A new look at Treasure Hunter

[Gaea]

MAJOR EDIT:

I created a new post that revises my theory if anyone is interested.
http://ami.calcobrena.com/2008/09/ff...56-debate.html

It's still the basic principal, the main difference is that I had to revise my theory of Treasure Hunter bonuses because I don't believe they'll be as generous as other Final Fantasy titles that use this ability that we HAVE dissected. Below is an excerpt from my latest blog entry regarding the issue.

The revised idea basically supports the concept that Treasure Hunter is likely more dynamic than simply doubling the original base value into a factorable result. This makes much more sense as it would be hard to customize FFXI's economy with my previously suggested linear design. Besides there's no reason why the values for n/256 can't give n any value they want.

To give an idea of how much more flexible this is as compared to a base 10 path of raw percentages, this method allows a difference as small as .390625% which is a little less than just 2/5 of a percentage point. A complex flat base 10 system wouldn't be as flexible. The n/256 model is also ideal for 32 bit computing as it allows the computation to be completed in a single basic clock cycle. Even today, because 32 bit computing is still the mainstream, SE is still releasing titles (that we can dissect since all the data is client side), using the n/256 mechanics. Older SE titles that were restricted to 8 bit and 16 bit computing utilized n/64 and n/128 mechanics to complete their computations for generating random results in a single clock cycle.

"Final Fantasy II" (IV in Japan), "Final Fantasy III" (VI in Japan) and ChronoTrigger all used n/128, and the 8 bit Nintendo version used the n/64. All 32 bit SE titles have used n/256 so there' s no reason to assume FFXI is different and it would explain alot of our frustration with trying to apply a raw base 10 system of percentages to the whacky mystery of FFXI drop rates and Treasure Hunter bonuses.

Anyways check out my blog post and you'll see what kind of information is needed to help us deduce an accurate formula for Treasure Hunter bonuses.

EXCERPT:
In my system of n/256, n is equal to x+x(a+b+cy). This formula follows the idea that Treasure Hunter (a), Treasure Hunter II (b), and Treasure Hunter +1" (c) have specific multipliers. However, it's possible that these values are actually fixed values and the formula is more simply n=x+a+b+cy. For clarity, x is the original number drop rate value, and y is how many pieces of equipment with "Treasure Hunter +1" are equipped. The first equation simply makes Treasure Hunter more dynamic than having a fixed value for each bonus.

[Gaea]

If you look at my blog, this is one of the many paragraphs. Haha, I'm sure I could have broken it down more but I wanted to kinda stay on topic without trailing off. I think what I really need to do is just break out MSPaint and simply I illustrate my model as most of today's youth seem to have an affinity for such a thing. I have a feeling my wall of text isn't going to get a lot of affection. Perhaps I'll even create a pop-up book for the stoners ^^ hehe

Right now, my theory is in its infancy and I don't want to go through all that work in breaking it down even further until I see I see how the community responds to it. I guess at the moment all I can do is apologize and hope that it will at least be analyzed.

[Jicent]

Or maybe a video with you zooming in on the text to let us figure it out.

[Gaea]

Originally Posted by Jicent

Or maybe a video with you zooming in on the text to let us figure it out.

Haha, you win my thread

[Andri]

Originally Posted by Gaea

Right now, my theory is in its infancy and I don't want to go through all that work in breaking it down even further until I see I see how the community responds to it.

If you don't make it easy to read and digest, no one will respond to it. Be kind to those you are asking help from.

That and the fact that you have the name for TH4 wrong, makes me skip down your entire post.

[Gaea]

I know, I know, it's so horrible I still haven't fully proofread it, haha. But if you could try to read a little into it you might want to actually finish it

Also, I'm not trying to have the name "wrong" I'm trying to explain it through a different perspective. For example at one possibility for this model you have actual "Treasure Hunter +4" meaning that 4 is added to the ratio of any rarity's given base value. A base value of 1/64 (1.5625%) would become 5/64 (7.8125%). Or in the other possibility that That Treasure Hunter II is applied separately after Treasure Hunter I, then you'd have "Treasure Hunter 4" boost the drop rate from 1/64 (1.5625%) to 3/16 (18.75%) The good news is, such a difference would be easy to test if we determined an item's drop rate to be 1/64 without treasure hunter.

[Sehbeh]

As a THF who has TH4 i will say this:

TH is a lie.

[Day]

Originally Posted by Gaea

I'd like to summarize by recapping the varying possibilities of this model.

I believe that any single item that can drop off a monster has a root value defining its rarity in a base 2 system.

These values are:

1/256 (.390625%)
1/128 (.78135%)
1/64 (1.5625%)
1/32 (3.125%)
1/16 (6.25%)
1/8 (12.5%)
1/4 (25%)
1/2 (50%)
1/1 (100%)

I believe that the Treasure Hunter job trait simply drops the rarity by a single tier effectively doubling the rate at which an item can drop. However, I believe that Treasure Hunter II may be more complex as it may simply add a bonus along with the first Treasure Hunter and create a ratio that simply can't be factored such as 3/256 (1.171875%) which would make more sense as rare items still seem to be rare even after several applications of Treasure Hunter.

However, there's also the possibility that the bonus is applied after the initial Treasure Hunter and that the rate drop rate is effectively doubled again. This could also make sense because when applied to the rarest possible drop rate of 256, that would give it a drop rate of 1/64 (1.5625%) as opposed to 3/256 (1.171875%).

The rarest possible items still remain pretty damn uncommon under the influence of either version of this formula. Treasure Hunter bonuses applied by the Thief's Knife and Assassin's Kote seem to point to the idea that original ratio itself is simply altered and result that can be factored is just a coincidence of the mechanics rather than a law governing the formula itself. That would mean that with "Treasure Hunter +3" your drop rate would be 4/256 which could be factored down to 1/64 and results in a percentage of 1.5625% when applied to the rarest possible drop rates. "Treasure Hunter +4" would likely reduce the rarity to 5/256 or 1.953125% making rare items still very rare even with with the highest possible rate of Treasure Hunter. Considering how some drop rates still seem to hover around 1% or 2% even with "Treasure Hunter 4" it is likely that this is the correct formula.

However, even under the most optimistic circumstances that Treasure Hunter II reduced the rarity of a 1/256 all the way down to 1/64 and the bonuses from Thief's Knife and Assassin's Kote are applied afterwards, that still only gives us a maximum bonus to the rarest items of (3/64) or drop rate of 4.6875% compared to the 5/256 ratio that grants us a 1.953125% drop rate. Such small differences would be difficult to test but if we applied either formula to a base rarity of 1/32 we'd have a difference of 3/8 (37.5%) or 5/32 (15.625%).

Finally, we would have to figure if multiple items are in the drop pool, if there's a cap on how many identical items can drop, and we have to find out if there are limitations that prevent fractions such as a drop rate of 1/4 being boosted to 5/4 or if that there are no limitations if this allows for an automatic drop rate or if the drop rate is rolled over into a new possible drop rate of an identical item. Of course, with an item of 100% drop rate that doesn't mean we'll get five them as I'm sure if this were the case there would be a cap imposed. Otherwise, items that drop 100% would have a drop ratio of 256/256 (1/1) and could be boosted to 260/256 or 5/1, depending on which formula for bonuses defines the model, and with no cap imposed we would be getting 5 of them under the latter formula and that just doesn't happen.
--------------------------------------------------------

Any thoughts would be appreciated, thanks.

­

[Bagel]

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

[Hawky]

Originally Posted by Gaea

After some realizations, I decided to publish my current theory of the Treasure Hunter and drop rate system. Unfortunately even if I post the general summary of it, it will still look like a wall of text but I think it could pan out to be an important contribution to the FFXI community.

For starters the entire outline of my theory can be found on my blog at: http://ami.calcobrena.com/2008/09/de...-and-drop.html

Here's a brief summary of that wall of text with this wall of text.

--------------------------------------------------------
I'd like to summarize by recapping the varying possibilities of this model. I believe that any single item that can drop off a monster has a root value defining its rarity in a base 2 system. These values are 1/256(.390625%), 1/128 (.78135%), 1/64, (1.5625%), 1/32 (3.125%), 1/16 (6.25%), 1/8 (12.5%), 1/4 (25%), 1/2 (50%) and 1/1 (100%). I believe that the Treasure Hunter job trait simply drops the rarity by a single tier effectively doubling the rate at which an item can drop. However, I believe that Treasure Hunter II may be more complex as it may simply add a bonus along with the first Treasure Hunter and create a ratio that simply can't be factored such as 3/256 (1.171875%) which would make more sense as rare items still seem to be rare even after several applications of Treasure Hunter. However, there's also the possibility that the bonus is applied after the initial Treasure Hunter and that the rate drop rate is effectively doubled again. This could also make sense because when applied to the rarest possible drop rate of 256, that would give it a drop rate of 1/64 (1.5625%) as opposed to 3/256 (1.171875%). The rarest possible items still remain pretty damn uncommon under the influence of either version of this formula. Treasure Hunter bonuses applied by the Thief's Knife and Assassin's Kote seem to point to the idea that original ratio itself is simply altered and result that can be factored is just a coincidence of the mechanics rather than a law governing the formula itself. That would mean that with "Treasure Hunter +3" your drop rate would be 4/256 which could be factored down to 1/64 and results in a percentage of 1.5625% when applied to the rarest possible drop rates. "Treasure Hunter +4" would likely reduce the rarity to 5/256 or 1.953125% making rare items still very rare even with with the highest possible rate of Treasure Hunter. Considering how some drop rates still seem to hover around 1% or 2% even with "Treasure Hunter 4" it is likely that this is the correct formula. However, even under the most optimistic circumstances that Treasure Hunter II reduced the rarity of a 1/256 all the way down to 1/64 and the bonuses from Thief's Knife and Assassin's Kote are applied afterwards, that still only gives us a maximum bonus to the rarest items of (3/64) or drop rate of 4.6875% compared to the 5/256 ratio that grants us a 1.953125% drop rate. Such small differences would be difficult to test but if we applied either formula to a base rarity of 1/32 we'd have a difference of 3/8 (37.5%) or 5/32 (15.625%). Finally, we would have to figure if multiple items are in the drop pool, if there's a cap on how many identical items can drop, and we have to find out if there are limitations that prevent fractions such as a drop rate of 1/4 being boosted to 5/4 or if that there are no limitations if this allows for an automatic drop rate or if the drop rate is rolled over into a new possible drop rate of an identical item. Of course, with an item of 100% drop rate that doesn't mean we'll get five them as I'm sure if this were the case there would be a cap imposed. Otherwise, items that drop 100% would have a drop ratio of 256/256 (1/1) and could be boosted to 260/256 or 5/1, depending on which formula for bonuses defines the model, and with no cap imposed we would be getting 5 of them under the latter formula and that just doesn't happen.
--------------------------------------------------------

Any thoughts would be appreciated, thanks.

So this is how AV is supposed to be killed.

[pugnax179]

need enmity to stick th

[Flight_93]

Originally Posted by Gaea

Here's a brief summary

brief.

1: short in duration, extent, or length2 a: concise b: curt , abrupt

[Gaea]

Ok, before my thread is completely derailed, here's an illustraton. (See attached Image)

In the first formula, it assumes that all Treasure Hunter bonuses equally alter the original drop ratio.

Let's say we have an item with the highest possible rarity of 1/256.

In the first formula, if you had the maximum possible number of treasure hunter bonuses (TH+4) you would get 1+4/256 or 5/256 which is a maximum possible drop rate percentage of 1.953125% for the rarest drops in the game.

The second formula illustrates the possibility that Treasure Hunter II is applied to the ratio that is calculated after Treasure Hunter I, thus giving us a maximum possible drop rate for the rarest possible value as 3/64 or 4.6875%.

The difference between the two formulas would be decided simply by figuring out when Treasure Hunter II is applied. This would give us an exact formula for discerning an item's drop rate. All we'd have to do is figure out if it's base drop rate is 1/256, 1/128, 1/64, 1/32, 1/16, 1/8, 1/4 or 1/2. The base drop rate would be its drop rate without any treasure hunter.

So, if an item has a "6.25%" base drop rate (1/16) then it would have a maximum possible drop rate of 31.25% with the first formula and 75% with the second formula. That means we could easily define which formula is correct by applying it to more common items.

The point also is that I'm sure you noticed how nothing likes to drop even with common items without any Treasure Hunter and then suddenly drops pour in even with something as basic as Treasure Hunter 1. That's because on more common items, the doubling affect is much more obvious. With Treasure Hunter II or "3" or "4" applied to common items, it almost seems like they drop "most of the time".

The point is, I don't think that drop rates are given raw percentages. I think they are assigned base 2 rarity values as they've been governed by in past FF games. Even the original Final Fantasy with its "1/64 chance" of encountering Warmech was a slave to this machine.

After analyzing other games including other FF games that also utilize Treasure Hunter, I've drawn these conclusions. Because FFXI is server side, it's been difficult for us to understand how Treasure Hunter and drop rates work and why they seem so random. On client side games, we can analyze the mechanics much closer and all I'm doing is applying proven facts and simple science to the mystery that is FFXI.

We don't have access to their server software and we can't look at what that software puts in memory and uses to determine variables, but we can analyze basic programming concepts, client side FF games, and other games that utilize similar systems to draw up a realistic theory and test a realistic model.

That's all I'm trying to do, and I'm just trying to make a connection that we've been unable to make for so many years.

[Gaea]

i'm still working on that part quannum. I want to use a treasure hunterless job to try to define the base rarity of a common drop and see if i can sort out which formula is actually correct. Then I want to start giving examples with items and the correct formula for people to test for themselves.

From there, once we have the correct formula, we can start seeking out the base value of more rare items such as those with base rarities all of 1/64, 1/128, and as high as 1/256.

[MaachaQ]

Originally Posted by Andri

If you don't make it easy to read and digest, no one will respond to it. Be kind to those you are asking help from.

That and the fact that you have the name for TH4 wrong, makes me skip down your entire post.

I think he means they are called "Assassin's Armlets" not Kote.

I will give the rest of your ideas a read, but hmm, needs more formatting to make it palatable...

[EdgeMoragan]

If it jumps from 50% to 100% how does this explain stuff like Sewer Syrup dropping the ring only about 90-95% of the time <_<? The specific behaviors of that NM's drop has always kind of stuck out to me.

[Ezell]

TH2+2=Just another reason for your LS to yell at you about "Broken" THFs.

If you want to make some head way, find out how the +1+1 affects 1% drop rates. The convential wisdom is the +1+1 is +1%+1%. What that means i have no idea, and according to the other live journal it's not a % value of 1 against bees. But those tests where far from inconclusive.

Or you can do what most THFs do, go to SSR and watch empty treasure pools all day long and wonder if it's really worth your XP bar and pride coming there with TK and AA equiped >.<

*Edit Cream Soda beat me to it^^

[Gaea]

Originally Posted by EdgeMoragan

If it jumps from 50% to 100% how does this explain stuff like Sewer Syrup dropping the ring only about 90-95% of the time <_<? The specific behaviors of that NM's drop has always kind of stuck out to me.

After toying around with the possibilities I'm beginning to draw certain conclusion to explain that phenomenon.

A: Its possible that that there's a cap on the drop rate at around 95% similar to accuracy. It's also possible that some items may even have different caps but I don't really think that's likely.

B: Treasure Hunter isn't applied to the most common items. (Unlikely but possible that very common items with a drop of rate such as 1/4 just aren't affected).

C: I believe certain items may have a set drop rate where Treasure Hunter doesn't apply. For example Soboro has a high drop rate of about (50%) and its drop rate seems unaffected by Treasure Hunter. With this model of Treasure Hunter, it would be difficult for it to affect it at all without virtually guaranteeing unless the item was given a separate cap than the "95%" one.

D: It's possible that since many monsters can drop multiples of the same item may have different rarity values for the same item. For example its possible that a specific NM crawler that can drop multiple silk threads may be have a 1/8 drop rate on one of the silk threads and a 1/4 drop rate on another silk thread giving you a 37.5% chance that at least one silk thread will drop and a 3.125% chance that both will drop. Toss in full Treasure Hunter bonuses and you would have an 87.5% chance of getting at least one silk thread, and a 15.625% chance of getting both.

[Gaea]

Originally Posted by Ezell

If you want to make some head way, find out how the +1+1 affects 1% drop rates. The convential wisdom is the +1+1 is +1%+1%. What that means i have no idea, and according to the other live journal it's not a % value of 1 against bees. But those tests where far from inconclusive.

My model doesn't have an actual "1%" drop rate, and the point of it is to illustrate how drop rates may be governed as a whole, and not just the effects of Treasure Hunter.

An example of a "1%" drop rate would probably be something like a 1/256 drop with basic "Treasure Hunter II". Under the first formula this would be a drop rate of 3/256 (1.171875%) The second formula would offer a 1/64 drop rate or 1.5625% with plain "Treasure Hunter II".


Now, as far as Usukane 35 head I'm sure its either a 1/256 drop or a 1/128 drop. Optimistically, if it has a 1/128 rarity that gives it less than .8% chance of dropping without any treasure hunter (0.78125%).

Assuming that Treasure Hunter can affect this drop, under the first model, the drop rate would be:
BASE: 1/128 (.78125%)
TH1: 2/128 factored to 1/64 (1.5625%)
TH2: 3/128 (2.34375%)
TH3: 4/128 factored to 1/32 (3.125%)
TH4: 5/128 (3.90625%)

As you can see Treasure Hunter bonuses at this level make such a small impression that sometimes it feels like Treasure Hunter isn't working. This also optimistically assumes that the base drop rate is a generous 1/128. If it were 1/256, you would have less than a 2% chance of it dropping even with max treasure hunter (which may actually be the case).

Now, let's look at an item with a base rarity of 1/128 with the 2nd formula which is actually a little more generous. It assumes that Treasure Hunter II is applied after Treasure Hunter I.
BASE: 1/128 (.78125%)
TH1: 2/128 factored to 1/64 (1.5625%)
TH2: 2/64 factored to 1/32 (3.125%)
TH3: 3/64 (4.6875%)
TH4: 4/64 factored down to 1/16 (6.25%)

Even with these most optimistic circumstances applied, the Usukane 35 head would only have a 6.25% drop rate with TH4. If the usukane 35 head has a rarity of 1/256 then that's drops us back down to a 4.6875% with maximum Treasure Hunter.

If you are pessimistic and think it has a 1/256 drop rate but believe the head is affected by Treasure Hunter then that still leaves you with either a 1.953125% chance of the drop with Treasure Hunter 4 using the first model, and 4.6875% with Treasure Hunter 4 using the second model.

Either way, you aren't getting any usu heads and Treasure Hunter appears to be broken. Personally, I believe that Salvage drops are affected by Treasure Hunter.

I don't want this thread to turn into a Salvage drop rate dicussion, though. The point is, I believe these are the actual values we've been searching for for assessing the rarity of an item and it makes much more sense than a flat 1%, 2%, 3% idea.

[Hiroko]

Tachi: Koki, obviously.