Talk:Mystic Clover/Drop rate

Notes

As far as I'm aware it is NOT possible to obtain T5 materials. I've added a table I think would be more useful for recording findings, but it might require doing the same for the 10 clover recipie.

Not sure if research is still going on, but here are my findings:

• 01:10x = 20 Crystals
• 02:1x = CLOVER
• 03:1x = CLOVER
• 04:1x = 2 Scales
• 05:1x = 2 dust
• 06:1x = 2 dust
• 07:1x = 4 totem
• 08:1x = 3 totem
• 09:1x = CLOVER
• 10:1x = CLOVER
• 11:1x = 1 bolt
• 12:1x = hidden treasure
• 13:1x = 2 planks
• 14:1x = CLOVER
• 15:1x = Corrupted Lodestone
• 16:1x = CLOVER
• 17:1x = 3 claws
• 18:1x = CLOVER
• 19:1x = 4 claws
• 20:1x = 2 scales
• 21:1x = 3 claws
• 22:1x = 2 ecto
• 23:1x = destroyer lodestone
• 24:1x = 3 claws
• 25:1x = 1 plank
• 26:1x = 2 ingots
• 27:1x = CLOVER
• 28:1x = 3 bolts
• 29:1x = CLOVER
• 30:1x = 3 bolts

76.105.103.27 10:51, 21 January 2013 (UTC)

Refine t6 mats

The page says 1-21 refined t6 mats. I am pretty sure that 21 refined mats do not come out from a 1x clover recipe. I do not know what the high end is though because I finished my clovers a while ago. C The 12:25, 12 February 2013 (UTC)

Isn't Putrid Essence basically a lodestone also? Why do we have it separated out? - Cait (talk) 11:55, 26 May 2013 (UTC)

Economic drop rate data

I recorded how many items I received after using the forge 100x, 100x, and 56x. I didn't bother recording whether I got 2 of something or 4. Here are my findings:

256 total attempts

• 75 clovers, i.e. 29.3%
• 271 fine mats, i.e. 1.06 per attempt (in alpha order of blood, bone, claw, dust, fang, scale, totem, venom: 35, 20, 43, 24, 37, 28, 53, 31).
• 122 common mats, i.e. 0.48 per attempt (in order of ancient, gossamer, leather, ori: 25, 44, 27, 26)
• 8 lodestones, i.e. 0.031 per attempt (1 corrupted, 4 crystal, 3 destroyer)
• 36 "none of above" (4 putrids, 15 coins, 8 ecto, 5 treasure, 4 obi)

Costs

At the time of posting, each attempt costs roughly 22s (ecto + mystic coin), 2,100 karma, and some skill points.

Output value, ~28s

The rough value of the output depends a lot on luck, whether you get mostly blood vs dust, ori vs leather, and how many lodestones. Assuming even distributions (which, even this small sample shows to be a bad assumption), the value is approximately:

• 21s in fine mats
• 1s in common mats
• 4s in lodestones
• 1s in "other"

Ignoring skill points and the desired clovers, this ends up delivering about 6s per 2,100 karma or ~2.8s per 1,000 karma. This compares to 3.6s per 1,000 karma via Lost Orrian Jewelry Boxes which is a lot less effort. 75.36.176.4 00:17, 30 June 2013 (UTC)

spoiled data

→ moved from Talk:Mystic Clover#spoiled data

Though the conclusions in the article made from the data might be a bit too mathematically for most people, I (as a maths student) like them. However, the data seems to have some errors, or rather: is lacking some review. I am pretty sure a few people used the template wrong and did not enter how many times they got something, but wrote done what their stack sizes said in the end. I did not give the 1-clover-table an in-depth look, but I saw two examples which struck me (clover rate wise): 31 out of 199 (~15.6%) by "VBMeireles" and 71 out of 513 (~13.8%) by Somohexual (split over 3 lines). With sample sizes of that magnitude in both cases, there is no way those are correct. Noticing that their mystic coin amounts are always divisible by 5 and their ecto/crystal/obsi amounts always by 2 adds the final piece. All other entries seemed reasonable on the first look, or had such a small size that nothing can be said. When leaving only those two candidates out, our total data changes from 1241/4422 (~28.1%) to 1139/3710 (~30.7%). However, I even tried to correct their data, and it worked quite fine: Dividing coins by 5, ectos/crystals/obsis by 2 and both material groups by 3 (guessed average of the 1-5 and 2-4 distribution), I get 31/88 (~35.2%) for VBMeireles and 71/229 (31.0%) for Somohexual. Including that in the totals, we get 1241/4027 (~30.8%). Only 0.1%-points difference compared to leaving them out. Also, the 1 and 10 recipe-data's confidence interval now overlap again. Tadaaa, or something. --aRTy 10:30, 5 October 2013 (UTC)

I reviewed the data again. Last time I just guessed the averages for fine and refined material with 3, this time I adjusted them until the data of both contributors matched the rough global average of 5% (for each item, both for fine and refined). The values were 3 (fine) and 2.4 (refined). This left me with a total of 4047, which includes 6 tries someone added to the table since my last post above. We end up with 1243/4047 ≈ 30,71%. I accordingly edited the article and also updated the calculations for the 10 recipe, as the collected data changed significantly.

I will now adjust the problematic data in the table, so future analyses are easier and readers find percentages in the table which actually match the article. --aRTy 02:37, 9 October 2013 (UTC)

I happened to notice one more candidate: Moiax with 7 clovers out of 73 tries, which is out of a 95% interval by far (~15 clover would be the lower end). The result is at the 3.9σ mark, so below 0.01% chance. Only flaw is the crystal amount of 1, which is not even, but I guess it could be a mistake, for instance if there were crystals in the inventory already.

If no one objects, I'd adjust that line too and merge it with the other reviewed lines.

As a side note: If anyone commented on this topic in general, I'd really appreciate it. Though I think it is reasonable to review and correct the data somehow, it's better with backup. --aRTy 16:53, 17 October 2013 (UTC)

Rearrange the table?

I'd like to change the order of the items in the table for better readability and easier comparision of the percentages by arranging them according to the "Alternate drops" in the mystic clover article (or like the "Possible drops" section below the tables). Mainly it's about having the 8 fine materials (bone, blood etc.) and the 4 refined materials (Wood, Bolt, Leather, Ingot) next to each other (not merged into a single column each). I think changing the table header and the copy-paste stencil in the code were sufficient. I'd say that also changing the order of every data line just to have the code match the display order is only of little benefit compared to the effort. --aRTy 17:14, 17 October 2013 (UTC)

you can rearrange the columns as you wish. The order of parameters doesnt matter and not enforced because of that. Changing the stencil is suggested.--Relyk ~ talk < 17:20, 17 October 2013 (UTC)

Details moved from main page

"Drop rate research shows the 1 clover recipe has a 30.71% chance of dropping in a sample of 4070 (+/- 1.42%, 95% confidence interval). Drop rate research shows the 10 clover recipe has a 31.71% chance of dropping in a sample of 615 (+/- 3.68%, 95% confidence interval). Since these values don't differ a lot and more importantly, their confidence intervals overlap largely, one can assume that both recipes produce clover with an identical chance (in theory), somewhere between roughly 30-32%. "

I moved this from the main article because it contains details covered on the drop rate article. Since the main article links this one, I think we can just post that both recipes have a 30-32% chance of dropping clover. That's really all people care about, as long as the wiki has or links to supporting evidence. 75.37.16.111 16:55, 21 October 2013 (UTC)

Submission from IP 71.75

{{SDRL | clover = 32 | blood = 22 | bone = 11 | claw = 18 | dust = 8 | fang = 12 | scale = 19 | totem = 17 | sac = 25
       | bolt = 4 | ori = 5 | leather = 6 | wood = 15 | ecto = 2 | lodestone = 7 | putrid = 0 | shard = 0 | coin = 0 | crystal = 2 | treasure = 2 | --[[Special:Contributions/71.75.110.206|71.75.110.206]] 17:52, 29 October 2020 (UTC) }}

I've removed this row as the edit summary stated the above data was supposed to be for 100 rolls of the "1 Clover" recipe. Data hasn't been recorded as # occurences, it's been recorded for number of items in the inventory. I could correct this, however I could only guess the amounts received in each roll, but this would be imo misleading and potentially mess up the other totals.

Example corrected data:

Clover 	32
Blood 	6
Bone 	4
Claw 	6
Dust 	3
Fang 	4
Scale 	6
Totem 	5
Venom 	7
Bolt 	2
Ingot 	3
Leather 	4
Wood 	8
Ecto 	1
Lodestone 	7
Putrid 	0
Shard 	0
Coin 	0
Crystal 	0
Treasure 	2

-Chieftain Alex 23:58, 29 October 2020 (UTC)