Treasure Hunter Demystified

I thought TH only affect the slot, not what's in it. the slot can be deemeds common rare or whatever but it's still only the slot.
then each slot has a pool which is unnaffected by TH.

is this all wrong ?

Hard to say now with the information released but the answer seems to now be “Both”

The slot itself as well as the contents apparently. But it’s reasonable why we never figured this out sooner; we had THOUSANDS of tests and Gigs of data collected on TH back when it stopped at +4 (was only +4 right?). Then SE added TH3 and stacking to the system and I don’t recall anyone taking on that mountain of testing again to find it. And according to the chart, the impact it has on rarer drops are low enough to where we’d all be certain it’s not doing anything (I.e. Omen Bodies)

So now we have to trust in our own findings and add the released information to it. Looks like Stacking to at least 12 will be a thing?

This is almost certainly correct, not wrong.

I know it's been commonly agreed for ages that TH affects the quantity of slots for drop and not the quality of the drop in the slot, but isn't it possible that it's actually false ? And that TH affects the % drop rate of a rare quality in its slot ?

KB :
D-Ring or Pixie Earring in the slot is guaranteed,
TH14 would grant a 5% chance to drop (ie. Pixie 95%) and if only TH4 is applied then it's 1.40% chance for DRing ?

This is sill wrong ?

Well if that is true, why will TH affect dring and not pixie earring? Instead, if you put TH14 on KB the chance for pixie earring will increase from.95% to 99%!

Idk, shouldn't be too hard to set rules such has TH prevails on the rarest items, and the rest if less rare items share the same slot. (Such as how QA>TA>DA works). It doesn't seems rocket science to program it this way.

I agree. Proceeding under assumption that it influences drops even from locked-pools we can assume that if the released information is correct it will predicate Defending Ring into the Rare Bracket while Pixie Earring stays in Common. Over thousands of drops we know that Pixie vs DRing is 95% vs 5% at default. I imagine TH14 should tilt those odds in DRing’s priority to 20% if it works in this case. Using DRing is only for a model purpose by the way; just to see if there’s any influence.

I could make a chart to show you are full of ***. But I think you already know that.

Umad?

No but I do appreciate the effort. Look I don't give a ***what you think. You are all equally *** stupid.

Here's the thing, if treasure hunter affected the chance of a rare item loading in a shared pool over the more common items we would have seen it by now. There have been tens of thousands of players bringing thief to king behemoth, omen bosses, lillith fights, etc over the course of the past few years alone. Many many times have they procced treasure hunter all the way to 12 or higher. So if treasure hunter really did affect item quality we have enough cumulative pooled community data that we would have seen it already.

Spoiler alert: We haven't.

We can't disprove a negative, but if TH influenced the ratio of drops within a shared slot by the amounts SE posted, there should be some indication of that variance in the existing data.

It's a safe assumption that many of the 15,268 KB kills recorded @ FFXIDB had some amount of TH. Either it's 5% and unaffected by TH, or it's naturally lower than 5% and the total (unknown)amount of TH used for those kills has resulted in an uncanny coincidence. Which do you think is more likely?

I always kill Lilith on thf. It's been my understanding that if the battle drops a coffer, TH means nothing. If the monster dies and you get drops, TH applies.

I've always been of the mindset that always bring thf no matter what, because +anything is always better than +nothing.

But FFXIDB shows the TH level up to 3, and there is an increase for the levels recorded for Dring. DRing lists at 4.0% at th0 4.5% at th2 and 5.2% at th3+

"I don't care what you think"

/Sittingmad

*but let's post to them to let them know I don't care, that'll show 'em*

Now I'll get back to the rest of us whom are equally *** stupid and play a game I still enjoy.

Well within expected variances for the sample size

The sample sizes for th0 and th2 are quite small compared to th3. (exact #'s are in the tooltip)

Yes, because SE said TH has no effect on King Behemoth's situation (and thus likely other scenarios of the same type).

From personal experience and polling various data sources and people, I'm under the impression that:

1.) TH does not affect drops that appear in a dedicated slot (even if multiple possible items can appear in that slot, like D.Ring/Pixie Earring).

However, I do subscribe to the idea that TH affects drops in other kinds of slots. Specifically:

1.) Drops with a dedicated slot that doesn't appear 100% of the time. In this case, it is suspected that TH affects the appearance of the slot, not the contents of the slot.

Similarly, for drops with no dedicated slot, i.e. can appear in any slot in the pool, I would refer to the table provided by SE. (This includes omen bodies shared in the 2nd slot of Omen boss drops, Volte armor, which can appear in any slot dropped a Wave 2 NM/Boss, etc. This situation is much like old-school NMs and not Escha-type NMs). As a result, I would hesitate to refer to the Treasure Hunter table for anything dropped in Escha zones, except maybe pulse weapons.

Edit: I can be wrong too and it is possible that TH operates in only two ways: TH affecting the non-100% slot only and not the item in the slot AND TH affecting the item in shared/non-dedicated slots. Either way, both cases are situations to apply TH according to the diminishing returns (if any) as specified by the table).

I'm surprised they gave us the actual rates for the tables. And yes FFXI's loot system is 100% based on drop tables with TH only effecting the chance at that table loading not the distribution of the items inside. A long time ago this was verified by hours long TH testing, the original post was somewhere on allahakzam.

Lets say a monster has four loot tables,

Table 1: 100%
Table 2: 100%
Table 3: 24% (very common)
Table 4: 10% (uncommon)
Table 5: 1% (very rare)

Applying TH4 would have them load as follows

Table 1: 100%
Table 2: 100%
Table 3: 64%
Table 4: 18%
Table 5: 2.5%

TH8 Makes it look like this

Table 1: 100%
Table 2: 100%
Table 3: 70.50%
Table 4: 22.50%
Table 5: 4.75%


We've actually seen monsters that have exact pattern. The one or two guaranteed slots full of cheap crafting materials. The one slot that has the kinda expensive crafting materials, and then the one or two slots where gear or desirable items drop.

The information provided actually matches observation. What kicks us in the nuts is when "trash" gear is in the same table as "useful" gear, with the terms "trash" and "useful" being subjectively defined by the community. I remember when I was farming for my argosy (a long time ago) I kept getting "trash" abjurations I didn't want. Awhile back I had to farm a different set and kept getting "trash" argosy that I didn't want. We can all relate to having experienced this before. We can't influence whats inside the table, only if the able loads in the first place. This is because the designers can't really decide which of the three items in the rare table the player wants to get.

No such thing, every slot has a table and tables do not share rolls. If an item appears in multiple tables it's because SE put it as a unique drop in each of those tables.

Example:

Table A:
Distilled Water
Salt
Iron Ore
Copper Ore

Table B:
Wild Onion
Crab Shell
Copper Ore
Brass Subligar

Those are two separate tables (that I totally made up) that are rolled separately with their own distribution. They just happen to both contain copper ore, there is no "universal table".

Thanks for the clarification! What is the relationship between drop slots and tables, when the number of drops slots can vary? (if you had to guess).

A table is just a list in a database, a slot is a pointer to that table. When a monster dies, it's on_death script will roll on each of it's slots and if a slot rolls true, it will then roll on table table to get the results.

This is a very common system in RPG's both tabletop and as a video game. Programatically it's easy to implement and update without changing any code. Just create the tables with the item distributions you want, then you reference them when creating the monster. You can even have multiple monsters reference the same tables or any combinations of those tables.

Too soon?

Lol you’re fine. I think the only reason I/we have been using DRing for this discussion is because it’s probably(definitely?) the most widely recorded Drop in this game’s history. So makes for an easily relatable discussion point for this.

*** this. lol It's like they could have told us this years ago but they wanted people to keep guessing and now since nobody gave that much of a ***anymore they're all. "Here's the real #s"

Just *** you Matsui... lol

Then you aren't trying hard enough, as mathematically you most certainly can.

I believe you given the desire to not change code and redeploy builds to change drop distributions. For non-dedicated slots, do you think TH contributes to the item distribution? I can see a randomly generated term influenced by TH, being used as the "dice roll" (lookup) into an drop distribution table for all the slots of a mob.

TH has zero effect on the distribution on the table, only if the table appears or not. There are some monsters in this game that have a table with only one item in it, see Ose. When it's a single item on a table, that item has 100% chance of appearing if the table is rolled true.

Think about it from SE's point of view, they don't know if anyone wants the Yamarang, the Dingir Ring, the Gin's Scale or the Ashera Harness. As players we obviously know the harness is the more desired drop, but the developers can't guarantee that and it would be unfair otherwise. In that table there is a total of 100% drop rate divided between those four items, if TH raised the distribution of the harness it must therefor lower the distribution of the other items. You can't have more then 100% of one item drop.

Th is very easy to observe if you test it on a set of regular monsters with several tables, it's quite noticeable.