Treasure Hunter %

[TacoTaru]

Forgive me if I've missed a previous topic, but a couple quick references of both wikies showed no results, nor did I find what I wanted out of a search. But a few quick questions concering Treasure Hunter.

1. Are the percentage of increase known for TH1 and TH2?

2. If no, can these be tested by farming X number of mobs for Y item without TH, by adding TH1, and TH2 (My logical answer is yes, but if there is some flaw in this I'd like to know).

3. What is a good number of killed enemies to go on for finding a good and credible average for these tests?

4. Is there a parser that parsers drop rate of items? I know Directparse (I think) parses the chat log, but does it parse the actual percentage of the drop rate?

Thanks in advance; I've asked numerous people for solid numbers on this, but most of them still don't believe that the knife and armlets only add 2%, so I figured I'd turn to more reliable people.

[spooky]

I don't think people know this information 100%. You could probably find some reasonable guesses though.
What IS known is that TH doesn't stack and to get the effect the THF must hit the mob (or just get on it's hate list?)
The problem with testing this is, while you could find % for TH0 TH1 TH3 and TH3 while farming somthing like silk thread. I doubt very much that it will have the same effect on say Dynamis mobs or HNM. The only real way to test would be to kill the same number of those as you did for silk thread (we're talking thousands). As you're unlikely to kill that many of the same HNM and record every single drop, with differnt levels of TH, (and oh yeah don't forget moon phase too lol,) it will forever remain a mystery.
Unless SE just out right tells us. >.>

[TacoTaru]

The idea was to generate the numbers based on something such as silk thread, beehive chips, etc. As for the stacking, I remember that mentioned from the SE announcement sometime last year. What I really wanted to learn was the approximate % of increase when going from TH0 > TH1 > TH2. On of the main reasons is to figure out how significant the trait actually is. Eyeballing it you can tell it's noticeable especially in Dynamis and Limbus, but to what extent is what I would like to know.

You said you didn't think the effect would carry over to HNM/Dynamis, but I'm curious as to why? Would it not be safe to assume that the percentage of increase on the trait would be a solid constant, and that the effect does not appear as significant on HNM/Dynamis because the drop rate of the items are so low? I don't know programming or code in the least, but I would think that it would be difficult to make it so that the job trait itself had a varying effect.

On numbers, I thought I remember an old test thread for something else saying ~1000 was a good number, but I was not confident enough to assume this. If I start testing, I want to be as consistent and accurate as possible while still making this as manageable as possible for the time constraints I have. Also seeing as you mentioned the lulz factors I guess it is also reasonable to ask whether or not those factors (day,moon, whatever else bullshit idea there is) should actually be recorded or is it safe to assume that over a large test such as this any changes that might effect it will be a non-issue due to the size?

[Thorny]

If I start testing, I want to be as consistent and accurate as possible while still making this as manageable as possible for the time constraints I have.

I'm not sure your time constraints, but making a consistant and accurate test for treasure hunter would require at the very least 10,000 kills with/without knife in the same game day/moon phase(that's assuming you don't believe there are any other controlled variables to monitor) to have a reasonable margin of error. If you're trying to do a test with 0 controlled variables you might as well just not bother because it lacks credibility unless you have an utterly obscene amount of kills >_>

[TacoTaru]

My time constraints limit me to mostly between 3-5hours a night, roughly 4 days a week. This is assuming that I spend my majority of my time conducting these tests.

The test also has the chance to allow room for clearing up other things. I think it's safe to say that moon phase can be tested in conjunction with an overall test of the job trait in itself. If over the course of all moonphases the percentage of drop is exactly the same, this rules out this theory of changing moon phases affecting the trait all the while building my sample size for the original test size.

The main problem I foresee is the amount of variables that can be linked to treasure hunter. The only one I personally have any qualms with bothering testing is day related drops. Back during inflation I farmed every single vanadiel day in KRT for bat wings, and while it was done completely eyeballed, there was absolutely no reason to believe day effected these drop rates. I don't remember the exact numbers now, but during a set amount of hours, I could accuractly predict a very close amount of how many stacks I would end up with by the time I left. This was at a time that I did heavily monitor day vs. drop rates because at the time I believe that certain days did warrant greater drop rates. And after over 3-4 months of nonstop farming, I was able to almost always able to guess my stacks obtained vs. time spent regardless of day.

I don't want to say day is not a factor at all, but over the course of the time I spent there and the time I spent focused on trying to find the absolute best days to farm, the opportunity never presented itself to show me one way or the other (on a side note, I killed well over 150+ spook pops there as well, and while this is a very shitty sample size, the results for his drops were mirrors of the bat wings). Would it be safe to generate what the community thinks are the staple effects of treasure hunter and work from there, or does this allow severe error even if the data numbers are still high (2k+)

[Darwinion]

I always believed that TH would add extra %age chance of all items dropping. All items have a given %age chance to drop, but some are rarer than others obviously.

A bee hive chip might have a 25% chance to drop from a Hornet for example, but with TH1 that might be boosted by say 15%. But that's not 25+15 = 40%. It's 15% of 25 added on = 28.75%. (Well it could be a flat amount like 25+15=40... who knows? Just seems a flat amount would be a bit too much or else we'd see AF2 drop way more often.)

Of course I'm totally making these figures up. Just using them to illustrate the way I believe it works. TH2 would obviously add more to the original chance. And with Dynamis items having such a piss poor drop rate anyway the added %age from TH adds a negligible amount to it. But I believe it is there. None of this takes into account moon phases etc.

Because of the vast array of different items and their drop rates, plus the same item having different drop rates from different mobs (lizard skins anyone?), plus variable other factors like moon phase, testing this would be an inhuman effort of epic proportions to actually prove. And quite frankly just not worth the effort.

[rath]

Was there not an advanced thread that attempted to do this way back when SE made that announcement? They first did tests to see what actions added TH to a mob, and then tests to see if zoning/dying/logging/outside party kill, removed it. But then they were going to attempt to do exactly what the OP plans to do, I stopped reading it by that point, but I am pretty sure that was the intended next step.

[Max™]

I don't remember where, but for the longest time I thought it was TH1 = Previous rate +10% of that rate (so 10% Drop rate=11%, a.k.a. basically placebo), TH2 = TH1+flat 10% increase, TH3/4 = TH2+flat 1% increase.

I know personally I'd love to see exactly what the TH1 numbers are, because for many things I've suspected I'd do better subbing a more DD sub and just killing X% faster than having X% better chance at drops.

[spooky]

You said you didn't think the effect would carry over to HNM/Dynamis, but I'm curious as to why? Would it not be safe to assume that the percentage of increase on the trait would be a solid constant, and that the effect does not appear as significant on HNM/Dynamis because the drop rate of the items are so low? I don't know programming or code in the least, but I would think that it would be difficult to make it so that the job trait itself had a varying effect.

Actually what I said was that it doubt it will have the same effect. Although it's meaningless to argue about it because you will never get a sample size large enough to test HNM.
TH could be done by adding an increase to the base drop rate, or it could just flat out add an extra chance, or it could do any number of things, you will never know for HNM because you can't kill 10,000 Kirins and give us the answer.

On numbers, I thought I remember an old test thread for something else saying ~1000 was a good number, but I was not confident enough to assume this. If I start testing, I want to be as consistent and accurate as possible while still making this as manageable as possible for the time constraints I have. Also seeing as you mentioned the lulz factors I guess it is also reasonable to ask whether or not those factors (day,moon, whatever else bullshit idea there is) should actually be recorded or is it safe to assume that over a large test such as this any changes that might effect it will be a non-issue due to the size?

If you really were going to do this you would need a sample of more than 1000, idealy something like 10,000 (for each TH level and after you've worked out the base droprate with another 10,000 of course). I would say it was worth recording moon though.
While this might be semi-achievable for beehive chips and the like, I don't see it happening any time soon for HNM. There is a chance you could do it with dynamis mobs in your lifetime I guess.

My point is though, why bother? It will take you fucking ages, people will take note of it and then all your efforts will be quickly forgotton. While it's all very noble to want to contribute something cool like this, it's just not practicle.

Best rule of thumb for TH is try to bring 1 person with the highest TH where it might be worth it. (No they don't need kill shot.)

[rath]

This is the most comeplete % testing I can find, I hope it helps.

http://khoisan.livejournal.com/15302.html

[Sepuku]

Very good informations.
From my own experience with TH or TH4 , the chances rise extremely if you get more people in your party.
I cannot explain why , but the treasure pool is always filled to the max with minimum 3 people.
Also Testimonies and other r/ex stuff drops are much better with additional 2 people in your thf party.
2 person parties do not do a significant raise in your drop chances for items.

Tested @ Labyrinth of Onzozo / Dangruff Wadi / Castle Oztroja / Castle Zvahl Baileys / Castle Zvahl Keep / Gustav Tunnel/ Quicksand Caves.

NMs tend to reliably drop r/ex stuff with a third person in the thf pt.
Regular mobs tend to reliably drop useless gear or crafting items.

tested on all NMs in the above area for a period of 3 months in each area.
examples:
Goblin Enchanters drop Gold ingots more frequently.
Goblin Miners tend to drop Gold Ores more frequently.
Mysticmaker Profblix droprate is at 23/32.
Shamans Cloak NM 8/15.
and so on

With Th4 ( test period 3 months 4-5 hours of killing each day)
Overall droprates on NMs are round about at 66% .
Overall droprates on regular mobs are at 2-3 items per kill.

Sorry i cannot give you exact numbers, but taking people with you will help a lot , if you really need an item to be dropped.

[Sechs]

Since we are in the newbie forum, I guess I will be forgiven for all the heresy I'm about to say.
Little is known about TH, and I'm sure we all wish SE could clarify it once and for all in every little detail, so people stop arguing with stuff like "killing shot" or "THF being in the party of the alliance that kills the NM". All things that might be true of course, but without proof they are just speculation.

About the way TH works, I remember reading a very interesting theory somewhere, a lot of time ago. I don't remember every detail now, but I do remember that back then when I read it, it seemed very likely, and much "SE-way".
In this theory, the guy who created it was supposing every monster has a list of possible drops, every time you kill it a "dice" is rolled (much like in the /random command) internally, according to the result of this dice, you would have received different things.
For example (very stupid example):
0-400: Nothing
300-400: Empress Hairpin
400-500: Damselfly Worm
500-750: Fly Lure
750-999: Insect Wing

Some mobs may have 1 roll, some may have 2, etc. Crystals are something on their own.
According to this guy, TH just "adds" another roll, giving you another chance to see the result being within the range you want it to be.
Of course there are several more variables to handle, like double item drops etc.

I don't remember all the details, but this guy was giving a rational and likely explanation to almost everything. It really convinced me back then when I read it, up to the point that I really thought the real system was something like that, rather than a "raw" % of drop system, that you would normally expect.
Nobody went on theorycrafting further this guy's theory? Or maybe it has been proved completely wrong meanwhile? :3

[sleekmotorwurkz]

There was another post a while ago debating this, rather then digging it up I will just say my response....

I believe that gilfinder might have the same % effect as treasure hunter. Everything is basically set up the same. Job trait, gear that has +1 to effect etc...

My thought was to go out and test on something that has a static gil drop with a decent respawn time (I never actually found anything so I tested on gobbies in Beaucedine - drop isnt static so was hard to tell what the base gil drop was)

Test with gilfinder 0, 1, 2, 3 and see if adding gilfinder 1-3 was a change of 1-3% or 10-30% etc.
Also testing the theory that 1 hit with gilfinder 3 then changing gear to gilfinder 1 would keep the "3" effect on - assisting in proof (even though SE pretty much confirmed) that possibly treasure hunter was a "debuff" and helping understand if the same person can overwrite their own TH4 with TH2 (without gloves and knife) for the kill.

Maybe I will get back around to it one day, nobody could really give me a mob with a static gil drop that was soloable and a reasonable respawn time, and since the gil drops in beaucedine were so erratic I pretty much lost interest.

[Alistaire]

If I start testing, I want to be as consistent and accurate as possible while still making this as manageable as possible for the time constraints I have.

I'm not sure your time constraints, but making a consistant and accurate test for treasure hunter would require at the very least 10,000 kills with/without knife in the same game day/moon phase(that's assuming you don't believe there are any other controlled variables to monitor) to have a reasonable margin of error. If you're trying to do a test with 0 controlled variables you might as well just not bother because it lacks credibility unless you have an utterly obscene amount of kills >_>

Not really. Just need 2 people, 1 with knife and 1 without. Keep pace killing the same types of mobs. If you don't finish your testing in one game day, wait for the next game day of that % moon to continue.

[solsovly]

There was another post a while ago debating this, rather then digging it up I will just say my response....

I believe that gilfinder might have the same % effect as treasure hunter. Everything is basically set up the same. Job trait, gear that has +1 to effect etc...

My thought was to go out and test on something that has a static gil drop with a decent respawn time (I never actually found anything so I tested on gobbies in Beaucedine - drop isnt static so was hard to tell what the base gil drop was)

Test with gilfinder 0, 1, 2, 3 and see if adding gilfinder 1-3 was a change of 1-3% or 10-30% etc.
Also testing the theory that 1 hit with gilfinder 3 then changing gear to gilfinder 1 would keep the "3" effect on - assisting in proof (even though SE pretty much confirmed) that possibly treasure hunter was a "debuff" and helping understand if the same person can overwrite their own TH4 with TH2 (without gloves and knife) for the kill.

Maybe I will get back around to it one day, nobody could really give me a mob with a static gil drop that was soloable and a reasonable respawn time, and since the gil drops in beaucedine were so erratic I pretty much lost interest.

I'm not 100% sure, but I used to solo the NM bard in SSG behind the ornamental door. I *think* he dropped a static amount of gil but it's been literally years since I've killed him. His respawn was something like 15-20mins. Can steal a gold coin and drops a couple decent bard scrolls so it would be decently profitable if you want to run your tests there.

[Genosync]

... this is something slightly odd, but of course I don't really give in to speculation like this, figured I'd just provide the information.

I've only seen NMs in Salvage drop multiple 35s (well, the ones that can, like Madames and Arrapago 5F Chariot) when the THF is in a PT that does have at least 3 people.

So, take that information as you will and go expand on it with testing if you've got the time--it's an interesting thought, but unfortunately I doubt it's law.

[evilpaul]

There was another post a while ago debating this, rather then digging it up I will just say my response....

I believe that gilfinder might have the same % effect as treasure hunter. Everything is basically set up the same. Job trait, gear that has +1 to effect etc...

My thought was to go out and test on something that has a static gil drop with a decent respawn time (I never actually found anything so I tested on gobbies in Beaucedine - drop isnt static so was hard to tell what the base gil drop was)

Test with gilfinder 0, 1, 2, 3 and see if adding gilfinder 1-3 was a change of 1-3% or 10-30% etc.
Also testing the theory that 1 hit with gilfinder 3 then changing gear to gilfinder 1 would keep the "3" effect on - assisting in proof (even though SE pretty much confirmed) that possibly treasure hunter was a "debuff" and helping understand if the same person can overwrite their own TH4 with TH2 (without gloves and knife) for the kill.

Maybe I will get back around to it one day, nobody could really give me a mob with a static gil drop that was soloable and a reasonable respawn time, and since the gil drops in beaucedine were so erratic I pretty much lost interest.

I'm not 100% sure, but I used to solo the NM bard in SSG behind the ornamental door. I *think* he dropped a static amount of gil but it's been literally years since I've killed him. His respawn was something like 15-20mins. Can steal a gold coin and drops a couple decent bard scrolls so it would be decently profitable if you want to run your tests there.

Gilfinder will still work if you fight a mob that's been fought by a person with it who wasn't in your party.

I was going to camp Sozu Sarberry one day and noticed a lack of non-Tonberry mobs in the area. Got over to it's spawn to see a bee fighting it and a BST running by. He sends me a /tell that its slaughtering everything he throws at it, so he's going to zone and I can have it. As he zones a Thief Knife RMT re-claims it with Steal. Gets blasted to low HP by Mijin Gakure, and then finally zones when Jubaku: Ni repeatedly fucks his Utsusemi casts up. I finally claim it, kill, and get my usual Tonberry Coat plus a larger amount of gil than usual. The RMT's TH sucked though, so no katana.

[Vedder]

There was another post a while ago debating this, rather then digging it up I will just say my response....

I believe that gilfinder might have the same % effect as treasure hunter. Everything is basically set up the same. Job trait, gear that has +1 to effect etc...

My thought was to go out and test on something that has a static gil drop with a decent respawn time (I never actually found anything so I tested on gobbies in Beaucedine - drop isnt static so was hard to tell what the base gil drop was)

Test with gilfinder 0, 1, 2, 3 and see if adding gilfinder 1-3 was a change of 1-3% or 10-30% etc.
Also testing the theory that 1 hit with gilfinder 3 then changing gear to gilfinder 1 would keep the "3" effect on - assisting in proof (even though SE pretty much confirmed) that possibly treasure hunter was a "debuff" and helping understand if the same person can overwrite their own TH4 with TH2 (without gloves and knife) for the kill.

Maybe I will get back around to it one day, nobody could really give me a mob with a static gil drop that was soloable and a reasonable respawn time, and since the gil drops in beaucedine were so erratic I pretty much lost interest.

I think there are a lot of NMs with a set gil drop amount. The problem, of course, would be finding ones that you could solo with all the different gilfinder traits on. But then again, even if you had additional people there, the amount of gil is just split amongst the participants.

Check out otherWiki for possible NMs... a lot of them have the exact gil drop amount listed.

[Max™]

Lowbie Beastmen NM's in Davoi/Bead/Oz?

The two Quads that drop the Charm/Rosary or whatever it was?

[Tordall]

I was under the impression that actual testing of TH was rather difficult, unless it's the effect of TH on a particular drop. Since so many things seem to affect drop rates. Various theories have speculated on particular weaponskills, game day, moon phase, jug pets, lucky eggs etc. There is also evidence that particular drops share a slot (impact knuckles and okotes for example), where TH likely has no effect, and mobs where TH seems to help, but there's still a limit on the number of a particular drop.

Anyway, comprehensive testing would likely have to be for a particular drop, take place over about a 2 week period (game year), and at the end there would have to be a fourier transform done to remove the game-year, game-day, and moon phase contributions.

You'ld also have to pay attention to annoying things like the statics of drops where you can get at most one (like emp. hairpin or bounding boots), or mobs where there is an 'either or' loot pool, some of which include 'nothing' as a valid loot option (kirin for example).