For The Shinies! A Guide For Thief

Again you're making assumptions and pulling numbers out of a hat. "should easily be 50% more" based off what? Gut instinct? I agree that cento should have a higher weaponskill frequency, but quantifying the difference by guessing isn't a good way to represent it.

Lol who said im guessing anything? I actually almost never guess, I use math most of the time, if thats possible.

Depends on exact samurai roll tp return on twashtar/cento rudra will be around 165, avg tp per round will be around 430. Capped delay for that combo is 72.4
I could assume 2 rounds on avg to get above 1000, but lets assume 2.5 rounds because sometimes you will need more with bad luck on multi attack proc. So with ws delay whole cycle would be around 5 sec.

Now Twashtar/Sari tp return will be around 188. Tp per round around 450, capped delay 77.4
To get into 2000 tp range you will need on avg 4 rounds, but sometimes more. Lets assume 4.5 rounds. With 2 sec ja delay whole cycle would be around 7.8 sec

7.8 / 5 = 1.56

There you go 56% more WSs with Cento with avg 2.5 rounds to 1000+ vs Sari 4.5 rounds to 2000+

Does wearing enmity gear when using Collaborator or Accomplice do anything? I haven't seen anything about it since it's just hate transfer but I wanted to be sure since it feels like Collaborator barely does anything even with Empyrean Head. Really wish Accomplice and Collaborator didn't share a timer...

those 2 ja's just take enmity from someone else. wearing extra wont do anything. it's just so that you can either do it to save someone or take it so you can dump it on the tank with a ta.

Well I know their function, I just wanted to make sure Enmity gear didn't like multiply the hate you took or something which would have been neat.

No, enmity gear does not affect collaborator or accomplice. However, our empyrean hat does. The base raider's bonnet does nothing, but the +1 and +2 both increase the enmity stolen by 10%, and the Ilvl versions raise that to 12% for the Ilvl 109, and 15% for the Ilvl 119 version. I carry my skulker's bonnet around just for that. Stealing 40% enmity every minute with collaborater, or 65% every 5 minutes with accomplice is a significant upgrade.

I find one of the best uses for collaborater to be in dynamis whenever we're fighting a beastmaster NM by the way. I like to steal my group's dark knight's hate before charm goes off, and since I sub /dragoon I can high jump the enmity away so I don't get charmed myself. It always prevents a couple deaths by ensuring the tank gets charmed rather than our best damage dealers. The beastmaster NM will die shortly thereafter, but even a single hit or two from a fully buffed damage dealer can fell almost any non tank in dynamis, especially if it's a critical or aftermath proc.

What would yield more TP on THF? 2% Triple Attack and 2% Quadruple Attack? Or 2% Triple Attack and 5% Double Attack?

I know Double Attack gets depreciated the more Triple Attack you have, and THF gets A LOT of Triple Attack. But 5% is actually still a decent amount, so I'm kind of torn.

Not to beat a dead horse into the ground... but if this is really the case and getting TH procs is your main priority, why not just offhand a Gandring? If on more of a budget, there's also Plunderer's Knife or even Skinflayer (I have a TH+2 Skinflayer that I sometimes use for max TH scenarios). You're likely not optimal DPS with any TH offhand, so... go for the best TH if proc ups are your priority? And it's not as if Gandring/Plunderer's are terrible offhands. Nor should it really matter to you anyway if the mob lives a little longer, if your stated goal is to get TH10+ (more time for procs, yo).

If what you care about hitting the baseline TH8 and then maximizing damage (this tends to be my main priority the vast majority of the time), there's no real need to get TH from your weapon slots. There's likely a better DPS weapon combo for your situation, be it Twash/Cento, Tauret/Shijo, Twash/Tauret, etc. So use those, tag the damn mob once with TH8+ gear, then swap back to better DPS gear and be done with it.

Gandring is a no-no because it costs a gazilion of gil.
Skinflayer TH+2 could be ok but: 1) I can't afford to use DM augs on dagger, need too many other things first, 2) Inventory+1! (whereas I already have many other things/daggers from other jobs, and Sandung from when it was actually a pretty nice option, before everything else got released)

I do have a TH14 set in TP, I'm still a bit unconvinced about using it, I'll be willing to change my mind when/if further tests showing it truly has an effect come out.

Atm I have 2x TH8 TP sets
One is: "Max DPS with TH8"
The other is: "Lotsa DT/Meva with TH8"

Clearly I use the second when I'm in situations where I'm gonna take damage and/or I don't have real dedicated support. The former is for when I want to do some damage but I often question that approach myself.
Given how DPS is something we really don't lack in my group 99% of the times, shouldn't I be actually striving to REDUCE my DPS so that the monsters last longer, hence giving me more TH level up proc chances?
Eh, I don't have an answer.
To be fair I think whatever I choose nothing is really gonna change though, given I'm doing group content with many other DDs. Reducing/raising my DPS would be kinda unnoticeable either way I'm afraid.

That seems to be the concensus.

1) Gearing for TH will NOT maximize your THF DPS.
2) Depending on content; tpbonus offhand is best.


.. this logic applies to any job that gears utility > dps right?

Either way, I found Madranta's THF spreadsheet which was moderately updated.
I added new gear, Tauret (with 50% Evis damage) and ran some tests.

Target: Apex Bat l36
A: ~3200 away from att cap
B: Att capped
Both sets are haste and acc capped.
Both sets using Red Curry Bun
No special rolls (Stp, DA, Crit etc)

Scenario: Spam WS as soon as possible.
WS used is Rudra for SA/TA.
Unstacked it's Rudra for all except for Tauret, where it's obviously Evisceration.

Results:
Note: to give Shijo a slight edge I swapped TP set belt in the tests with Shijo. It's Reiki Yotai for all other weapons, Windbuffet +1 for Shijo.


Now for some lenghty comments:
This test is a bit unrealistic for the same reasons Melphina explained, which I found to be true for myself as well.
It's very rare to be able to WS exactely at 1000TP when you offhand Centovente.
This is even more true if you have Regain in idle, Adloquium from SCH and/or Regain roll from COR.
I experienced this on my DNC when I used Twash/Centovente.
It was still quite awesome and ahead of other options for me, but the difference was clearly smaller than what the spreadsheet nubmers suggest.

Furthermore we need to discuss the difference between Sari and Tauret offhands, which are very very close with pretty much with any MH.
Tari tipically gives better WS numbers.
Tauret compensates lower WS numbers with the lower delay (and respectable white damage, the base value is the same!)
I think this means if you find yourself doing content where you have some form of regain and pauses between pulls, Sari will probably pull ahead of Tauret even if it's slightly behind.

A similar thing can be said about Twashtar main hand.
I tested with AM3, AM1 and no AM.
AM3 will have a very high uptime but it's not realistical to assume it's gonna be 100% on.
AM1 will have almost 100% on instead since you can overwrite it with any other AM level.
AM0 is not very realistic either because you're always gonna have some AM at least 75% of the time, imho.
Regardless of this premise, Twashtar builds clearly gain a lot from their white damage and the OTD procs.
So, again, if you're doing content where you don't melee a lot, Twashtar might fall behind other options in terms of damage.

Tauret is of course spectacular in low buff situations because of how crits work.
I'm a bit confused to see Shijo behind Sari though. I suspect this might be because I was already at 100% crit rate and hence the 5% crit on Shijo doesn't get counted? I dunno, but this sounds a bit strange.
To be fair I was expecting Shijo to be ahead of Twashtar OH as well, not behind, but whatever...
For Tauret tests I also simulated it's TP>Crit effect artificially by adding a 20% average crit rate to the melee sets (didn't add to the WS sets)



To conclude
Centovente is a force to be reckoned with. Granted I feel the "real" difference you're gonna see in-game is clearly much smaller than what the spreadsheet suggests, but it's still gonna be hot.
For all other combinations, I don't think the OH makes such a huge difference honestly, except for Twashtar OH for Aeneas and Tauret maybe, that's clearly quite ahead of other options.
Another interesting thing, but we knew that already, is that the difference between R0 and R15 Aeneas is truly truly small. Not sure I'd call it worth it the 250m of gil it requires.
Clearly it becomes more relevant in a scenario where you need acc (in my tests acc was capped by default)

To answer myself: What would be the best MH to use if you're stuck on Perf Sari OH?
The difference is small and hardly relevant.
Twashtar is gonna win if you melee a lot and make use of the Aftermath.
Tauret is gonna win if you're really underbuffed (For instance I'm not even in the DD pt when I do Dynamis, most of the time, so I guess this should be my choice?)
Aeneas is not really that far behind though.

One last note: this test is a "WS spam" one, which I don't think properly reflects the way I play THF in some circumstances.

For instance many times I find myself holding on TP while I wait for various cooldowns to be ready again.

With this "holding" style of play, which option would win?
That's a good question.
To be fair someone could say: "If that's how you're forced to play for whatever reason, who even cares about DPS?"
And I can't disagree with that, who cares about whatever you choose in such a scenario.

Regardless I think:
1) If you're engaged and meleeing, Twashtar MH builds (any) are gonna win because of the AM impact on white damage
2) If you're not gonna engage or barely, Aeneas MH builds are gonna win

Man where is all the r15 vajra love.

I wasn't interested in Vajra, I apologize D:
I think it holds more potential than most people in this thread think it has, but for pure DPS I'm not sure it's gonna be ahead of the powerheads you can see in the test above.

Furthermore, a proper Vajra test would require to take into account other things:
1) the time to set up AM3 every 3 mins and the X seconds in between re-applications of AM3. Can't abuse despoil all of the time for that because it's a 5 mins CD.
2) the TP gear would have to be different, with less emphasis on multiattack and more on STP I suppose
3) for the second style, the one where you melee less and/or hold TP a lot waiting for CDs, Vajra could be nice because of the SA/TA boosts and the AM3 in between.

If that spreadsheet is the same as most spreadsheets, then it doesnt simulate WSing at 1000TP but ASAP. When you set up 1000TP on front page it's MINIMUM that it will WS at. If you go to WS page you have AvgTP there

For example in MNK sheet on Victory Smite setup when I set 1000TP as minimum and 0 TP round overflow (it simulate lags) Im getting 1553.00 TP avg, which is 1303 without moonshade (which is calculated in there). So people arguing that spreadsheet calculates WS usage at exactly 1000TP don't understand how it works.

2% QA is gonna win.

Even ignoring triple attack, Quadruple Attack is base swing plus 3 extra swings. Double attack is base swing plus 1 extra swing. You are going to need 3x the amount of DA in order to match the extra swings from QA.

So even with zero TA, 2%QA > 5%DA. Add in any TA, and the gap just widens.

That summs everything up nicely. Treasure hunter gear always sacrifices on max DPS. The more you equip, the more dps you cut. And whether or not centovente is better than sari or tauret offhand depends on accuracy. If accuracy is capped centovente offhand wins, otherwise the Ilvl accuracy bonus is better.



Vajra is a strong dagger and malignance gear makes it even better. There hasn't been a whole lot of data on vajra since malignance came out, but the general facts are this

--Twashter max damage TP set still gets more swings in than Vajra malignance set, and with empyrean aftermath it's white damage is higher

-- Twashter's sneak or trick attack rudra's are on par with vajra's sneak or trick attack mandalic stabs. Rudra's has stronger dexterity and TP mods than mandalic stab, but mandalic stab has an attack modifier. As long as you're getting buffed correctly from your group twashter has the better weaponskill, and even though vajra boosts man stab by 30% and twashter only boosts rudra's by 10% the end result is a wash.

--Twashter gets better mileage from centovente than vajra for the same reason. Harder TP scaling on its weaponskill.

I actually think Vajra is better than Tauret, and probably on par with aeneas. We just don't mention it much because it's more of a a sidegrade than an actual upgrade, and it's more of a niche weapon. The mythic trials are also a royal pain the the rump, so it's easier to recommend someone build twashter as their first REMA, but that's beside the point. The scaling puts R 15 vajra somewhere in the ballpark of aeneas territory, but Twashter still has the edge. This could change in the future of course. If they scale REMA's up to rank 25 and vajra's boons are significant enough it's posotioned to dethrone twashter as the top dagger. But in their current state it just isn't quite there yet.

Better than Tauret, worse than Twashtar and on par with Aeneas? You are implying that Aeneas is better than Tauret. Sechs post shows they are the same. My own sheet shows Tauret is better. At least for WS spamming.

Mandalic Stab is boosted by 49.5% with R15 Vajra, not 30%.

You compare SA/TA Mandalic with Vajra and SA/TA Rudra with Twashtar, but I think Vajra(MH)/Twashtar(OH) actually has stronger SA/TA Rudra than any combination with Twashtar main hand outside of low TP with Centovente OH.

Main problem with Vajra is same as with every Mythic. It's power is strongly dependent on AM3, so every scenario that force you to lose AM3 uptime is bad. The main advantage of Vajra is obviously Mandalic Stab skillchain property. If you need Fusion to skillchain with someone or want to solo light SC, Vajra will easily outperform any other dagger.

I am not a numbers person. I do however play THF quite abit and have all 3 daggers r15.

Aenaes is in my locker, and i alternate between vajra or twash mian hand depending the mob and skillchain.

I've never gotten twash to hit a SA+ws number like mandalic on some things though. Like WoC I can hit 80-90k SA+mandalics for example.

SA or TA Mandalic Stab simply has superior attack, so if you cant hit about the same SA or TA Rudra with Twashtar as you do SA or TA Mandalic with Vajra, then its an attack problem. They should be about the same.

Now if you go for pure damage, try SA or TA Rudra with Vajra offhanding Twashtar. That should be stronger than with Twashtar in main hand.

EDIT: Actually Vajra/Centovente should also have a little stronger SA or TA Rudra, than Twashtar/Centovente. Using Cento with Twashtar is easier tho, because 70 dex is around 52 accuracy and R15 Vajra only has +30 accuracy.

My two setups are...

vajra/twash and twash/tauret at the moment.

Im usually attack capped but crazier things have happened.

The numbers I provided need to be taken with a grain of salt.
They are likely dependant from the gear I used, which is good but not BiS in every slot.

Either way my numbers show that Tauret (Evis) is better than Aeneas (Rudra), even R15 Aeneas.
The different is really small at capped attack, we could say they're pretty much the same, with this difference being slightly bigger if we're talking about R0 Aeneas.

This difference is much bigger when you're att starved.
Then again in that scenario Evis is better than Rudra even with Aeneas MH (but still worse than Tauret's Evis, of course)



Pretty lame if you ask me, but I guess that's how things are :x
I was honstly more surprised by Twashtar beating Shijo for Evis spamming. As I said before I wonder if that was because I was already at capped Crit or for whatever other reason.

Were you? I had Twashtar OH leading for Tauret Evis too, so you should expect the same :P

Wellllll it's basically 4% crit, 5% Cdmg, 2% TA vs 35 DEX.
I wasn't expecting the latter to beat the former for a crit based WS, but hey! What do I know...

That doesn't surprise me. Tauret's freakin nasty. I believe you.

Probably hit the sweet spot on dDex. From BG:

that there is an extremely significant jump in critical hit rate when dDEX is increased from 40 to 50 (from 4% at dDEX=39 to 15% at dDEX=50), but no further boost to critical hit rate once dDEX is greater than 50.

Shouldnt be the case here. Afaik Tauret/Shijo Evis set has over 355 Dex even with 0 DEX on herculean feet and thats the lowest dex feet available for that set. He used Apx Bats 136 as target and in the sheet it has 303 AGI, so ddex was already capped.

So you can expect Twashtar leading even more if DEX from it would actually help with dDEX.

I think Vajra could actually win.
It has the strongest SA and TA (both with Rudra and Mandalic). Depends how long you wait tho. If you wait to 3000TP on both Twashtar and Vajra then maybe Twashtar AM3 would pull ahead. If you wait to like 2000 on Twashtar then it would be probably like 2300+ on Vajra, which would help Vajra catch up a little at unstacked WS damage and would obviously pull even more ahead for SA/TA.

Adhemar Bonnet +1 (B) vs Pillager's Bonnet +3

I guess the latter wins for 1hit WS like Rudra, but the former should win for Evisceration?

seems legit sechs.

on a different note, i started building a twash a little while ago and then i met glavoid.. is there an easy way to farm him?