Drop Rate Analysis and Treasure Hunter

[Motenten]

TH1 (Mnk/Thf)

Savanna Dhalmel (Killed 100 times)
        4 dhalmel hide                 [Drop Rate (#):   4.00 %]  [Drop Rate (any):   4.00 %]  [Max #: 1]
        1 giant femur                  [Drop Rate (#):   1.00 %]  [Drop Rate (any):   1.00 %]  [Max #: 1]
       29 slice of dhalmel meat        [Drop Rate (#):  29.00 %]  [Drop Rate (any):  23.00 %]  [Max #: 2]
        9 tuft of dhalmel hair         [Drop Rate (#):   9.00 %]  [Drop Rate (any):   9.00 %]  [Max #: 1]

       Dropped  0 items    65 times ( 65.00 %)
       Dropped  1 items    27 times ( 27.00 %)
       Dropped  2 items     8 times (  8.00 %)

    Drop set: Nothing
       Count:  65  [ 65.00 %]

    Drop set: slice of dhalmel meat
       Count:  15  [ 15.00 %]

    Drop set: tuft of dhalmel hair
       Count:   8  [  8.00 %]

    Drop set: slice of dhalmel meat, slice of dhalmel meat
       Count:   6  [  6.00 %]

    Drop set: dhalmel hide
       Count:   4  [  4.00 %]

    Drop set: slice of dhalmel meat, tuft of dhalmel hair
       Count:   1  [  1.00 %]

    Drop set: giant femur, slice of dhalmel meat
       Count:   1  [  1.00 %]

Percent kills with drops: 35%, vs predicted ~37.5%. Supports theory.


Max number of drops per kill (observed):
TH0: 2
TH1: 2
TH2: 4

Does not support theory of +1 slot per TH level, but probably still applies with just TH2 having an effect. Would correlate with ram drop rates when I farmed them with TH2 vs TH0/1, where there seems to be a drastic increase in the number of drops possible. So, slots appear to matter, just the degree to which they matter, and how they are determined, is yet to be ascertained.



Percent of drops:

        w/TH0   w/TH1   w/TH2
hair:     39%     26%     35%
femur:    13%      3%   12.5%
grass:     9%      0%      1%
hide:      9%     11%      5%
meat:     57%     66%     75%

Not as close as the TH0 sample, but this is again a very small sample set. The progression of meat percentages is somewhat interesting, and I think can be traced to the number of available slots. The number of dual-meat drops was quite high. With a limit of 2 drops per mob, earlier high-chance checks may impact the odds of seeing later drops in the list. IE: a greater number of double meat drops meant that several times there was just no chance for femurs or hair to drop. Of course that depends on whether the drop list is random or ordered, and would need its own testing set.


For now the TH1 info supports enough of the theory that I'm willing to go with it, and will consider how the results stack up. Comparing the TH0 and TH2 data will be interesting due to the apparent difference in the number of slots available. Will need to put some thought into that. From now on I'll be adding to the TH0 data.

[Toksyuryel]

Perhaps that is the effect TH2 has? Maybe TH1 = increase rate that items will drop, and TH2 = increase slots for items to drop in? And then TH+1 would simply affect TH1, further increasing the chance of a drop. Requires more research though.

[kwontess]

300 more data points. Dhalmel with TH2. I'm not going to bother compiling statistics on these until we get the full 1000 (at 700 now so it shouldn't be too long). Anyone out there with TH2 want to do a couple hundred :D

nothing: 50,41,36
meat: 17, 21,18
meat+hide: 1,0,0
meat+hair: 4,3,6
meat+femur: 1,6,4
meat+femur+hair: 1,0,1
meatx2: 3,7,12
meatx2+hide: 0,0,1
meatx2+femur: 3,3,3
meatx2+hair: 2,3,1
meatx2+femur+hair: 2,1,0
hair: 13,12,12
hair+grass: 0,1,0
hide: 1,1,3
femur: 2,0,2
grass: 0,1,1

[Ezell]

This is awesome guys!!!!
GJ

[Motenten]

Summary of data for TH0 on dhalmels so far:

Savanna Dhalmel (Killed 400 times)
        3 dhalmel hide                 [Max #: 1]  [Items/Kill:  0.008]  [Drop Rate:   0.75 %]  [% of Drops:   3.09 %]
       11 giant femur                  [Max #: 1]  [Items/Kill:  0.028]  [Drop Rate:   2.75 %]  [% of Drops:  11.34 %]
       43 slice of dhalmel meat        [Max #: 2]  [Items/Kill:  0.108]  [Drop Rate:   9.75 %]  [% of Drops:  44.33 %]
        2 sprig of papaka grass        [Max #: 1]  [Items/Kill:  0.005]  [Drop Rate:   0.50 %]  [% of Drops:   2.06 %]
       38 tuft of dhalmel hair         [Max #: 1]  [Items/Kill:  0.095]  [Drop Rate:   9.50 %]  [% of Drops:  39.18 %]

       Dropped  0 items   317 times ( 79.25 %)
       Dropped  1 items    70 times ( 17.50 %)
       Dropped  2 items    12 times (  3.00 %)
       Dropped  3 items     1 times (  0.25 %)

    Number of times each group of items dropped.

      317 [ 79.25 %] -- Nothing
       31 [  7.75 %] -- tuft of dhalmel hair
       30 [  7.50 %] -- slice of dhalmel meat
        6 [  1.50 %] -- giant femur
        4 [  1.00 %] -- slice of dhalmel meat, slice of dhalmel meat
        3 [  0.75 %] -- giant femur, tuft of dhalmel hair
        3 [  0.75 %] -- slice of dhalmel meat, tuft of dhalmel hair
        2 [  0.50 %] -- dhalmel hide
        1 [  0.25 %] -- dhalmel hide, sprig of papaka grass
        1 [  0.25 %] -- sprig of papaka grass
        1 [  0.25 %] -- giant femur, slice of dhalmel meat
        1 [  0.25 %] -- giant femur, slice of dhalmel meat, tuft of dhalmel hair

Of note, there was a 3-item drop, so the max number of items at TH0 is not 2. Therefore that part of the theory is no longer valid (though it remains to be seen if 4 items can drop with TH0).

More importantly, though, as I was reviewing and trying to process the results, I realized that some aspects of this just don't add up.

Given the presumed drop rates for meat and hair, there is no way that the proportion of 2-drop kills should be as low as it is. Meat and hair alone should generate about 60% 2-drop kills if there is no restriction on slots. I have some more ideas on how this might work, but it will take some time to write up, and I need to consolidate Kwontess's data as well.

[kwontess]

One thing to note, if the game really does this as two rolls, the first to see if anything drops and the second to see what drops then you will need twice as many kills with TH0 to get equally accurate numbers.

[Motenten]

Another update.

Savanna Dhalmel (Killed 700 times)
        9 dhalmel hide                 [Max #: 2]  [Items/Kill:  0.013]  [Drop Rate:   1.14 %]  [% of Drops:   5.14 %]
       17 giant femur                  [Max #: 1]  [Items/Kill:  0.024]  [Drop Rate:   2.43 %]  [% of Drops:   9.71 %]
       69 slice of dhalmel meat        [Max #: 2]  [Items/Kill:  0.099]  [Drop Rate:   9.00 %]  [% of Drops:  39.43 %]
        2 sprig of papaka grass        [Max #: 1]  [Items/Kill:  0.003]  [Drop Rate:   0.29 %]  [% of Drops:   1.14 %]
       78 tuft of dhalmel hair         [Max #: 1]  [Items/Kill:  0.111]  [Drop Rate:  11.14 %]  [% of Drops:  44.57 %]

       Dropped  0 items   545 times ( 77.86 %)
       Dropped  1 items   137 times ( 19.57 %)
       Dropped  2 items    16 times (  2.29 %)
       Dropped  3 items     2 times (  0.29 %)

Taking:
Hide1: 1%
Hide2: 1%
Femur: 2.5%
Meat1: 5%
Meat2: 5%
Grass: 0.5%
Hair: 10%

Chance of 0 drops: 77.23% (observed: 77.86%)
Chance of 1 drop: 20.64% (observed: 19.57%)
Chance of 2 drops: 2.03% (observed: 2.29%)
Chance of 3 drops: not calculating.

Chance of all 7 dropping: 1 in 320 billion.

No 'slot' mechanism necessary to reproduce observed behavior. Bit of a relief, actually, because once I started thinking about how a slot mechanism would have to be implemented, I realized it got overly complicated very quickly.

Further, no "pre-determination" needed to see if 'something' drops. While it was an interesting possibility, the theory fell apart in certain areas, so just as well that it's not needed.


Note on the drop rate choices I picked above:
Meat: 5% because that roughly fulfills both mechanisms for estimating base drop rate.
Hair: 10% because that's about what it stayed at for most of the sampling period. Iceday produced an abnormally high 20% drop rate on hair which skewed the total upwards a little. It dropped back to normal on Lightningday.


I'll do another run tomorrow to finish up to 1000 samples, but am currently comfortable with this. I will note, though, that the totals so far indicate that TH is not applied equally to all items in the drop pool. Will get more specific when we're up to 1000 samples each.

[Motenten]

Final TH0 sample for dhalmels.

Savanna Dhalmel (Killed 1000 times)
       11 dhalmel hide                 [Max #: 2]  [Items/Kill:  0.011]  [Drop Rate:   1.00 %]  [% of Drops:   4.42 %]
       22 giant femur                  [Max #: 1]  [Items/Kill:  0.022]  [Drop Rate:   2.20 %]  [% of Drops:   8.84 %]
      100 slice of dhalmel meat        [Max #: 2]  [Items/Kill:  0.100]  [Drop Rate:   9.00 %]  [% of Drops:  40.16 %]
        3 sprig of papaka grass        [Max #: 1]  [Items/Kill:  0.003]  [Drop Rate:   0.30 %]  [% of Drops:   1.20 %]
      113 tuft of dhalmel hair         [Max #: 1]  [Items/Kill:  0.113]  [Drop Rate:  11.30 %]  [% of Drops:  45.38 %]

       Dropped  0 items   784 times ( 78.40 %)
       Dropped  1 items   188 times ( 18.80 %)
       Dropped  2 items    23 times (  2.30 %)
       Dropped  3 items     5 times (  0.50 %)

    Number of times each group of items dropped.

      784 [ 78.40 %] -- Nothing
      100 [ 10.00 %] -- tuft of dhalmel hair
       68 [  6.80 %] -- slice of dhalmel meat
       12 [  1.20 %] -- giant femur
        8 [  0.80 %] -- slice of dhalmel meat, slice of dhalmel meat
        7 [  0.70 %] -- dhalmel hide
        6 [  0.60 %] -- slice of dhalmel meat, tuft of dhalmel hair
        4 [  0.40 %] -- giant femur, slice of dhalmel meat
        3 [  0.30 %] -- giant femur, tuft of dhalmel hair
        2 [  0.20 %] -- giant femur, slice of dhalmel meat, tuft of dhalmel hair
        1 [  0.10 %] -- dhalmel hide, sprig of papaka grass
        1 [  0.10 %] -- sprig of papaka grass
        1 [  0.10 %] -- dhalmel hide, dhalmel hide
        1 [  0.10 %] -- slice of dhalmel meat, slice of dhalmel meat, tuft of dhalmel hair
        1 [  0.10 %] -- dhalmel hide, sprig of papaka grass, tuft of dhalmel hair
        1 [  0.10 %] -- giant femur, slice of dhalmel meat, slice of dhalmel meat

Going to assume more precise drop rates for final comparison. Note, once again (due to when the last batch was done), hair drops were higher on Iceday than others. Iceday had 19 drops out of 125 kills for 15%; Lightningday had 10 drops out of 100. I'm highly inclined to believe that drops have a base rate, and then get boosts on certain days. Can't verify that until I get the temporal filter built.


Taking:
Hide1: 0.5%
Hide2: 0.5%
Femur: 2.5%
Meat1: 5%
Meat2: 5%
Grass: 0.25%
Hair: 11%

Chance of 0 drops: 77.34% (observed: 78.40%)
Chance of 1 drop: 20.65% (observed: 18.80%)
Chance of 2 drops: 2.04% (observed: 2.30%)

Chance of selected groups // observed:
tuft of dhalmel hair: 9.6% // 10.00%
slice of dhalmel meat: 8.1% // 6.80%
giant femur: 2.0% // 1.20%
dhalmel hide: 0.78% // 0.70%
2x slice of dhalmel meat: 0.21% // 0.80%

Margin of error: 3.2%.

All data falls well within the margin of error. Almost all data is within 1% of predicted values. Selected drop rate percentages are of course subject to error, but the most direct implementation of drop percentages seems to apply. The 'slot' theory appears to be invalidated. The theory I proposed earlier for 'predetermination' appears to be invalidated.


Next will be a run with TH1 to see if there is any indication that TH is anything other than a direct modification to drop rate values (eg: multiple rolls against the loot table).

[zantes]

I believe that mobs have groups of possibilities for treasure. I've been keeping track of the Tremor Rams that I've been killing lately and watching the order the drops occur. What I'm seeing so far is as follows:

0-3 Ram Skin
0-1 Ram Horn
0-3 Ram Skin
0-1 Ram Horn

What I'm not sure about yet is whether each group of ram skins is 1 group or 3 individual chances to get ram skins. I'll keep recording the number and order of drops on rams to see what I can determine.

[Valyana]

I'm thought that Gilfinder affected the party?

Gilfinder is a debuff effect like TH. Once when I was leveling BST in SSG, I watched a THF fight a Sahagin for a while and then flee away. I took over when it went yellow and got significantly more gil than usual.

Certainly some drops are grouped: For instance, skeletons in Arrapago Remnants drop either 0 or 2 stat cells, never 1.

[Mojo]

I've been falling short in collecting data for these last few days, but everyone else has been doing a fantastic job. I've updated the spreadsheet with all of the new stuff. Go ahead and review it to make sure there are not mistakes. I organized the dhalmels from single > double > triple > quadruple drops, and added some base drop rate info at the end. At this point I think what would be most helpful is TH2+1 and TH2+2 data on dhalmels. Going to add some more stuff later.

[Toksyuryel]

You missed my 12/100 @ TH0 on crawlers

[Mojo]

Whoops, updated.

[Gunitsoldier]

Srsly, if anyone should be a mod, it's Mojo. So much great contribution that he brings here, i'm surprised he isn't one yet, you'll def have my vote if there's an election!

Ontopic: Nice info!
I'll stop kissing ass now

[Motenten]

You used my first report of TH0 on crawlers, but not my updated one (though you got the crow total). See post #58.

[Kimiko]

Mob: Bull Dhalmel (Watersday, 62%)

Kill #        Meat    Hide    Bone    Grass        Total Drops
-------------------------------------------------------------------------------------------------
01        1    0    0    0        1
02        1    2    0    0        3
03        0    0    0    0        0    
04        1    0    1    0        2
05        1    0    0    0        1
06        1    0    0    0        1
07        1    0    0    0        1
08        0    1    1    0        2
09        1    0    0    0        1
10        0    0    0    0        0
11        1    0    0    0        1
12        1    2    0    0        3
13        0    0    0    0        0
14        0    1    1    1        3
15        0    0    0    0        0
16        2    1    0    1        4
17        2    0    0    0        2
18        1    0    0    0        1
19        0    0    0    1        1
20        1    0    0    0        1
21        0    0    0    0        0
22        2    1    1    0        4
23        1    0    0    0        1
24        0    0    0    0        0
25        1    0    0    0        1
26        2    0    1    0        3
27        0    0    0    0        0
28        1    1    0    0        2
29        1    1    0    1        3
30        0    1    0    0        1
31        2    0    0    0        2
32        0    0    1    0        1
33        0    0    0    0        0
34        1    0    0    0        1
35        1    0    1    0        2
36        2    1    1    0        4
37        0    0    0    0        0
38        1    0    1    0        2
39        1    0    0    0        1
40        1    0    0    0        1
41        1    0    0    0        1
42        1    0    1    0        2
43        2    1    1    0        4

Drop Instance    29    11    11    4        35
# of Drops        36    13    11    4        64
# of 1-item drops    22    9    11    4        17
# of 2-item drops    7    2    0    0        8
# of 3-item drops    -    -    -    -        5
# of 4-item drops    -    -    -    -        4        
% of 0 Drops    32.56%    74.42%    74.42%    90.70%        18.60%    
% of 1-Item Drop    51.16%    20.93%    25.58%    9.30%        39.53%    
% of 2-Item Drops    16.28%    4.65%    n/a    n/a        18.60%
% of 3-Item Drops    -    -    -    -        11.63%
% of 4-Item Drops    -    -    -    -        9.30%

It's a small sample size so far, but I'll be working this for the next several days while farming Dhalmel Hides. Also, as part of this test, I've obtained the Dhalmel Saliva drop ahead of time (r/ex item) so this is not figured in.

[Motenten]

TH1 on dhalmels, Mnk/Thf. Cumulative with previous TH1 report.

Savanna Dhalmel (Killed 350 times)
       10 dhalmel hide                 [Max #: 1]  [Items/Kill:  0.029]  [Drop Rate:   2.86 %]  [% of Drops:   5.81 %]
       12 giant femur                  [Max #: 1]  [Items/Kill:  0.034]  [Drop Rate:   3.43 %]  [% of Drops:   6.98 %]
      105 slice of dhalmel meat        [Max #: 2]  [Items/Kill:  0.300]  [Drop Rate:  23.43 %]  [% of Drops:  61.05 %]
        2 sprig of papaka grass        [Max #: 1]  [Items/Kill:  0.006]  [Drop Rate:   0.57 %]  [% of Drops:   1.16 %]
       43 tuft of dhalmel hair         [Max #: 1]  [Items/Kill:  0.123]  [Drop Rate:  12.29 %]  [% of Drops:  25.00 %]

       Dropped  0 items   220 times ( 62.86 %)
       Dropped  1 items    93 times ( 26.57 %)
       Dropped  2 items    32 times (  9.14 %)
       Dropped  3 items     5 times (  1.43 %)

    Number of times each group of items dropped.

      220 [ 62.86 %] -- Nothing
       46 [ 13.14 %] -- slice of dhalmel meat
       35 [ 10.00 %] -- tuft of dhalmel hair
       18 [  5.14 %] -- slice of dhalmel meat, slice of dhalmel meat
        7 [  2.00 %] -- giant femur, slice of dhalmel meat
        7 [  2.00 %] -- dhalmel hide
        4 [  1.14 %] -- slice of dhalmel meat, tuft of dhalmel hair
        4 [  1.14 %] -- slice of dhalmel meat, slice of dhalmel meat, tuft of dhalmel hair
        4 [  1.14 %] -- giant femur
        2 [  0.57 %] -- dhalmel hide, slice of dhalmel meat
        1 [  0.29 %] -- giant femur, slice of dhalmel meat, slice of dhalmel meat
        1 [  0.29 %] -- dhalmel hide, sprig of papaka grass
        1 [  0.29 %] -- sprig of papaka grass

While it isn't complete yet, some peculiar trends are emerging relative to the TH0 data.

Looking primarily at the Items/Kill ratios:

          TH0     TH1
Hide:    1.1%    2.9%
Femur:   2.2%    3.4%
Grass:   0.3%    0.6%
Meat:   10.0%   30.0%
Hair:   11.3%   12.3%

First, the femur drop rate seems to have increased in line with the commonly accepted +50% rate, going from 2.2% to 3.4%. Ok so far.

Hair, though, seems to have hardly budged at all. I'll note that there's no Iceday drops in there, and if day affects things we might see enough of a boost doing that to push it close to a +50% rate. Wouldn't count on it though. At best it looks like a +25% rate.

Meat, on the other hand, has tripled in drop rate. That's beyond even the +100% rate that TH2 is being given at the moment. Hides also seem to be close to tripling in rate, but it's harder to say for sure on them.


Basically, there's no single TH formula application that can explain all three of those changes at the same time.


Further, the idea of independant drop rates for multiples of the same type (meat, in this case) doesn't jive with the numbers. Using the sum total approach, the numbers imply a 15% drop rate per meat. However that would indicate that the rate of 2x meat drops should be 2.25%, and current results are 6.6%.


Let me add in the current TH2 totals (of 700 kills):

          TH0     TH1    TH2
Hide:    1.1%    2.9%   2.9%
Femur:   2.2%    3.4%   7.7%
Grass:   0.3%    0.6%   0.7%
Meat:   10.0%   30.0%  51.1%
Hair:   11.3%   12.3%  18.6%

Drop rate of 2x meat is:
1.0% @ TH0 -- a bit high for 2 independant drops (should be 0.25%)
6.6% @ TH1 -- too high for 2 independant drops (should be 2.25%)
12.6% @ TH2 -- too high for 2 independant drops (should be 6.25%)


We also see that going from TH1 to TH2, hide drop rates don't change, femurs double, and hair gets a 50% boost.


I think that once we hit 1000 each on TH1 and TH2 for dhalmels, we need to find another mob with a mix of similar 'types' of drops to see if a comparable pattern emerges. Drop types: bone, leather, thread/cloth, cooking. At least 2, preferably 3 of those types from the same mob, with a reasonable drop rate on each (5%+).

[Necronus]

Update to my data, this is a continuation of what I posted earlier, not a seperate data set:

Maze of Shakrami: TH0

Carnivorous Crawler (Killed 538 times)
       76 spool of silk thread         [Drop Rate:  14.13 %]


Maze of Shakrami: TH1

Carnivorous Crawler (Killed 376 times)
       60 spool of silk thread         [Drop Rate:  15.96 %]

[Enedin]

Update to my data, this is a continuation of what I posted earlier, not a seperate data set:

Maze of Shakrami: TH0

Carnivorous Crawler (Killed 538 times)
       76 spool of silk thread         [Drop Rate:  14.13 %]


Maze of Shakrami: TH1

Carnivorous Crawler (Killed 376 times)
       60 spool of silk thread         [Drop Rate:  15.96 %]

<3 this thread.

Remember, the extremer the drop rate, the more you need to kill to get an accurate result. After killing 500 Gigantobugards I found my Tusk drop rates were skewed for some reason, the drop rate is just so low that it would take a lot more mobs for the percentages to normalize.

Let's say item A has 2% drop rate and item B has 50% drop rate. If you kill that monster a 100 times, there's a pretty big chance you'll end up with 40-60 of item B. 60, 50, it's not much difference per se. But with a 2% drop rate on item A, chances that you'll end up with 1-3 of item A are reasonable, but you just might end up with none of item A.

It's kinda hard to explain, pretty sure Bernoulli has something to do with it, same for extremely HIGH drop rates. Fastest results (read: when killing 100 additional mobs the results won't vary wildly) will be gotten through 50% drop rate items.

[Motenten]

Ok, finished the TH1 data set. Lots of things to say here; this will be long.

TH1 data

Savanna Dhalmel (Killed 1000 times)
       19 dhalmel hide                 [Max #: 3]  [Items/Kill:  0.019]  [Drop Rate:   1.70 %]  [% of Drops:   3.39 %]
       46 giant femur                  [Max #: 1]  [Items/Kill:  0.046]  [Drop Rate:   4.60 %]  [% of Drops:   8.20 %]
      350 slice of dhalmel meat        [Max #: 2]  [Items/Kill:  0.350]  [Drop Rate:  27.00 %]  [% of Drops:  62.39 %]
        6 sprig of papaka grass        [Max #: 1]  [Items/Kill:  0.006]  [Drop Rate:   0.60 %]  [% of Drops:   1.07 %]
      140 tuft of dhalmel hair         [Max #: 1]  [Items/Kill:  0.140]  [Drop Rate:  14.00 %]  [% of Drops:  24.96 %]

       Dropped  0 items   598 times ( 59.80 %)
       Dropped  1 items   272 times ( 27.20 %)
       Dropped  2 items   101 times ( 10.10 %)
       Dropped  3 items    29 times (  2.90 %)

    Number of times each group of items dropped.

      598 [ 59.80 %] -- Nothing
      143 [ 14.30 %] -- slice of dhalmel meat
      101 [ 10.10 %] -- tuft of dhalmel hair
       57 [  5.70 %] -- slice of dhalmel meat, slice of dhalmel meat
       22 [  2.20 %] -- giant femur, slice of dhalmel meat
       18 [  1.80 %] -- slice of dhalmel meat, tuft of dhalmel hair
       17 [  1.70 %] -- slice of dhalmel meat, slice of dhalmel meat, tuft of dhalmel hair
       12 [  1.20 %] -- dhalmel hide
       12 [  1.20 %] -- giant femur
        6 [  0.60 %] -- giant femur, slice of dhalmel meat, slice of dhalmel meat
        4 [  0.40 %] -- sprig of papaka grass
        3 [  0.30 %] -- giant femur, slice of dhalmel meat, tuft of dhalmel hair
        2 [  0.20 %] -- dhalmel hide, slice of dhalmel meat
        1 [  0.10 %] -- dhalmel hide, sprig of papaka grass
        1 [  0.10 %] -- dhalmel hide, giant femur, slice of dhalmel meat
        1 [  0.10 %] -- giant femur, slice of dhalmel meat, sprig of papaka grass
        1 [  0.10 %] -- dhalmel hide, dhalmel hide, dhalmel hide
        1 [  0.10 %] -- giant femur, tuft of dhalmel hair

TH0 (for reference)

Savanna Dhalmel (Killed 1000 times)
       11 dhalmel hide                 [Max #: 2]  [Items/Kill:  0.011]  [Drop Rate:   1.00 %]  [% of Drops:   4.42 %]
       22 giant femur                  [Max #: 1]  [Items/Kill:  0.022]  [Drop Rate:   2.20 %]  [% of Drops:   8.84 %]
      100 slice of dhalmel meat        [Max #: 2]  [Items/Kill:  0.100]  [Drop Rate:   9.00 %]  [% of Drops:  40.16 %]
        3 sprig of papaka grass        [Max #: 1]  [Items/Kill:  0.003]  [Drop Rate:   0.30 %]  [% of Drops:   1.20 %]
      113 tuft of dhalmel hair         [Max #: 1]  [Items/Kill:  0.113]  [Drop Rate:  11.30 %]  [% of Drops:  45.38 %]

       Dropped  0 items   784 times ( 78.40 %)
       Dropped  1 items   188 times ( 18.80 %)
       Dropped  2 items    23 times (  2.30 %)
       Dropped  3 items     5 times (  0.50 %)

    Number of times each group of items dropped.

      784 [ 78.40 %] -- Nothing
      100 [ 10.00 %] -- tuft of dhalmel hair
       68 [  6.80 %] -- slice of dhalmel meat
       12 [  1.20 %] -- giant femur
        8 [  0.80 %] -- slice of dhalmel meat, slice of dhalmel meat
        7 [  0.70 %] -- dhalmel hide
        6 [  0.60 %] -- slice of dhalmel meat, tuft of dhalmel hair
        4 [  0.40 %] -- giant femur, slice of dhalmel meat
        3 [  0.30 %] -- giant femur, tuft of dhalmel hair
        2 [  0.20 %] -- giant femur, slice of dhalmel meat, tuft of dhalmel hair
        1 [  0.10 %] -- dhalmel hide, sprig of papaka grass
        1 [  0.10 %] -- sprig of papaka grass
        1 [  0.10 %] -- dhalmel hide, dhalmel hide
        1 [  0.10 %] -- slice of dhalmel meat, slice of dhalmel meat, tuft of dhalmel hair
        1 [  0.10 %] -- dhalmel hide, sprig of papaka grass, tuft of dhalmel hair
        1 [  0.10 %] -- giant femur, slice of dhalmel meat, slice of dhalmel meat

TH2 so far (hand-compiled)

Savanna Dhalmel (Killed 700 times)
       20 dhalmel hide                 [Max #: 2]  [Items/Kill:  0.029]  [Drop Rate:   2.57 %]  [% of Drops:   3.53 %]
       54 giant femur                  [Max #: 1]  [Items/Kill:  0.077]  [Drop Rate:   7.71 %]  [% of Drops:   9.52 %]
      358 slice of dhalmel meat        [Max #: 2]  [Items/Kill:  0.511]  [Drop Rate:  38.57 %]  [% of Drops:  63.14 %]
        5 sprig of papaka grass        [Max #: 1]  [Items/Kill:  0.007]  [Drop Rate:   0.71 %]  [% of Drops:   0.88 %]
      130 tuft of dhalmel hair         [Max #: 1]  [Items/Kill:  0.186]  [Drop Rate:  18.57 %]  [% of Drops:  22.93 %]

       Dropped  0 items   331 times ( 47.29 %)
       Dropped  1 items   222 times ( 31.71 %)
       Dropped  2 items   104 times ( 14.86 %)
       Dropped  3 items    35 times (  5.00 %)
       Dropped  4 items     8 times (  1.14 %)

    Number of times each group of items dropped.

      331 [ 48.50 %] -- Nothing
      129 [ 18.43 %] -- slice of dhalmel meat
       75 [ 10.71 %] -- tuft of dhalmel hair
       52 [  7.43 %] -- slice of dhalmel meat, slice of dhalmel meat
       25 [  3.57 %] -- slice of dhalmel meat, tuft of dhalmel hair
       20 [  2.86 %] -- giant femur, slice of dhalmel meat
       14 [  2.00 %] -- giant femur, slice of dhalmel meat, slice of dhalmel meat
       14 [  2.00 %] -- slice of dhalmel meat, slice of dhalmel meat, tuft of dhalmel hair
       10 [  1.43 %] -- dhalmel hide
        7 [  2.00 %] -- giant femur, slice of dhalmel meat, slice of dhalmel meat, tuft of dhalmel hair
        6 [  0.86 %] -- giant femur
        3 [  0.43 %] -- giant femur, slice of dhalmel meat, tuft of dhalmel hair
        2 [  0.29 %] -- giant femur, tuft of dhalmel hair
        2 [  0.29 %] -- sprig of papaka grass
        1 [  0.14 %] -- dhalmel hide, tuft of dhalmel hair
        1 [  0.14 %] -- dhalmel hide, slice of dhalmel meat, tuft of dhalmel hair
        1 [  0.14 %] -- dhalmel hide, dhalmel hide, sprig of papaka grass
        1 [  0.14 %] -- dhalmel hide, dhalmel hide, slice of dhalmel meat
        1 [  0.14 %] -- dhalmel hide, giant femur, slice of dhalmel meat, tuft of dhalmel hair
        1 [  0.14 %] -- slice of dhalmel meat, sprig of papaka grass
        1 [  0.14 %] -- dhalmel hide, giant femur
        1 [  0.14 %] -- dhalmel hide, slice of dhalmel meat
        1 [  0.14 %] -- dhalmel hide, slice of dhalmel meat, slice of dhalmel meat
        1 [  0.14 %] -- sprig of papaka grass, tuft of dhalmel hair

Notes

A quick note that 3 hides dropped from 1 dhalmel. That puts max hides at 3 now instead of 2.


Analysis

Data continues to be confusing, especially with meat drops. I've approached it a number of different ways, but every single time, the observed drop rate for 2x meat is far higher than theoretically predicted for 2 independant drops. One particular analysis, though, did lead to a possible solution.


Using the assumed independant drop rates (so for TH1, 17.5% for meat, 14% for hair), what are the odds of getting 1 meat + 1 hair compared to getting 2 meat? This question occured to me since it seemed like the former would be more common, but the latter showed up more often.

Odds of 1 meat + 1 hair: 49/1000
Odds of 2 meat: 30/1000

Observed 1 meat + 1 hair: 21/1000
Observed 2 meat: 74/1000

Not only reversed of expectations, but drasticly so. However it also led me to consider one other approach using the drop rates I have at the top that Kwontess was confused about (27% for any meat drop). Assuming that meat was a single drop item, what are the odds of any meat + hair combo?

Odds of n meat + 1 hair: 38/1000
Observed n meat + 1 hair: 38/1000


And finally there's a result that matches observations. The question is, does it hold up across all data sets?

@ TH0:

Odds: 10/1000
Observed: 9/1000

@ TH2:

Odds: 50/700
Observed: 51/700


And the answer is yes. (Unlike the 50%/50% per meat taken from the TH2 data, which fell apart in TH0 and TH1 scenarios.)


Can we confirm this with other mixtures?


TH0 meat + hide:

Odds: 1/1000
Observed: 0/1000


TH1 meat + hide:

Odds: 5/1000
Observed: 3/1000


TH2 meat + hide:

Odds: 7/700
Observed: 5/700


TH0 meat + femur:

Odds: 2/1000
Observed: 7/1000


TH1 meat + femur:

Odds: 12/1000
Observed: 33/1000


TH2 meat + femur:

Odds: 21/700
Observed: 45/700


Unfortunately.... No. Meat+Hide works, Meat+Femur doesn't. Why not?

Seems that femurs apparently are quite shy about showing up on their own. Predicted number of solo femurs with TH1 is 28, but only 12 are observed. 17 are expected with TH0, but only 12 drop solo. And with TH2, with 26 expected only 6 drop by themselves. While technically well within the margin of error, there may be something else going on here.

In fact, 45/54 giant femurs in the TH2 set drop in association with meat; 7/22 with TH0, and 33/46 with TH1. That seems to be an abnormally strong link, getting stronger at higher TH levels. Unfortunately it's been way too long since my stats classes, and I can't dig up the formulas for calculating correlation on stochastic data. For the moment, though, it looks like there's a dependency link between those two drops (ie: if you get a good cut of meat, there's a decent chance of keeping the bone to go with it, or something like that).




Conclusion 1

Multiple drops of items are generally not independant items in the loot pool, but rather a single item in the loot pool with the possibility of dropping more than 1 at a time.


Conclusion 2

Each single drop item in the pool has the ability to drop, at most, 3 of that item. It may also be further restricted to 1 or 2 of that item dropping.

Examples: 3 hides on dhalmels, 2 sets of 3 hides on rams.


Conclusion 3

There may be dependencies among the items in the loot pool. Unconfirmed, but the data is highly suggestive.



Treasure Hunter influence

If we take the given drop rates (disregarding number of items dropped), what is the apparent influence of TH?

Drop rates:

            TH0      TH1      TH2
Femur      2.2%     4.6%     7.7%
Grass      0.3%     0.6%     0.7%
Hair      11.3%    14.0%    18.6%
Hide       1.0%     1.7%     2.6%
Meat       9.0%    27.0%    38.6%

Increase per TH:

            TH1      TH2
Femur     +100%    +350%
Grass     +100%    +250%
Hair       +25%     +65%
Hide       +70%    +160%
Meat      +200%    +330%

The grass and hide drop rates are really too low to make decent estimates on, and the possible dependency issues on femurs means I'll drop those as well. Of the remainder, we see increases of 25% for hair vs 200% for meat with TH1, and 65% for hair vs 330% for meat with TH2. Neither drop follows any of the presumed drop rate increases for TH1 or TH2.

So at this point I still don't know how TH works, but at least we have some solid baselines for previously confusing information.