TH testing: final results

[Masamune]

Placebo, eyeball etc.

k so should i translate this lolanswer as "no, nothing have been tested and there are not a single reason for, like if we f?cking cared, gtfo or test yurself"

If i'm correct, then i would remind that most poeple care about TH mostly for Rare items, which are usually dropped from unsoloable mobs.
All theories debunked in this thread (good job on the effort btw) were done SOLO, hence my question.

I really wish to see some evidence about that TH killing blow myth bringing nearly NOTHING to droprate enhancement.

[Meian]

I really wish to see some evidence about that TH killing blow myth bringing nearly NOTHING to droprate enhancement.

Didn't SE say killing blow by THF does nothing?

[Enedin]

k so should i translate this lolanswer as "no, nothing have been tested and there are not a single reason for, like if we f?cking cared, gtfo or test yurself"

If i'm correct, then i would remind that most poeple care about TH mostly for Rare items, which are usually dropped from unsoloable mobs.
All theories debunked in this thread (good job on the effort btw) were done SOLO, hence my question.

I really wish to see some evidence about that TH killing blow myth bringing nearly NOTHING to droprate enhancement.

It's a good theory: it would force THFs to be good at what they do and contribute in a way that is useful. I doubt that SE would go through the hassle of coding this though. But since you have my solo results, you can test the same mobs with 2 people, and see if the results differ.

SE has kept on giving clues about how TH works, even giving us a second trait of TH just to prove it's there. Then they proceed by giving information which trashed a few hard rumours, such as TH stacking or overwriting, or the final blow.

To answer to the SOLO topic... It most likely doesn't matter, since the highest TH trait is used (SE). I just gave out ALL information about the testing method that was consistent across all tests (like, the solo killing), but again, most likely it doesn't matter whether you're solo or not.

[Masamune]

kinda already "tested" (no parsing but very clear eyeballing) while farming my euvhi organs with Brutala(BLM):
- When BLM solo-nuked ~ 90% of Euvhi's HP, then i come in to finish it (SA WS or whatever even with an Accomplice), drops (HQ or NQ organs, or cluster) were kinda meh.
- When i pulled + soloed ~ 60% of Euvhi's HP, then BLM comes in to nuke finish it (+ me accomplice or Collabo), drops were clearly better.
...that over the course of 2 full afternoons.

... but then again one would call BS on such test since no numbers...

EDIT: i remembered another test done between a SMN and a THF @ Crawler's Nest on Lizards (i don't remember tough where did i read that test...), leading to same conclusion as above.

Another example would be my Askar Head and Macha Slops: i litterally ended up tanking NM when they dropped (6 men PT). Such occurences is very rare given fact that usually tanks get the most hate over me on most NMs...

The opposite is clearly visible @ Salvage when usually poeple puts the THF solo in a 2nd pt... no buffs and can't collabo/accomp. This is even more "feel-able" on skellies where mnks owns easily THF's hate (hi2u arrapago shiity cell droprate).

My personal opinion on TH topic is that when you are :
1: soloing
and
2: on a mob lower lvl than yours,
then your TH effects are applied at full, ie capped. In that case, all results posted here applies.
BUT when one of the 2 conditions above isnot true (or both usually), results goes to floored effects if THF's hate at ToD is shit.
In any case, theory saying TH as a modifier% applied to each "drop slot" each mob have is correct IMHO.
Again, this is my personal opinion.

[S N K]

I really wish to see some evidence about that TH killing blow myth bringing nearly NOTHING to droprate enhancement.

Some people still believe this bullshit even now. All I remember was SE telling people that TH is active as long as the THF is on the hate list. The THF could die, homepoint, change jobs and come back but as long as the mob doesn't go passive then whatever TH it was hit with will remain active.

I used to remember someone here seriously thinking a THF's Enmity effected Treasure Hunter as well...

[noodles355]

They've started doing drops based on how much the player did, like in WoE, Abyssea etc, so it could be possible they've done something similar with TH for the past years and had it hidden...

That said, I seriously doubt it. Although god, I would love it if it was true and the Thief had to be good to increase drop rate. It would mean my main job couldn't be replaced by a dual-boxed mule naked with armlets/knife.

[Shamaya]

What I suggested earlier (this isn't actually my original idea, I think Shamaya was the first one to suggest it in a thread long ago, dunno though) was that the notion of "no drops" or "some drops" does hold true. //
This is rare though as most mobs have synthesis item pools that can support more than just one item. In these cases, both TH1 and TH2 have some effect that results in more multiple drops, although there isn't sufficient data to say how this works. This would explain the inconsistencies in Enedin's drop rates (where the total number of items coming off a mob with a multiple drop loot pool goes up by more than the expected 50%.)

I'm actually not sure what exactly you mean about 'notion' here, but giving credit where due, Shulula (similar name as mine of course, and on same server) was the first to distinguish between a mob dropping absolutely nothing, and dropping something/anything as part of their data. I cited them / used this in my thing as well. Btw, I don't think you posted on the thief forums about your effort. You should :3 (though, saying that, Enedin did the same and didn't get much help).

That second point about TH having a large effect on creating "multiple drops" in the treasure pool. Important to note that I've never farmed and I never tested TH as far as the potency of it goes so this is complete hearsay, but I've heard a lot of people that farm say this. Namely I remember, that if you kill sheep/rams in Konschtat, you'll get double skins or double skins and more at ridiculously improved rates. Or that you essentially couldn't get double skins without TH.

The level has to be suitable. Higher level mobs drop more items, but take longer to kill.

Just wanted to know if you had any confirmation of this. You don't mean that drop rate decreases as you get higher in level with respect to the mob, right? Just that lower level mobs in general have smaller pools, right?

Just one question: any tests done regarding droprates and THF's Emnity in PT/alliance? // I really wish to see some evidence about that TH killing blow myth bringing nearly NOTHING to droprate enhancement.

Hey Masa. Don't worry, because SE has commented on these very things. Not directed at you but moreso at about 50% of the ff population,

VOODOO NONSENSE ->

I only stress this because I've seen these same questions asked over and over again for years. If you ever have any thought-provoking questions about the game mechanics and seek clarification, wiki is a great place to start.
-As others have mentioned, killshot doesn't matter
-Enmity levels make no difference
-TH does not have any effect on treasure coffers
-TH has an effect on 'rare' & 'r/ex' drops
-& much much more

To expand on what Viewpoint was saying, here is concisely what we know about activating TH:

"Treasure hunter 'comes into play' when:
(1)The thief (or /thf) produces enmity on the mob in question
(2)The thief performs an action on that target mob"

Pretty much ver batim from SE, though it is unknown if conditions (1)&(2) are supposed to be concurrent or can be separate, but my guess is concurrent (i.e. engaging mid fight, missing an attack (0 enmity), then /whm curing an ally, and disengaging and afk'ing, does not activate TH).

Viewpoint takes it one step further and mentions that the thief could die, HP, change jobs, drop part, and log off, and TH would remain unless the mob cleared its hate list and regened a tick. He's most likely right. I won't say 100% due to some of the vague words SE uses (i.e. 'comes into play' as opposed to 'TH is incorporated into the calculation of drops at time X'). Better safe than sorry, but he's probably right.

* * *

There are a lot of good facts in that wiki article. But one I'd like to bring up briefly before finishing, comes from a really old interview. Like 2003 pre-NA release old. But in it, it says,

Does having a thief increase the possibility of a drop even in key/quest items (aprons, skull, wadi grass, etc.)? Of course! Only the thief has the ability to acquire the Treasure Hunter job ability. There is an increased chance in obtaining dropped items.

The link to this can be found here.

-I don't exactly know what 'wadi grass', but I do know that the Crab Apron is a r/ex drop.

-As for the "key/quest items", I must confess that I edited into the otherwiki that TH affects key item drop rates.
That's how I read it back when. But looking at it now, it probably just means chest keys.
So if anyone has been depending on TH for VNM abyssite conversions because of this, I apologize Unless I'm the only one that thinks they're talking about chest keys now.

[Masamune]

Some people still believe this bullshit even now.

Indeed very sadly.... see poster just below your post confirms also...

All I remember was SE telling people that TH is active as long as the THF is on the hate list. The THF could die, homepoint, change jobs and come back but as long as the mob doesn't go passive then whatever TH it was hit with will remain active.

My point is not exactly that, there is a difference between remaining active with shitty TH potency, and active with strong TH potency.

Sadly i don't have a 2nd char to test more thoroughly, and honestly not really my field of interest (i'm more oriented damage testing). Just adding my 2cents at this topic hoping someone would debunk the totally unparsed TH in PT/alliance.

EDIT: lol Sham, posted at same time XD
Anyways, do you have a link to tests or SEannoucement to this claim "-Enmity levels make no difference" please ?

[Mojo]

Most people were aware of what you meant regarding potency changing with a THF's hate but most people also recognize how fucking retarded the idea is in the first place. Don't derail a potentially useful thread by presenting theories that have no factual foundation ("What you've noticed" does not qualify as this.) If you really think there is some merit to your theory, then perhaps you could come up with some evidence first instead of asking people to waste their time proving your theory wrong.

[Shamaya]

Clarifying how TH activates is a slight de-rail since the thread is talking about potency. I can say, Masa, that while SE has never talked specifically about potency, if you read through all of their commentaries it gives a very strong impression that it's only activating it that is crucial. I'm not gonna say you're flat out 100% wrong, but while they never used words like 'potency', I think it's a good idea to read the SE commentaries first (they're linked on wiki discussion page). Won't take long to read.

Also in addition to what Viewpoint said about TH and "the hate list." SE doesn't specifically say "hate list." That's a player-invented term, I believe. Misconceptions can occur through use of such a term. Because if it were true, zoning or dying would remove you from the 'hate list' in the Kaeko sense,

Spoiler: show

* The "CE/VE Enmity Reset" (most call this a full hate reset") is caused by only 1 situation - the mob must attempt to target you and find you unavailable (this can be for many reasons such as K.O. status, logging off, zoning, Goblin Footprint, etc.). This also means that the mob must first attempt to hit you, however, so simply dying or zoning does nothing but give you the "hate list" reset, not the "CE/VE value" reset. In some situations involving slept mobs or linked hate, it will not even grant you the "hate list" reset.

. In addition as just quoted, Kaeko mentions a specific situation where, if and only if, the monster attempts to target you and finds you unavailable, you will incur a "full hate reset." If we're to believe that the thief either/or needs to maintain a spot on the hate list, or maintain some TE value above zero for TH to remain active, then we might be falsely lead to believe that there are situations where TH becomes inactive. I.E. if the thief is at the top of the hate list and dies, or if the thief is duo'ing and dies, and his partner bios the mob and zones. Idk if these satisfy Kaeko's conditions above, but TLDR: SE has always been very vague. Best to be very careful interpreting what they say (if you're 'OCD' like me).

[Masamune]

Most people were aware of what you meant regarding potency changing with a THF's hate but most people also recognize how fucking retarded the idea is in the first place. Don't derail a potentially useful thread by presenting theories that have no factual foundation ("What you've noticed" does not qualify as this.) If you really think there is some merit to your theory, then perhaps you could come up with some evidence first instead of asking people to waste their time proving your theory wrong.

Agreed, that's why i've added some comments on my posts to enphasize how "What you've noticed" does not qualify as evidence. But i'm honestly not thinking it would be a waste of time proving it wrong (or right), since that's what most poeple think about "lolthf" like Shamaya sumed up pretty well:

Not directed at you but moreso at about 50% of the ff population

Also, to clarify more my previous posts werenot talking AT ALL about activating TH effect (is not even the topic Enedin opened) but about the step ahead, ie how to improve TH effect effectively applied.

Regarding the term "hate" used, yes i was thinking "CE/VE", not the general and vague overall hate used when speaking about resetting, since i think pretty much everyone understood that dying zoning or warping = CE/VE @ 0 + removed from mob hate list. (did i write correctly :S).

[Mojo]

I've met people who simply refuse to believe we have good approximations of how weaponskill damage functions. They insist that we have no way of knowing Warrior's won't do more damage on Firesday. It hasn't been tested for specifically, so they are 'right' in a way, but it's still a stupid idea devoid of any rational thought. The same can be said about hate affecting Treasure Hunter. Although SE has never directly stated it, it's been implied that hate doesn't affect Treasure Hunter (from what they've said and from what we know about the rest of the game.) Please stop shitting up this thread.

[Enedin]

Just wanted to know if you had any confirmation of this. You don't mean that drop rate decreases as you get higher in level with respect to the mob, right? Just that lower level mobs in general have smaller pools, right?

Nah, I don't have any confirmation, that was just my feel of it. After testing about 3 different kinds of mobs, I noticed that some items (on higher level mobs) either had:

- a higher chance of dropping (base%+)
- multi-drop (whereas lower level mob didn't)
- a larger drop pool

For lizards, I found this to be pretty much true. It's also a bit true for Leeches. So I guess it's a bit true, it's also pretty logical.

To contribute to the current discussion...

All ideas are welcome, but unless there's some numbersto support the theory, we won't be convinced. Even after killing over 9000 mobs myself, some people are still not convinced XD

[Mojo]

I've got a few questions about your Stirge testing, Enedin.

Which items can Stirges double drop (if they can at all?)
What's the most number of items you received from a Stirge.

[Enedin]

I've got a few questions about your Stirge testing, Enedin.

Which items can Stirges double drop (if they can at all?)
What's the most number of items you received from a Stirge.

<3 Stirges in Ranguemont Pass.

They can't double drop any item. The most common drops were Bat Fang and Bat Wing. They also drop Fiend Blood, but it's been so long ago that I don't remember anything about that. I'm pretty sure I didn't count it because it has such a low drop rate anyway.

So, the most number of items I would have been able to receive was 3, and most likely I got that a few times.

[Mojo]

I figured as much. This could help explain your data. You would get at least a 50% increase on either item (Bat Fang/Bat Wing) with TH1 according to the pool drop rate theory. The other increase comes from the idea that almost every time a double drop proc'd, the item that didn't drop would drop (since the Fiend Blood almost never appeared.) If the effects of TH1 on the double/triple drop rate was high (which it would have to be to account for your results, unless the pool drop rate theory is wrong of course) then this could easily result in much higher total number of items obtained, especially in cases where the single drop rate for a particular item is close to the combined single drop rates for all of the other items and doubly especially if, in addition to this, that there were only two possible items in the pool (or approximately two, since Fiend Blood is so rare.) I hope that makes sense...

[S N K]

Here's another thought, Why do so many fucking THF's think that not mainhanding a Thief's Knife won't give Treasure Hunter?

The amount of THFs I've seen mainhanding that knife with a Blau or higher base DMG Dagger in their offhand makes me really wish SE hadn't nerfed Sozu Rogberry because some Thief's obviously do not want to listen.

The only real noticeable thing I ever saw from having TH4 was farming bats in Ifrit's for Dragon Blood. There was a slight and I mean slightly higher droprate on Dragon Blood then when I only had TH2 at the time my THF was 75. I ended up with a lot more Wings tho then anything.

[Byrthnoth]

So, I guess it's worthwhile to look at potential models for how TH works. I'm going to lay out the theory from solid part to shaky part as I understand it, please correct me.

1) Monsters have up to 10 drop "slots" like BCNMs, that have chances to be certain things.
Ex. http://wiki.ffxiclopedia.org/wiki/The_Wyrmking_Descends

2a) Treasure Hunter positively influences these drop rates by some mechanism as long as the Thief has performed an action/is on the monster's hate list. However, it is not capable of adding slots.

2b) Treasure Hunter positively influences these drop rates by some mechanism as long as the Thief has performed an action/is on the monster's hate list and can add slots. Aka, if 3 lizard skins is the max you can get off CN Lizzies with TH0, you could maybe get 4 from them with TH2.

-- Now it gets more murky, there are several options for how TH could increase drop rates within slots --

3a) The odds of a drop increases statically, so like +20% or something. -- This seems unlikely. TH4 Thieves still deal with 1% drop rates.

3b) The odds of a drop increase proportionally to the original drop rate, so drop rate*1.5 or something.

Assuming it isn't the first option (because all items in the game don't seem to get a minimum drop-rate of a set amount when a TH Thief hits it, crap drop rates are still crap drop rates.)

4a) If you fail to get a drop in one of the slots, that slot is re-rolled a maximum of X times where X is your TH level.

4b) If the monster fails to drop anything, (potentially barring crystals and seals), all the slots are re-rolled.

Anyone see anything to add/change? Please copy this and modify it as you see fit.

[Shamaya]

I kind of like the idea of a possible slots/rolls theory of drop rates for all mobs now. Bryth's explanation is for some reason much more believable than others I've read. I've seen at least two good proposals for how TH might affect drops in this thread. And as unfortunate as it would be to still not have an answer on the potency and exact means of efficacy that treasure hunter has, I think the '+50%, +100%' conclusion does look unfounded, so far.

But as is obvious and as people have been saying, what we do need is a sufficiently large sample. If it is truly sufficient, then it would seem that discovering how TH operates is only limited by our imagination/analysis; the correct model should be discoverable.

Unfortunately since it seems that the "slot hypothesis" is a possibility, we're in a bit of trouble. I wonder that it may be possible, using our current data, to disprove a "slot-based" TH theory in favor of a theory where drops are improved by static percentages alone (i.e. "PostTH-DropRate = PreTH-DropRate*1.5" or +50%). But I find it unlikely that the slot-based theory could be proven false, because..

Our data samples (Enedin's, and most others I've seen) lack recording for the variables that we would need to verify (and alternatively, to disprove) a slot-theory; namely:

-"No drop" rate
-All possible treasure-pool compositions

I believe Shulula was the first to record drops in this way. As an example, let's say we're fighting Wild Kobayashi's. Kobayashi can drop the Kobayashi Hot Dog, the Kobayashi Ketchup Packet, the Kobayashi Plastic Cup, and the rare quest-item "Kobayashi's Dinner Napkin." At first we might think it's all well and good to record only the number of kills and the drop% of each individual item. But if we'd been prudent and recorded all possible drop pool combinations, we'd find that Kobayashi can drop the hot dog, ketchup packet, and plastic cup at the same time. But, never can he drop the plastic cup and the coveted dinner napkin at the same time. Also he can drop up to two hot dogs or ketchup packets per kill, but never three. And never do you see a single ketchup pack when you get 2 ball-park willies, and vice versa.

I'm not even sure how to connect that to what I said before and what Byrth and others have been saying. But TLDR: We should record all possible treasure pool compositions, or the correct model for TH and drop rates in general may never be found.

Thorny should be postin his big sample some time soon as well, hopefully. But I'm not so sure if Kparser keeps logs of drops on a fight-by-fight basis when you save a parse. If not, we may find ourselves with insufficient data. If it does, though, I believe all 9k+ of those fights would have to be hand-parsed 1-by-1 in order to list all the different pool compositions.

* * *
PS: If anyone's interested in looking at Divisortheory's old TH+1=Drops+1% test, you can check it out here.

[Enedin]

3b) The odds of a drop increase proportionally to the original drop rate, so drop rate*1.5 or something.

That's how I think it works. But I think it's either 3B or 4A, and not both.

4a) If you fail to get a drop in one of the slots, that slot is re-rolled a maximum of X times where X is your TH level.

This is what Miraun deducted from my results, and I support this theory. Not sure if you've seen the spreadsheet he made on page 1/2 somewhere, but if you input my numbers, on most mobs it's very close, and on all mobs it's within the error margin. So I'm very much leaning towards this.

4b) If the monster fails to drop anything, (potentially barring crystals and seals), all the slots are re-rolled.

Assuming this is how it works, it only happens once, right? Otherwise a mob always drops something XD Tbh, I don't believe this is true (but as I said earlier, all ideas are welcome). I just want to know why you think this is true, maybe I missed a post in here.