Treasure Hunter Testing

[Insaniac]

Running totals

•Colibri feathers

TH0 - 41/161 = 25.46% +/-7.2%
TH2 - 97/157 = 61.79 +/-8.6%
TH3 - 313/542 = 57.74% +/-4.6%
TH4 - 71/113 = 62.83% +/-10.2%

•Colibri beaks

TH0 - 30/161 = 18.63% +/-6.4%
TH2 - 69/157 = 43.94% +/-8.5%
TH3 - 97/188 = 51.59% +/-7.9%
TH4 - 23/46 = 50.00% +/-14.4%
TH3/4 - 102/211 = 48.34% +/-7.4%

Due to the large amount of effort on my part (90-120 minute per session) I decided not to separate the TH3 and 4 beak drops because, quite simply, it's annoying. I'm not running a bot to do this and a decent amount of the recording is done with pen and paper so if you would like to see this test done differently I encourage you to do it yourself . If you have data you would like added to the running totals please try to post it in the same format as mine.

Example:

Total kills - Total number including both TH3 and TH4
TH4 kills - Number of TH4 kills
Total feathers - Total number including both TH3 and TH4
TH4 feathers - Number of feathers that dropped on TH4 kills
TH3/4 beaks - Total number of beaks including both TH3 and TH4 kills

Edit: Arcon will be supplying separate TH3 and 4 numbers for beaks and will be added to the running totals separately. I will continue to report my beak drops as TH3/4.

Aura statue testing- Credit to Vold
The goal here is to try and discern any noticeable difference between drop rates on rare items between TH2 and high levels of TH.

TH2
Golem shards - 9/100 - 9%
Mythril ores - 47/100 - 47%
Raxa - 0/100 - 0%
Diorite - 11/100 - 11%

TH6
Golem shard - 7/100 - 7%
Mythril ore - 53/100 - 53%
Raxa - 0/100 - 0%
Diorite -7/100 - 7%

[Insaniac]

Batch 1 - 8/9

Total kills - 210
TH4 kills - 26
Total feathers - 129
TH4 feathers - 14
TH3/4 beaks - n/a

[Insaniac]

Batch 2 - 8/10 (Not included in running totals due to dramatically higher drop rates due in theory to a TH bonus from sanction)

Total kills - 182
TH4 kills - 25
Total feathers - 127
TH4 feathers - 23
TH3/4 beaks - n/a

[Insaniac]

Batch 3 - 8/11

Total kills - 211
TH4 kills - 41
Total feathers - 125
TH4 feathers - 26
TH3/4 beaks - 102

[Babekeke]

If you're not going to use a better format with stricter guidelines, you're wasting your time and might as well stop now. The sheer fact that you're not even counting beaks means that your informations is useless. What if the difference between TH3 and TH4 is simply more chance for a 2nd item to drop?

[Insaniac]

If you're not going to use a better format with stricter guidelines, you're wasting your time and might as well stop now. The sheer fact that you're not even counting beaks means that your informations is useless. What if the difference between TH3 and TH4 is simply more chance for a 2nd item to drop?

The goal of this test is to see what exactly TH3 does compared to TH2. For the moment I'm comparing it to an 11k sample of TH2. That's the only reason I'm separating TH3 and 4 on the feathers. The beaks are really irrelevant to the test but someone in the other thread wanted me to post them so I did. Eventually I will get around (or someone else will) to retesting TH2. Honestly as far as TH4 goes we will probably never have a large enough sample size to nail down exactly what the TH+1 from TH4 does. Though, we will get a sense of the effectiveness (or lack there of) once I start getting some TH2 numbers from ???/THF kills.

Edit: Just to be more clear the original goal of this test was to prove that TH3 was actually just TH2+1. So far that's exactly what it's looking like. I only decided to keep adding to the sample to lower the margin of error so comparing it to a new TH2 test with a smaller sample size would be possible.

[Byrth]

Any TH undertaking is always massive. This is a useful datapoint whether he tracks TH4 vs. TH3 or not.

Theory in practice at the moment is this:
1) TH1 matters, gives a re-roll
2) TH2 matters, gives a second re-roll
3) TH+1 and the TH proc system, as far as any of us can ascertain, do all of jack shit.

So it doesn't really matter if he tracks the proc system, as long as number 3 is true. If we'd like a second way to measure this, I've started working out the loot pool for procced NQ Dynamis monsters:
Slot 1: Currency 100%
Slot 2: Currency ???%
Slot 3: Currency ???%
Slot 4: AF2/D1 Weapon
Slot 5: Crafting item
Slot 6: Currency ???%

I'm betting Slots 4-6 aren't affected at all by !! Proc, and Slots 1-3 are added.

[Arcon]

Personally, I still find it very hard to believe that even gear-TH works differently from trait-TH, let alone different tiers of trait-TH working differently from each other. For that, it really would matter if he tracks the procs. I will redo some TH2 testing when I get to it, to see if Thorny's data is still accurate.

I just started doing some TH0 kills last night, but was called away. However, I did manage to get this:
THX
Lesser Colibri
Name Drop rate Drop rate (%)
Colibri Feather 41/161 25.5%
Colibri Beak 30/161 18.6%

It's a relatively small sample, but it falls pretty clearly within Thorny's TH0 test. I'll focus on TH2 next.

If you could do me a favor, when you add the drop rates to your running totals, could you seperate those clear TH3 and TH4 kills, and those unknown TH3~? I would really appreciate that, and you can still add the pure TH3 kills into the TH3~ section, if you think TH4 doesn't matter. Anyways, I'll be doing a TH2 test soon (WAR/THF), and then move on to TH3 without WS/gearchange.

Edit:
I've updated my parser. It now seperates different TH procs. However, there are some limitations, since it works off of the chatlog alone:
- You can't kill more than one mob at a time (AoE kills do work, just not randomly changing target and procing TH on both mobs for example)
- If you die it won't know if the TH effect on the mob has been reset, so it will assume whatever TH was on before is still on
- It won't show drop rates for items that didn't drop at all (if you kill leeches it may not show a Carbuncle's Ruby if none drops), so if you know a mob is supposed to drop something else, you'll need to manually calculate 0/x for that item
- It doesn't know what the "base" TH is, so for drops without visible TH procs, it will say "THX"

Note that I'm still working on it, and I don't have a dedicated testing site, so you might notice some strange output occasionally, but it won't mess up the drop rate data, that it does accurately by now.

[Arcon]

This time WAR90/THF45. Made sure to get TH on every mob by hitting them directly.

THX
Lesser Colibri
Name Drop rate Drop rate (%)
Colibri Beak 69/157 43.9%
Colibri Feather 97/157 61.8%

Just a small sample again, but as before, it falls very close within Thorny's reported stats, a bit higher on feathers, a bit lower on beaks. Assuming all items receive the same boost from TH (which isn't even proven yet, but I'll assume that's the case for "common" items, or non-NM drops), it evens out nicely to his reported drop rate.

I'll do TH3~ next.

[Arcon]

Next batch cut short, because some people can't behave. These results are clean though:

THX (assumed TH3)
Lesser Colibri
Name Drop rate Drop rate (%)
Colibri Feather 109/188 58.0%
Colibri Beak 97/188 51.6%

TH4
Lesser Colibri
Name Drop rate Drop rate (%)
Colibri Feather 31/46 67.4%
Colibri Beak 23/46 50.0%

TH5
Lesser Colibri
Name Drop rate Drop rate (%)
Colibri Feather 2/2 100.0%
Colibri Beak 1/2 50.0%

No gearswaps, no weapon skills, all TH procs accounted for. THF90/NIN45, no TH gear (TH3 base), all upgrades from melee hits. Another interesting tidbit... 4845 melee hits, 50 TH upgrades. I don't know if there was any testing done for this yet (there may have been, I'm a lazy researcher), but it seems the chance from melee procs to increase TH is 1%.