you can and want to solo them without sleep bolts.  A good example would be VT Ul'Phaubos (sharks).  You may or may not need to sleep them, depending on skill and swaps.  If poison is a problem, there are fortunately a lot of nice solo/tanking dagger alternatives now.  There's Ermine's Tail from MMM, Azoth from Ixion, and a tier2 ZNM dagger with evasion as well.
Edited December 30th, 2008


A personal update
I haven't been doing much with forum community lately, as far as testing/analysis go.  There are a few things that I've done already, or have been meaning to do but simply haven't bothered posting about it yet.  If I come up with anything worthwhile I'm sure I'll write on it.

IRL, I've recently graduated with a BA from the University of Florida, entered the work force, and realized that I'm going to need to continue my education again at some point soon.  I've done a lot of little things in game, though, in the past three months since I got Mandau.  I got my alt character to Rdm 75.  Still working on missions and misc stuff for her.  I attempted a Genbu solo Thf/Nin, which was clearly failing so it ended up being a Rdm/Whm + Thf/Nin duo.  I messed around with Zeni NM's a lot.  Worked on assaults, and almost have Captain rank.  I attempted hard-core, 5-6 day a week, 6-8 man salvages for a little over a month.  Despite doubters, when I go back to doing salvage, I intend to keep it almost always 6-man.

I also led my LS's einherjar events for about a month, but we stopped doing it when I realized that we simply can't low man Einherjar the way I'd intended, meaning that with the numbers in our LS we won't be able to do Odin cycles.  I'm unsure about upgrading Vajra, simply because I don't know if the method I used to fund Mandau will be repeatable to the extent which I'd need it to be (as if that doesn't sound cryptic enough).  Gearwise, I'm trying to focus on getting my Love Torque and Assassin's Armlets (lol), and have been totally lazy about making gil since stage5 and am basically broke.

Most recently, I've been finishing off much needed subs (drk, sam, dnc, sch, and blm for Shamaya, and nin, drk, blm, and sch for Tepic) and almost have Rng up as my 3rd 75.  My fourth will most likely be Blm, that also being Tepic's 2nd 75 to-be.  I've been taking advantage of the new level sync system, and dual or triple box, it's absolutely broken if you continue to spam Korroloka Tunnel -> Qufim Island, and repeat.  Broken, I tell you.  Maybe after I ungimp myself I'll do something a bit more interesting.

Last edited December 30th, 2008

Hurrah.  Moar2come.

Two months ago, I decided to do a little TH testing.  As my first test, I thought it'd be best to see if TH was cumulative--that is, to see if adding more thieves to the alliance would increase drops.  I've talked to a number of people about the topic, and most seem to be quite sure of their conviction, either quite sure that it is the case, or quite sure that it is not the case.  When asked how they were so sure, they could cite scanty anecdotal evidence, at best, to support their claims.  I found that among those that I talked to, as well as on forums, there was no clear consensus on the issue.  So I decided to do this to amend that.  I also thought that the results would be of some practical value.  If the test showed that TH  is cumulative, LS's could ensure higher drop rates by bringing more thieves.  If it showed that TH was not cumulative, LS's could save themselves from the trouble of the old "make sure to throw a thief in every party."  So in absence of an official answer and a consensus, and with a practical value, I decided to do this. 

Further, despite recent commentaries by SE, it was still unclear to some (though not so much to me) as to whether or not the party with the thief in it had to get the killshot in order to ensure TH's activation, so I figured I'd also test that while I was at it.

To those reading this for the first time, this post is a follow-up.  You can read a lot about this in my (first entry - link).  Also, as a background for those that are interested in drop-rates on these mobs with no TH, you can check out Shulula's LJ.  The reason I wanted to follow up was, firstly, because I felt that the initial sample sizes might have been too low.  Secondly, because the drop-rates in each phase weren't as close to equal as I'd have liked for confidence purposes.  Thirdly, because I simply felt that one batch of samples taken on a single day wasn't enough.  In any case, I'm confident enough now to stand by these results and make them public.

My test included 3 phases: 1.Two parties, 1 thief, the party with TH getting the kill.  2.Two parties, 1 thief, party without TH getting the kill.  3.Two parties, 2 thieves.
The first set of bars on this graph is a combination of the first two aforementioned phases.  Looking at the total drops, they're quite close, with the test phase including 2 thieves showing lower drops.  This isn't because adding more TH decreases drops, but simply because the samples, though I think they're sufficient, weren't large enough to make the drop-rates completely equal.
In these next two graphs, you see the 3 phases separately.  The point of the first graph was to see if perhaps adding extra TH might increase the chance of the mobs dropping multiple items as opposed to a pure increase in drop-rates.  The second graph looks at actual drop rates.  The sample sizes taken were 406, 167, and 310 in the order shown on the graph.  Thus, the total sample of mobs killed with 1 thief in the alliance was 573, and 310 for 2 thieves.

The sample sizes are unequal, but I believe them to be sufficient--if there was any sort of cumulative TH effect worth a salt, it should have shown up in samples of this size.  Further, previous samples still pointed to non-cumulative TH, with further samples only enforcing this trend.  I conclude that -

• TH does not appear to be cumulative; If it is, the cumulative effect is so small as to be hardly measurable and hardly worth the effort o maintain.
• By non-cumulative, it does not increase drop rates, nor the chance of multi-drops.
• The party with TH in it does not need to get the kill.

I imagine this is the case in dynamis, etc, as well, where people outside a given alliance can attack the same mob.  I'd like to get more samples, maybe equal them out a bit, but I'm confident enough in these results, and unfortunately its unlikely I'll have time to continue this further.  If time permits, I'll do a test on something else, such as whether or not the thief leaving the party removes Treasure Hunter, or something otherwise non-TH related.

Questions/comments/criticism are very welcome.  Hit me up any time.
-Sham

Additional, more specific data on drop-rates and drop-events for the various phases can be found in this photobucket.
Raw Data (Excel Docs): 1, 2

Sir VTEC is a cry baby.

Some people have been interested in some of the dats I use.  Some like them a lot, others think they look terrible.  Here are the dats I'm currently using, and links to where I got them, since it's important to note the hard work of the creators.














And yes, I love blue.

After a lot of juggling in my schedule, everything for this is finally complete.  There's plenty enough to say about the test, if anyone has questions.  But I'll try to keep it brief here and get straight to the meat of it: appealing to the best evidence, (1)it appears that TH does not stack within alliances.  (2)Further (although not the main goal of the test), it also appears that the party which includes the TH does not need to get the kill shot.

Excel file containing all the data.

Data on Proportions
TH2x1, TH0 Kill - Drop Types
TH2x1, TH0 Kill - Drop Events
TH2x1, TH2 Kill - Drop Types
TH2x1, TH2 Kill - Drop Events
TH2x2 - Drop Types
TH2x2 - Drop Events
Drop Type: IE beehive chip, honey, etc.
Drop Event: IE no-drops, single drop, double drops, etc.

Conditions Target Mob: Death Jacket; Yuhtunga Jungle Moon: 40%->55% First Quarter Moon (Must be noted that a small portion of the first phase was in 38%WXC) 3 Test Phases: 1xTH2(TH0 Kill), 1xTH2(TH2 Kill), & 2xTH2. 2 Parties, 1 person each: 75Thf + 75Mnk, 75Thf + 75Brd, 75Thf + 58Thf Each PC hit each the mob at least once. Unknowns and Possible Sources of Error Game-day Sample size Human error Time of day, job level difference, direction faced, and other voodoo nonsense that shouldn't matter. Need input here, anything I'm forgetting?

MethodSE has recently made the following commentaries, which have stimulated a lot of TH talk recently.  Here's what they said:

Sources: (1), (2)

Everyone wanted to know the answer to the question in the first quotes, but it's also obvious that players also wanted to know if TH had any sort of cumulative stacking effect within an alliance.  I take it that the intent of the question in the first quote was to get an answer in the context of party and alliance, but notoriously vague SE only used the word 'party' in their response.  For this reason, the multi-TH stacking within an alliance has been a matter of debate in the past month since the first fan-festival commentary.  Without clarity, the best that we could do was discuss or go and test.

So I went ahead and, with input from several friends and Thf forum members at KI and Allah', brainstormed up a test and ran with it.  Thank you all for your valuable input, you know who you are.  Although I should say that I'm not going to make this public on forums until deciding on whether or not I need to enlarge my samples, and if so, getting those samples.  So if you're reading this, try to keep it all under-wraps for awhile.

The best conclusion I can give after doing these tests is that it appears, to the best analyzed evidence, that TH does not stack within an alliance (and most likely also multiple alliances, as in Dynamis).  This further supports the whole "debuff theory" idea that people have discussed at length on the forums, as opposed to the "party buff" theory.  While the bar graph in the first pic above shows declining drop rates when the phase went from 1xTH (TH0 Kill) to 1xTH (TH2 Kill) and to 2xTH, common sense would have it that this downward trend is due to small sample size, and that if the sample were infinitely big, the drop rates should be equal.  However, since they did decline, it seems evident that there is no correlation with stacked-TH and increased drops, and thus stacking TH in an alliance does not improve drops.

Other thieves reading this may feel like QQ'ing.  Feel free to.  These kinds of results are not the kind we'd like for our job class.  For others, these results should be pleasing.  This means that there is less of a burden on alliances.  All that needs to be done is to have one thief in the alliance.  However, some will say that, since this proves the debuff theory, this means that a thief can be invited in to hit the mob and then thereafter kicked.  This is a possibility, but the debuff theory is not proved here.  Even if SE stated that if the thief dies, TH stays in effect, it does not follow that if the thief is leaves the party, TH remains (although to the best of our knowledge it is at least possible).

My recommendation to linkshells is to keep a single thief in the alliance, full-time, or at least near the end of the fight before the mob dies.  For dynamis, while it is theoretically possible that you'd only need one thief in order to "stick TH" on all of the mobs, I don't think it's a good idea to bring only one thief to dynamis.  It's a difficult task for a thief to land TH on all mobs, even if he's not pulling.  If the thief is pulling (as they should be), they'll have even less time to do this.  I think at least 3 thieves for a dynamis is advisable, just to make sure that TH is on everything, while at the same time ensuring that there is a good supply of pullers and sac-pullers.

I'll go ahead and stop here.  Questions and comments are more than welcome.  If you have questions, comments, or criticism, you can comment here, reply on another forum thread, PM me, /tell me, etc.

I'm not generally the type to keep track of the lootz and such, and the general purpose of this LJ is not meant to be a blog of accomplishments or a diary of sorts.  But I figure that since these last two weeks in the game have been pretty big for me, I'll just post a sum of most of the loots I obtained recently.

It may have been a bad choice, but I kinda went nuts with some of the gil I've been bringin in for relic.  Well, I suppose when I get much closer, I can bazaar/sell to get that last bit of currency.  Estimating, I think it'll be this upcoming July, at the very earliest, that I'll have enough currency for Mandau.  I want it so bad :(

Will begin to post here shortly.  My main aim of these tests is to see whether or not stacking thieves in multiple parties within the alliance(s) has any effect on drop-rates.  I was out for many hours this weekend and killed a total of ~435 mobs or so for this.  I have all of the data currently analyzed, but I won't post it here until I make the appropriate charts and graphs.  And I won't release it publicly until I do more testing.  More to come, cheerz ^_-