Talk:Honeycomb

Item rates

So, I opened 100 honeycombs, I got 35 honey, 37 dead bees, 17 pollen, 11 royal jelly. Not sure how helpful this would be, so just putting it out here. xD Acknowitall 21:13, 19 August 2010 (UTC)

So the success rates are approx. 35% Fresh Honey, 35% Dead Bee, 15% Pollen, 15% Royal Jelly ~kotarou3 ^TALK_CONTRIBS 11:31, 12 September 2010 (UTC)

I just got 5 Honey, 2 Royal Jelly, 2 Pollen and 1 Dead Bee from 10 combs. That makes it 40 honey, 38 bees, 19 pollen and 13 royal. I think the chances are more of 35% Honey and Bee, 20% Pollen and 10% Royal ~kotarou3 ^TALK_CONTRIBS 11:43, 12 September 2010 (UTC)

The results still aren't definitive, its looking like 35~40% Honey, 34~38% dead bee, 13~19% polen and 10~13% royal jelly. 200 more results should narrow it down. Tellos 14:12, 12 September 2010 (UTC)

With zeph's set of 200 below it comes out to a breakdown of 112, 86, 72, 40, for a total of 310.
Meaning Honey 36%, Bee 28%, Pollen 23%, Royal 13%. More testing needed? -- ladywinter ~{talk | contribs}~ 18:09, 13 September 2010 (UTC)

So far, as I'm doing this experiment, Royal Jelly and Pollen seem to be relatively constant. Dead Bees and Honey, on the other hand, seem to be fluctuating a lot. Will be testing more~ --Zephyre^TALK 18:35, 13 September 2010 (UTC)

Testing (Please don't touch this data)

Since my inventory doesn't allow it, I'm conducting a super number crunching experiment in 10 parts. --Zephyre^TALK 18:33, 13 September 2010 (UTC)

Test 1
Fresh Honey	Dead Bee	Pollen	Royal Jelly	Total
38	27	24	11	100
Test 2
Fresh Honey	Dead Bee	Pollen	Royal Jelly	Total
34	21	29	16	100
Test 3
Fresh Honey	Dead Bee	Pollen	Royal Jelly	Total
27	40	21	12	100
Test 4
Fresh Honey	Dead Bee	Pollen	Royal Jelly	Total
33	27	24	16	100
Test 5
Fresh Honey	Dead Bee	Pollen	Royal Jelly	Total
39	25	19	17	100
Test 6
Fresh Honey	Dead Bee	Pollen	Royal Jelly	Total
42	25	19	14	100

Okay, so I'm getting really bored smacking a tree a couple thousand times, so I'm going to stop now (lol). But taking the data I gathered, and Acknowitall's data, we come to the conclusion of:

Cumulative Tests
Fresh Honey	Dead Bee	Pollen	Royal Jelly	Grand Total
248	202	153	97	700
35.4%	28.9%	21.9%	13.8%
35%	30%	20%	15%

The last set of percentages is the "Theoretical Percentage". If 300 more honeycombs were ripped in half to reveal their insides, the resulting numbers would probably look something like that. --Zephyre^TALK 21:10, 13 September 2010 (UTC)

lol, but why does the theoretical percentage have to be a multiple of 5? XD I mean, if you look at the way elemental defense stacks up (it's bizarre if you actually get to the percentages of hard numbers) and the rate of shock page drop and metallurgy drop rates and the tiny multipliers on most repair cost increases (in enchants) that mabi doesn't necessarily go by round numbers, especially when you go into the nitty gritty of it. The raw percentages of your cumulative tests actually are similar to the theoretical numbers I posted with just two of your trials plus Acknowitall's data. =\ Why couldn't Royal jelly have a 13% drop rate? -- ladywinter ~{talk | contribs}~ 21:42, 13 September 2010 (UTC)

You didn't use the standard deviation to account for probable error. It should be closer to 37% honey, 27% dead bee, 22% pollen, 14% Jelly. The results are still not completely accurate (the values are approximately 80% accurate). In order to get the results to near exactness you'd need to do open ~250 more. Tellos 22:23, 13 September 2010 (UTC)

Factoring in previous tests it appears to be 38% honey, 27% dead bee, 22% pollen, 13% jelly (accurate to within +-.6%) Tellos 22:31, 13 September 2010 (UTC)

You forget though, that percentage is actually a number derived from a set of data formed by the average occurance of something for every hundred tests. Although that (<-) could have been worded much better, the gist of it is that the percentage is something we get from testing for every hundred of something. Hence "Per Cent" (cent is the french word for hundred). Though I do agree with you when you say that we should have an even 1000 for better test results, keep in mind that I did take the percentage of the numbers relative to the 700 tests I've done.
To answer Lady up there, I just like whole numbers a lot more :) And the 0.86% Drop Rate of Shock Page 2 seems to be a manual test that someone's done, most probably doing the Shadow Mission 1000 times and only get 86 pages. If that was the case, then I would assume that the drop rate would actually be around 1% with the fluctuation of the Random Number Generator that devCAT uses, and we all know how well that works out *cough*Ferghus. But then again, it could be a non-whole number as a percentage. Heck, each individual Honeycomb might even have a set RNG of its own. We'll never know ;) --Zephyre^TALK 22:41, 13 September 2010 (UTC)

The drop rate of shock and its location actually came from a Japanese Magazine that was advertising the new Sword of the Gods update. One of its articles talk about shock and it gave a nice chart on where it drop, difficulty, and the percentage. --Miyuna 23:19, 13 September 2010 (UTC)

Probability is the rate at which something occurs in the long run. Even if each comb had its own randomized RNG, if you take enough sets the data will normalize (that is to say it will begin to show a rate of occurance). The 5 multiple thing is just something that we as humans are comfortable with, most games program in probabilitys down to the first or second decimal. The prupose of a large number of trials is to minimize the chance that the probability rate is inacurate (much like it'd be innacurate to say that a coin toss will result in always head if a coin were tossed and landed on heads 3 times in a row). With 800~1000 trials we can easily find the average rate of obtaining certain outcomes for the honeycombs, ie the probability. The common arguement is based on the logical falacy "If something will occur 1 out of 3 times, then it will most certainaly occur after 3 attempts" when in actuality it'd have a 99% chance to have occured AT LEAST ONCE in all attempts (ie most certainlty occured) after 36 attempts. Also, a found percentage is said to be less accurate when the standard deviation between the sets is higher, and more accurate when it is lower. The numbers in the previous section are enforced by the numbers found in this section with the percents I listed as being MOST LIKELY accurate (that is to say that the chance that they are wrong is EXTREMELY low). Statistics are never 100% accurate, they just attempt to minimize any error. Also Zephyr, we would not assume it is "1%", in terms of statstics it'd be more like 2~.1% as that number of trials is insufficient to determine that small a percentage. Tellos 23:37, 13 September 2010 (UTC)

Don't have an interest in mass-testing this (and if I was a smart noodle, I probably would have recorded stats from when I was doing the r6 Campfire quest, but oh well.

https://imgur.com/a/WCjG4m5

20 Honeycomb;

• 2x Royal Jelly
• 7x Fresh Honey
• 2x Dead Bee
• 1x Pollen
• 8x Wax

Of the Wax rolls,

• 3x 5
• 1x 4
• 2x 3
• 1x 2
• 1x 1

Dunno if the wiki has a 'info gathering' google worksheet anywhere? Might be valuable to just have people dump stats (along w/ screenshots for proof maybe) and have auto-sum, as opposed to trying to do this math via a bunch of comments.

https://imgur.com/a/WKaSuls

19 Honeycomb;

• 1 Royal Jelly
• 5 Fresh Honey
• 2 Pollen
• 4 Dead Bee
• 7 Wax
  • 0x 5
  • 1x 4
  • 2x 3
  • 2x 2
  • 2x 1

I'll toss in mine but an image capture isn't gonna work cuz the log doesn't record 100 entries

100 Honeycomb:

• 28 Royal Jelly
• 12 Fresh Honey
• 12 Pollen
• 18 Dead Bee
• Wax (30 Draws):
  • 5x: 6
  • 4x: 7
  • 3x: 4
  • 2x: 6
  • 1x: 7

I collected a large amount of pollen for Fynn Crafting, so I decided to log my results and compile the data for all openings since the update that added wax. Below are the results.

Contributor	Date	Fresh Honey	Dead Bee	Royal Jelly	Pollen	Wax x 1	Wax x 2	Wax x 3	Wax x 4	Wax x 5
Bronzebreak	05-Nov-2020	7	2	2	1	1	1	2	1	3
Bronzebreak	10-Dec-2020	5	4	1	2	2	2	2	1	0
Jllai	26-Apr-2021	12	18	28	12	7	6	4	7	6
Hematin	11-Jun-2021	115	105	49	54	33	41	46	43	42
Hematin	19-Jun-2021	97	78	30	42	28	27	23	29	17
Hematin	20-Jun-2021	60	43	21	18	15	8	17	18	14
Hematin	30-Jun-2021	120	79	39	46	35	29	28	32	33
Hematin	04-Jul-2021	33	21	14	10	11	6	1	5	12
Hematin	29-Jan-2022	78	53	26	25	22	24	21	20	17
Hematin	30-Jan-2022	125	110	53	50	32	27	40	34	37
	Sum	652	513	263	260	186	171	184	190	181
	Sum (%)	25.1	19.7	10.1	10.0	7.2	6.6	7.1	7.3	7.0

With this data, assuming the developers round to 5% denominations as they tend to do with other content, I believe we can conclude that the rates are as follows:

• 25% Fresh Honey
• 20% Dead Bee
• 10% Royal Jelly
• 10% Pollen
• 35% Wax (with an even distribution between 1-5, inclusive)

If anyone else has a significant number of logged data, and is confident that it is free from statistical bias, feel free to reply to this comment and I will update it when I see it. For now, I will update the actual page with these assumed rates.