Lab Test 22 - Lucky Enchant Stones

[Juji]

~Request by SoulMagicX~
Testing: Lucky Enchant Stones - Armor and Weapon R and maybe another grade, pick one :) (not requesting rates of enchanting normally, just the "luck" in the Lucky Stones
Question: Test overenchant success/failure rate with/without stone.
Reason: There is a wide belief that these stones actually are the reverse, ie: they reduce your luck to the point that it is more successful without it.Item Tested: Lucky Enchant Stone: Enchant Weapon (R-Grade)
Weapon Test Trials: 50 Apocalypse Cutters Enchanted to +4 | 50 Apocalypse Cutters Enchanted to +4 with Lucky Enchant Stones
Armor Test Trials: 50 Twilight Breastplates Enchanted to +4 | 50 Twilight Breastplates Enchanted to +4 with Lucky Enchant Stones
Note: Due to time constraints, the number of weapon and armor enchants was set to 100 for this lab. This test does not accurately provide weapon and armor enchant rates.

Apocalypse Cutters Enchanted to +4
Success: 25
Fail: 25

Apocalypse Cutters Enchanted to +4 with Lucky Enchant Stones
Success: 32
Fail: 18

Twilight Breastplate Enchanted to +4
Success: 32
Fail: 18

Twilight Breastplate Enchanted to +4 with Lucky Enchant Stones
Success: 38
Fail: 12

[SoulMagicX]

Thanks Juji :) Good to know they are not bugged!

If possible could you run this test on the Lucky Armor stones too?

[Paul Senior]

lucky weapons really works, but the people in general don't trust in the lucky armor stone, can try a test with armor?

[Juji]

lucky weapons really works, but the people in general don't trust in the lucky armor stone, can try a test with armor?

Thanks Juji :) Good to know they are not bugged!

If possible could you run this test on the Lucky Armor stones too?
http://i50.tinypic.com/2e64j9f.png

[Laby]

Haha Juji you made my day :D

[Truxardor]

This
I just did the test, but enchanting 1,000 or even 500 is very time consuming just to show that these stones work. I did 100 instead. It may be low, but that is the most you guys will get out of me for this. Consider yourselves lucky! :rolleyes:
+ This
http://i50.tinypic.com/2e64j9f.png
Made my day.:D

[SoulMagicX]

http://i50.tinypic.com/2e64j9f.png

lol :D Thank you very much madface Juji! Now I can blow stuff up! lol :o

[OrochiGirl]

Lab Request Restrictions:
Due to game design and the intentions of developers and producers in the Lineage II franchise, there are some topics and features in the game for which the Juji's Lab testing is not provided:


No Quests (Requires extensive testing)
No Enchant Rates
No Critical Rates
No Raid Boss Drops
No Event Rewards
No XP Table Stats
No Requests for Development (Use the Game Suggestions section (http://boards.lineage2.com/forumdisplay.php?f=179))



You told me in a PM that the lucky stones fall under the "No Enchant Rates" category. WTH?

[Juji]

You told me in a PM that the lucky stones fall under the "No Enchant Rates" category. WTH?
Yes, that's right. However, I did not provide the rates for the stones and this test was to simply see if using the stones give a higher chance of success when enchanting.

[Untagged]

So to sum up:
enchanting weapons beyond +3 = 50%
enchanting armor beyond +3 = 64%
using lucky enchant stone on weapon = 64%
using lucky enchant stone on armor = 76%

or
Lucky enchant stone on weapon = 14%
Lucky enchant stone on armor = 12%

[SoulMagicX]

So to sum up:
enchanting weapons beyond +3 = 50%
enchanting armor beyond +3 = 64%
using lucky enchant stone on weapon = 64%
using lucky enchant stone on armor = 76%

or
Lucky enchant stone on weapon = 14%
Lucky enchant stone on armor = 12%

Not really, the test sample was far too small to give such conclusions but at least it shows the lucky stones do work as some were thinking they were bugged :)

[Untagged]

Just a conclusion based on Juji's numbers. If you have other test numbers to disprove then do so. Players have been writing on this forum that they think the enchant R grade weapon numbers are around 50% without the lucky stones. Juji's number confirm that, even with a small sample.
I agree that a larger sample would be better, but this is what we get, so barring new data, Juji's data and my statement stand as valid.

[Juji]

I'm afraid that SoulMagicX is correct. This test was too small to give an accurate enchant rate for weapons and armor.

[Untagged]

I'm afraid that SoulMagicX is correct. This test was too small to give an accurate enchant rate for weapons and armor.

You just invalidated your test by saying any numbers you posted may not be accurate. By your admission, lucky stones and normal enchant rates may be Korean Random.

Easy answer is to increase the sample size as you have expanded on other tests. The larger the sample size, the higher the degree of confidence in the statistics. Not to be rude, but you put forth the numbers, so the onus is on you to prove or disprove the data you posted. If 50 samples is not enough to guarantee significant percentile numbers, then increase the sample size to improve your degrees of confidence. I would suggest that any sample for any future testing you put forth be of 1000 samples (minimum) to achieve confidence in the statistics you post.

[OrochiGirl]

Yes, that's right. However, I did not provide the rates for the stones and this test was to simply see if using the stones give a higher chance of success when enchanting.

My PM didn't ask if you can test the lucky stone rates. My PM asked if a comparison (using a lucky stone vs. not using one) was allowed.

[Juji]

You just invalidated your test by saying any numbers you posted may not be accurate. By your admission, lucky stones and normal enchant rates may be Korean Random.

Easy answer is to increase the sample size as you have expanded on other tests. The larger the sample size, the higher the degree of confidence in the statistics. Not to be rude, but you put forth the numbers, so the onus is on you to prove or disprove the data you posted. If 50 samples is not enough to guarantee significant percentile numbers, then increase the sample size to improve your degrees of confidence. I would suggest that any sample for any future testing you put forth be of 1000 samples (minimum) to achieve confidence in the statistics you post.
I said that the test would not accurately give the enchant rates for weapon and armor, which was in my original post. This test was to determine the success/fail rate for using the stones and to verify that they are indeed working as intended. Of course, increasing the sample size to 1,000 would give a more precise number, but that is not feasible.

[Untagged]

My apologies, I read more into what was posted than was actually there.

[MichalisKrdm]

I don't really believe in Lucky Enchant Stones, not from the view that they don't work but from the view that they do not increase your luck. If karma doesn't want you to enchant something, you won't.
I made my Dual Blunts +7 without any Destruction/Blessed Scroll or Lucky Stone because of that reason.

@Juji Also, you should add a Like or Thumb up button here hahaha.
http://i.qkme.me/35f2uw.jpg

[SoulMagicX]

Juji! As a result of your test I had the nerve to try +4ing my armor tops!

Suceeded 5 R armor tops to +4 first try no fails with the Lucky Stones from this Draco event! :D

Have a Lucky Weapon R Scroll on my Rogue but already elemented and SA'd the dual daggers so Im chicken ><

btw this Draco is by far my favorite event Ever! :D Got so much useful stuff to get my toons fixed up and some things to sell (I'll never afford 10 of those Lvl5 Legendary Dyes but I got 3 of them to sell :D lol)

[AzMoDaN]

Is chance for mages weapons (apocalypse buster/caster) the same or not? I've heard it's lower…

[Osito2Dancer]

I'm afraid that SoulMagicX is correct. This test was too small to give an accurate enchant rate for weapons and armor.
But you can still apply statistical approaches to calculate an accurate enchant rate, by setting a standard mean study (the famous gaussian distribution N(U,O) that can be analyzed with stat-inference) and establish rates that will not be that far from the current setup of the server (as long as you did more than 30 tries, you can deduct it by the principle of "central limit theoretics").

[Juji]

But you can still apply statistical approaches to calculate an accurate enchant rate, by setting a standard mean study (the famous gaussian distribution N(U,O) that can be analyzed with stat-inference) and establish rates that will not be that far from the current setup of the server (as long as you did more than 30 tries, you can deduct it by the principle of "central limit theoretics").
I already posted the enchant rates in the Lab Requests topic. 1/3 for Armor and 2/3 for Weapons.

[Osito2Dancer]

I already posted the enchant rates in the Lab Requests topic. 1/3 for Armor and 2/3 for Weapons.
Still, your tests indicate that the armor chance is larger than 1/3 success rate (is rather 1/3 of failure as it seems). Weapons seem to be deviated from the 2/3 though, but its expectable. Also, you didnt try with magic weapons and fighter weapons, since they do have different enchant chances (or so i've "heard" :rolleyes:).

But i digress; to establish a numerical value on the chance of getting your +6 armor +15 weapon is being mean to those who find "enjoyable" to do over-enchanting (such as myself), one of the most fun aspects of the game is to break an important weapon at +4 and get a +15 D grade weapon meant to be Crystal Supply.

[BurnCycle]

Still, your tests indicate that the armor chance is larger than 1/3 success rate (is rather 1/3 of failure as it seems). Weapons seem to be deviated from the 2/3 though, but its expectable. Also, you didnt try with magic weapons and fighter weapons, since they do have different enchant chances (or so i've "heard" :rolleyes:).

But i digress; to establish a numerical value on the chance of getting your +6 armor +15 weapon is being mean to those who find "enjoyable" to do over-enchanting (such as myself), one of the most fun aspects of the game is to break an important weapon at +4 and get a +15 D grade weapon meant to be Crystal Supply.
From what i've tested (and i tested a lot) there's a different rate in armor enchant from +3 to +4 (i'm guessing it's 2/3) the rest is in line to what Juji said. (2/3 to weapon, 1/3 to armor higher then +4)
Why Juji still says the rate from +3 to +4 is 1/3, even after he kinda disproved in this topic it's a mystery to me. :D (Sorry Juji ;))

[Dreamer]

ChildofHeaven (GM at the Asia) made a test that showed a way lower OE rate. Especially once you OE the Armor +4 -> +5 and +5 -> +6.

[Yidao]

About halfway down here you will find the test:
http://tw.myblog.yahoo.com/l2_friend/article?mid=28816&prev=29088&next=28587&l=f&fid=5
一般狀況　means enchanting a Bound Apokalypse Cutter from +3 to +4 without Lucky Stone 200 times, success rate ５０。５％.
輔助石 means enchanting a Bound Apokalypse Cutter from +3 to +4 with Lucky Stone 200 times, success rate ５６。５％
Juji doesn't say what is written in the description of the Lucky Stones he used. Child of Heaven used Event Stones/武器用強化輔助石-R級(活動) which improve the chance by 10%. In Korea there are also Stones which improve the chance by 20%:
http://nshop.plaync.co.kr/lineage2/goods/detail?goodsNo=4633&categoryPath=1&sort=1&sortOrder=DESC

And that's where it gets tricky. Those 10% resp 20% are taken from the base probability of an unaided enchantment attempt. R-Grade physical weapons have a base probability of 50%. Add to that 10% of 50%, i.e. 50+5, and you are at ChildOfHeaven's result. Armor, on the other hand, has a much lower probability when you go beyond +4:
http://tw.myblog.yahoo.com/l2_friend/article?mid=16912&next=16405&l=f&fid=5
So enchanting a Twilight Breastplate only to +4 is a bit misleading. Since the base probability of going from +4 to +5 is only 33%, 10% of that added would only give you 36% :p

By the way, there seem to be people around who seriously believe that mage weapons have the same enchantment success rate as physical weapons. If Juji could add a test with a magical weapon (Apocalypse Buster or whatever) this might help to avoid a lot of frustration among new players :)

[piRule]

http://i50.tinypic.com/2e64j9f.png

AHAHHAHAHAHAHAHAHHAHAHAHAHAHAHHAHA
God, how ppl can be as stupid.

[Cegua]

this is synthethic test or what? how it posible that suces with weapon is 50%? i make 8 fail to enhcnat apo dual dagers to +4... 8 times.... and i know ppl who make 18 fails to make +4!! 18 times!! what is going here?

[Laby]

this is synthethic test or what? how it posible that suces with weapon is 50%? i make 8 fail to enhcnat apo dual dagers to +4... 8 times.... and i know ppl who make 18 fails to make +4!! 18 times!! what is going here?

Bad luck ?

Attribute stone success rate was 50% during Gracia Final. A lot of tests proved it, and some illegal and stolen server files showed it also. But I was able to fail 10 stones in a row, then success 10 stones in a row.

[Punky]

Is chance for mages weapons (apocalypse buster/caster) the same or not? I've heard it's lower…

I would like to know this myself...any chance of a test?

[KingBuff]

I would like to know this myself...any chance of a test?

+1 I would love to see some tests on mage weapons as well to satisfy my curiosity a bit.

[BurnCycle]

I already posted the enchant rates in the Lab Requests topic. 1/3 for Armor and 2/3 for Weapons.

Now that this myth created by juji was "destroyed" by the euro version of this lab test, (Would like to know why they let them test enchant rates but not here) can we at least get an apologie for misleading us into believe in false rates?

[Juji]

Now that this myth created by juji was "destroyed" by the euro version of this lab test, (Would like to know why they let them test enchant rates but not here) can we at least get an apologie for misleading us into believe in false rates?
What myth? You still have a 2/3 chance at each enchant stage to get a successful enchant. So if you're going from +3 to +4 it should be 2/3 then for +3 to +5 it is 2/3 * 2/3. As you go higher it gets harder. The enchant rate for magic weapons is much lower than 2/3, but I don't have the exact number for that one.

If you wanted to do a REAL test for specific enchant probabilities for each level then you would have to enchant 10,000 weapons or higher. Anything lower will not yield the exact probabilities. This is the reason why I never bothered with completing that request.

[Laby]

What myth? You still have a 2/3 chance at each enchant stage to get a successful enchant. So if you're going from +3 to +4 it should be 2/3 then for +3 to +5 it is 2/3 * 2/3. As you go higher it gets harder. The enchant rate for magic weapons is much lower than 2/3, but I don't have the exact number for that one.

If you wanted to do a REAL test for specific enchant probabilities for each level then you would have to enchant 10,000 weapons or higher. Anything lower will not yield the exact probabilities. This is the reason why I never bothered with completing that request.

https://forum.4game.com/showthread.php?t=14278

1000 weapons got OE'd : 1000 bow, 1000 retributer

[Juji]

https://forum.4game.com/showthread.php?t=14278

1000 weapons got OE'd : 1000 bow, 1000 retributer
Again, 1,000 weapons is not enough for an accurate test. The enchant rate is the chance you have to get a successful enchant at each level and this does not change. What does change is the probability from going to different enchant levels.

For Example:
+3 to +4 is 2/3
+3 to +5 is 2/3 * 2/3
+3 to +6 is 2/3 * 2/3 * 2/3

[Spudnik]

I don't really believe in Lucky Enchant Stones, not from the view that they don't work but from the view that they do not increase your luck. If karma doesn't want you to enchant something, you won't.
I made my Dual Blunts +7 without any Destruction/Blessed Scroll or Lucky Stone because of that reason.

@Juji Also, you should add a Like or Thumb up button here hahaha.
http://i.qkme.me/35f2uw.jpg

sure. but when *I* do it, its trolling.

[Laby]

I don't know if you took some time to read the link in fact There are numbers for going from +3 to +4, then +3 to +5, +3 to +6 & so, exactly as you show your math.

But the numbers are far from this value.

[BurnCycle]

He probably didnt read at all. (Juji, doesn't look like you :D)
The funny thing is that Juji own tests supported the results on this lab made by Innova ppl. He just keep saying the % was the same.
We now know that the % of going from +19 to +20 is 20% chance in weapons. (20% is a bit diferent from 2/3... maybe i don't know math at all)

[Juji]

I don't know if you took some time to read the link in fact There are numbers for going from +3 to +4, then +3 to +5, +3 to +6 & so, exactly as you show your math.

But the numbers are far from this value.
I did read it and the numbers are far from the value I gave because the number of weapons enchanted in that test is not enough to give accurate data. If I were to enchant over 10,000 weapons then the number I would get for +3 to +4 would be around 66%.

[BurnCycle]

I did read it and the numbers are far from the value I gave because the number of weapons enchanted in that test is not enough to give accurate data. If I were to enchant over 10,000 weapons then the number I would get for +3 to +4 would be around 66%.

I also put the data in a table:


Enchantment Chance
+3 - +4 52,5%
+3 -+5 26,6%
+3 - +6 14,5%
+3 - +7 4,6%
+3 - +8 1,5%
+3 - +9 0,4%
+3 - +10 0,2%

And here is the success rate for each enchantment stage:

Stage Chance
+3 - +4 52,5%
+4 - +5 50,7%
+5 - +6 54,5%
+6 - +7 31,7%
+7 - +8 32,6%
+8 - +9 26,7%
+9 - +10 50%



2/3 is NOT 52%. I would advise you to check your math.

Edit: A real 2/3 chance is more in this test:
https://forum.4game.com/showthread.php?t=14437

With your theory this is an aberration:
https://forum.4game.com/showthread.php?t=13922

Honestly i dont understand why you keep insisting in something that simply is not truth.

[Juji]

2/3 is NOT 52%. I would advise you to check your math.1,000 weapons is NOT 10,000 weapons.

Stage Chance
+3 - +4 52,5%
+4 - +5 50,7%
+5 - +6 54,5%
+6 - +7 31,7%
+7 - +8 32,6%
+8 - +9 26,7%
+9 - +10 50%
This should be 66% for all.

[BurnCycle]

1,000 weapons is NOT 10,000 weapons.


This should be 66% for all.

lol. Never seen a case of denial like this!
Basically your saying that 1000 is to low test case. So if you do the 10k test and see the results are the same you will say you need a 100k test? I find your reaction hard to believe! :confused:

EDIT: I had some statistic classes and in my honest opinion a 1k case study is actually a decent case study that should be close to correct result.

EDIT2: according to you a 1k case study would give a more accurate result:
I said that the test would not accurately give the enchant rates for weapon and armor, which was in my original post. This test was to determine the success/fail rate for using the stones and to verify that they are indeed working as intended. Of course, increasing the sample size to 1,000 would give a more precise number, but that is not feasible.

[Juji]

lol. Never seen a case of denial like this!
Basically your saying that 1000 is to low test case. So if you do the 10k test and see the results are the same you will say you need a 100k test? I find your reaction hard to believe! :confused:

EDIT: I had some statistic classes and in my honest opinion a 1k case study is actually a decent case study that should be close to correct result.

EDIT2: according to you a 1k case study would give a more accurate result:Their test was to determine whether the enchant rate is lower for magic weapons and it is, but if they wanted to accurately test the probability (not the rate) for each enchant level then they would need to enchant more weapons.

Indeed, at that time I thought 1,000 was the right number to test for enchant rates and probabilities, however that is not the case according to the information I have currently.

[BurnCycle]

Their test was to determine whether the enchant rate is lower for magic weapons and it is, but if they wanted to accurately test the probability (not the rate) for each enchant level then they would need to enchant more weapons.

Indeed, at that time I thought 1,000 was the right number to test for enchant rates and probabilities, however that is not the case according to the information I have currently.

It's funny that for S grade weapons the number of case tests seem to be enough to get the result you want. (clearly the enchant rate depends on the weapon/grade you using)
https://forum.4game.com/showthread.php?t=14437

What you saying is a statistic aberration. There is no way that with 10k case sample you will get a % close to the 2/3 you mention. (around 66.6%)

I don't understand why you insisting in keep this idea that is more then disproved with this 3 tests made, looks like there's clearly some other obscure reasons behind your position and makes me sad you keep trying to give the players an idea that is simply not true.

EDIT: Forgot to mention that from +3 to +4 the case study is THE SAME for both, physical and magician. (1k weapons) the rate is clearly different. (https://forum.4game.com/showthread.php?t=14437)