A new look at Treasure Hunter

[Shuemue]

Man, poof and toksyuryel working together, what a dream team!

[Amele]

OK, well this is kinda detracting from the actual thread. You guys keep going on and on about your main processors and you are totally missing the point. I say "clock cycle" and you completely disregard memory clock cycles and bus bandwidth and just start an epeen argument over processing power. Full 256 bit memory addressing itself is still relatively new and you mostly only see it in the latest 600 dollar GPUs.

a flop is a floating point operation per second. this is a complete pipeline start to finish. (and is a good example of how much activity you can do)

to discuss 'clock cycles' when considering operational speed in a pipelined multi-stage modern co-processor is like discussing how many horses(the animal) you need to go 60miles an hour with a combustion engine.

and do you know how long a 256 bit word is? it can store numbers as large as: 1.159x10^77 (so large I'm not going to bother writing it) and that has no bearing on the discussion of whether or not it's necessary to constrain the minimum resolution to 8bits instead of 10 or 16 or 32 in a 64 bit system.

[Gaea]

I didn't say a 256 bit word I said 256 bits. I also know what a flop is. I'm not going to debate the issue of modern computing any longer. Whether or not the system is n/256, n/512 or n/1024 is irrelevant because the differences are minimal and the formula would relatively universal as long as its applied in a base 2 system.

Anyway, I was attempting to create a base 2 model for determining drop rates and applying Treasure Hunter because I think the base 10 theory is fundamentally flawed. Instead, I'm getting rejects with 24 posts trying to derail the thread.

In the mean time, I'll continue to do my own research, and if the model does pan out, I guess I'll just accept full credit rather than making it a community effort. Since I'm relatively sure that this system will actually pan out after observing SE titles released post and prior to FFXI, I thought it would be worth sharing this information and conducting some community research.

[Toksyuryel]

OP is being a dick, but I think seriously researching this is important so don't let him scare you off. We've already come up with some interesting data and I'm sure it'll get even better once sample sizes are sufficiently large.

[Gaea]

I suppose. I've hit report a couple of times and hopefully Sonomaa will clean out the thread and ban if necessary. It's rare I feel really strong about a "new idea" in FFXI, especially 6 years after it was released, but I can't help but feel that way now. The fact that the mechanics governing this phenomenon have yet to be proven leads me to believe that we were going in the wrong direction. Hopefully, looking at this in a different light, and taking it a new direction will reward us with some real progress.

[Wizerd]

I suppose. I've hit report a couple of times and hopefully Sonomaa will clean out the thread and ban if necessary. It's rare I feel really strong about a "new idea" in FFXI, especially 6 years after it was released, but I can't help but feel that way now. The fact that the mechanics governing this phenomenon have yet to be proven leads me to believe that we were going in the wrong direction. Hopefully, looking at this in a different light, and taking it a new direction will reward us with some real progress.

I personally believe that no amount of testing will result in any real progress when it comes to Treasure Hunter. We can make educated guesses but it will be similar to Magic Accuracy at best.

Due to the large samples required and the tediousness of obtaining said samples, finding concrete information is nigh impossible.

To get a *decent* understanding of Treasure Hunter, I would recommend killing one monster at least 3000 times with each level of Treasure Hunter, and make sure this monster has at least 2 drops at different rarities. Then repeat this test for at least 10 different types of monsters. Observe how the effects of Treasure Hunter differ from item to item and mob to mob.

Then, and maybe then, will we be able to understand Treasure Hunter to a satisfactory degree. The reason I say 3000 kills is because the time it would take to kill 3000 monsters of the same type in any reasonable scenario is at least 12-16 hours. This covers a broad range of days and spans across a decent amount of moon cycles. It also requires you to split up your testing over a period of a few days per set, which allows the moon cycles to sway in and out.

I don't believe in moon or day affecting drop, but for those that do, something to this degree should satisfy even them.

Of course, the amount of kills I describe above is 150,000. Have fun.

[Toksyuryel]

I think if we work together on the same mobs, even 10,000 mobs per level shouldn't be out of reach, if say 10 people do 1000 each. The more the better. I think to start we should choose a mob with only one drop and that can't double drop it, just to get a baseline. I've suggested Crawlers in West Sarutabaruta [S] for this purpose, as there are a lot of them clustered together on fast repop timers and the item they drop is worth a decentish amount of money. We can branch out from there.

[carve]

what i fail to understand is why does the base 2 system change things ?

all your theory is suggesting is that there exists a (simple) formula which tells us how much % the base drop rate has been increased by

this percentage is not even depending on the base we use, only the base drop rate is

however this formula seems realistic and figuring it out by killing thousands of mobs is a good thing, but figuring out in which base the drop rates are coded server side is useless imo

[DAKPluto]

As a THF who has TH4 i will say this:

TH is a lie.

This.

It's a placebo so THF could do more endgame than just flee and die.

[Silenka]

This.

It's a placebo so THF could do more endgame than just flee and die.

^

[Aphelidia]

I suppose. I've hit report a couple of times and hopefully Sonomaa will clean out the thread and ban if necessary. It's rare I feel really strong about a "new idea" in FFXI, especially 6 years after it was released, but I can't help but feel that way now. The fact that the mechanics governing this phenomenon have yet to be proven leads me to believe that we were going in the wrong direction. Hopefully, looking at this in a different light, and taking it a new direction will reward us with some real progress.

Im not sure whats the point in going through all these formulas, testings, and collecting datas about TH. Lets assume you figure out the exact effect TH has on drop rates. Then what?

Items either drop or they dont.
Knowing its drop rate wont give you a better chance at a drop. This thread is therefore, completely useless.

[Toksyuryel]

How is finding out what the hell TH+1 really means in any way useless?

[Gaea]

Well, I don't think its useless, especially if certain items have caps placed for their drop rates. For example if you wanted to get a certain item with a maximum drop rate of 64/256 and you could attain that drop rate with Treasure Hunter I, then you could save yourself a lot of time just in that respect.

Also, depending on the way Treasure Hunter is applied, you could gauge if a certain level of Treasure Hunter is even worth the bother. The % increase for items with really low drop rights might not even be worth the hassle of higher levels of Treasure Hunter.

what i fail to understand is why does the base 2 system change things ?

all your theory is suggesting is that there exists a (simple) formula which tells us how much % the base drop rate has been increased by

this percentage is not even depending on the base we use, only the base drop rate is

however this formula seems realistic and figuring it out by killing thousands of mobs is a good thing, but figuring out in which base the drop rates are coded server side is useless imo

A base 10 system wouldn't be uniform for drop rates of different rarity values. The bonus applied to Treasure Hunter would vary greatly between items with different rarity values. This is a common issue for people trying to "debunk" Treasure Hunter. My suggestion is that they are just going about it the wrong way and there is a uniform way to apply Treasure Hunter bonuses that will be consistent between items with different rarity values.

[Shuemue]

In the mean time, I'll continue to do my own research, and if the model does pan out, I guess I'll just accept full credit rather than making it a community effort.

You're not curing cancer you fucking pompous admin: word removed, it's a floundering video game.

[BlueMage]

Sorry Poof, BG mods won't sleep with you the same way GMs do. Everyone knows TH does nothing.

[DancingEdge]

Should really just parse Greater Colibri Beak drop rates.

Wiki already has rates on it, do it with a burn party, then a burn party with thf.

super fast kills, lots of data.

[Amele]

A base 10 system wouldn't be uniform for drop rates of different rarity values.

^ mathematically speaking, there's no difference between a base 10 and a base 2 system, unless you lock yourself into forms like 01 010 0100 01000{base x} which is still going to be potentially uniform, just wildly larger in variance between 'tiers'.

videogames use base 2 systems because (surprise) digital electronics is base 2. edit: (note, mostly humans express base 2 in computers as base 16; which is a convenient demonstration of my point)

humans use base 10 systems because (surprise) human being's counting systems are either base 10 (hands) or base 20 (if you're barefoot).

since the translation rules exist to do perfect mathematical translations between two arbitrary bases, all assuming base 2 really does is give you a convenient jumping off point for locking into exact instead of approximate values.

but being base 2 doesn't automatically mean x/256 (a point that I see you've now conceded so I guess this is a case of violent agreement)

[Kaeko]

I would proceed extremely cautiously if you truly want to 'test' and 'prove' this in the correct manner (a.k.a. in a scientific manner). The fundamental 'rule' of starting any sort of test is to make a SIMPLE hypothesis, NOT a full blown THEORY. You start SIMPLE, and move towards the complex IF and only IF the data supports it.

As someone else so wisely pointed out you're not curing cancer, you're testing some formula in a video game, but the way you go about doing it is really the exact same way - scientific method. In order to make claims like what you have in this thread, you MUST have data to back it up or you won't convince anyone and it won't matter if you're right or wrong because no one will follow it. It is up to the RESEARCHER to convince the average person, NOT the average person to treat your theory with automatic support. The fact people here are being harsh to your theory is what is supposed to happen. If people accepted shit to be true without being critical the real world would get nowhere. While this is just a game, the same concept applies universally.

Right now, what I see is you proposing a full blown theory before getting any data, then trying to find data to support your theory. The fact you came up with the theory (of base 2s) BEFORE collecting data automatically predisposes you to bias. Even worse, you completely disregard Khoisan's data, which was performed in almost perfect scientific fashion (moving from hypothesis to data to data analysis). You're supposed to look 10x more closely to things that contradict your own ideas, not disregard them.

I think it's nice that you want to test something as difficult as this, but you're doing it the wrong way - no one will ever believe you without the data and proper procedure, whether you turn out to be right or wrong. Also, lots of things work on base 10 in this game. For instance, I personally felt enmity would be base 2, but I was not so hell bent as to ignore data saying otherwise. It turns out enmity is in base 10. Keep an open mind, don't take short cuts, and be most thorough in the DATA COLLECTION process - NOT the theory. I feel you're taking short cuts to get to the 'glory' without putting in the work right now. The fact you mention wanting all the credit is really off base as well. You're 1) not going to get a freaking Ph.D. from this, it's a game, and 2) there is 0 way you can fully credit yourself because someone (Khoisan) already conducted a formal TH test. To give yourself full credit now would involve disregarding Khoisan, which would place significant black marks on anything you eventually come up with.

I'm not trying to be harsh but just everything about the way you're doing this is completely against the way it's been generally accepted you do any testing for the last like 600 years - thank god people actually don't try to cure cancer this way.

[BlueMage]

Poof has GM friends that she shows naked pictures of herself to in return for special treatment. Said GM friends make her so special and all knowing. Don't mess with the high and mighty Poof, Kaeko! SHE'LL GET YOU BANNED!!!!!!111ONE!

That's her proof, she sexed up some GMs (well known fact on Fairy).

[payu]

love you gaea