I have managed to fail 7 times in a row in tat neckless!! and by in row i mean Click Click Click Click Click Click Click! also it was a month that i crafted 18 items all S + and no fails!!
Just luck my friend! Dont get upset!
Ive failed 5 out of 6 crafts on various pieces for the set..
It happens with every 27th player if the success rate is 60%. And every 250th player fails all 6 crafts.
That's one of the things that contribute to the infamous Korean Random in a Korean MMO.
you have to test craft rate for the whole server not for just one char ..
for example i have recently crafted whole dynasty robe set without fails
and you failed the whole set
if we add these two cases we get correct success ratio
if you will craft really lot then eventually you will get a few successes in a row which will compensate for your lots of fails.
the problem is that these successes might be for items worth a lot less than your fails, but technically success rate will be the same still.
this is why this system is r.tarded. they should just double production cost in both key mats and resources and recipes and make it 100%. or make it a choice for those who are willing to risk for possible profit - either risk and spend 1x all stuff for 60% or spend 2x or even 3x but craft with 100% probability.
for a list. They confirm many things players suspected for a long time, especially about OE success rates. I won't quote everything, but some success rates they calculated from hundreds of OE attempts were e.g.:
Most knew this already from own experience (e.g. "if you want OE nuker weapon, then enchant a dagger and trade for a nuker weapon at mammon"), but it is nice to see this finally confirmed by NCSoft employees in their own tests, instead of constantly being pointed at the "all OE is 2/3 success rate" crap.
Success Rate and Success Chance are not defined the same. You could have a 99.99% chance to succeed and still fail 100% of the time on 1,000 tries. I don't know why you people keep going on about this crap.
Yes, but the CHANCE is still nice to know. As you can see e.g. from the official experiments linked above, the OE chance for nuker weapons is lower than stated in game on the enchant scrolls. So it could very well be that the 60% chance for crafts is wrong, too (though apparently it isn't). This has nothing to do with a player being totally unlucky, that can always happen. But if the chance has been experimentally proven to be 60%, then that player at least knows that it's not the game itself which is broken and can stop sending support tickets :)
Success Rate and Success Chance are not defined the same. You could have a 99.99% chance to succeed and still fail 100% of the time on 1,000 tries. I don't know why you people keep going on about this crap.
They keep going on because they think they understand statistical terms, but they don't. For example, they think they understand the term 'random' from it's use in popular culture, but statistics defines it precisely and technically. So the behavior in L2 follows the statistical definition of 'Random' but not the popular definitiion, and non-quant L2 players get their shorts all bundled up as a result.
I suspect that most who fuss about this don't really understand the concept of statistical independence. Somewhere buried in their mind is an expectation that a series of outcomes (usually 'fails') influences the next trial.
The whole debate probably arises because technical terms, such as random and independent, are also used in popular speech. If that were not the case, people would not misunderstand as easily. If, for example, we were talking about calculating gravitational slings in astronomy, folks would again misinterprete technical words and misunderstand what was going on...even though what was going on was mathematically correct.
I am sure that those who do not have a graduate degree in statistics will disagree with what I say. But those who do have such a degree will understand.
Yes, but the CHANCE is still nice to know. As you can see e.g. from the official experiments linked above, the OE chance for nuker weapons is lower than stated in game on the enchant scrolls. So it could very well be that the 60% chance for crafts is wrong, too (though apparently it isn't). This has nothing to do with a player being totally unlucky, that can always happen. But if the chance has been experimentally proven to be 60%, then that player at least knows that it's not the game itself which is broken and can stop sending support tickets :)
You're still confusing success chance with success rate. A 60% success chance (success chance) does not mean you will succeed 60% of the time (success rate). The experiments in question only show a success rate, that success rate does not dictate success chance.
Take a 2-sided coin, one side heads the other side tails. Flip it once. There is a 50% chance it will land heads up. Flip the same coin 9 more times (10 total flips). Write down the results for all 10 flips. Assume heads up is success.
I did my little experiment while writing this. My coin landed tails up 6 times, heads up 4 times. According to you it should have landed heads up 50% of the time, or 5 times. It did not.
If you doubt my numbers then collect your own. I will state, despite all this, that is indeed possible to have a 50% success rate, but not probable.
A 60% chance ought to be a 60% chance not some average done on an entire server.. thats crazy... and to complicated.. i dont remotely believe that.
You fail statistics.
I wonder if NCSoft implemented a "fun" distribution like Cauchy in there (instead of the more common Gaussian). The results from this one tend to really confuse laymen. :D
EDIT: Specifically, distributing the amount of successes per person according to a Beta-Binomial distribution (means, if you fail, you're more likely to fail the next few; and if you succeed, you're more likely to succeed the next few as well) without changing the mean would be downright evil. :D
You're still confusing success chance with success rate. A 60% success chance (success chance) does not mean you will succeed 60% of the time (success rate). The experiments in question only show a success rate, that success rate does not dictate success chance.
Take a 2-sided coin, one side heads the other side tails. Flip it once. There is a 50% chance it will land heads up. Flip the same coin 9 more times (10 total flips). Write down the results for all 10 flips. Assume heads up is success.
I did my little experiment while writing this. My coin landed tails up 6 times, heads up 4 times. According to you it should have landed heads up 50% of the time, or 5 times. It did not.
If you doubt my numbers then collect your own. I will state, despite all this, that is indeed possible to have a 50% success rate, but not probable.
There is no need to explain this to me, I KNOW WHAT YOU MEAN >.<
I make it easy for you, so that YOU know what I mean, although I am sure we just mean the same thing.
At some point, somewhere in the game, there HAS to be some code which
a.) rolls a random number
and
b.) checks if this number is equal to or lower than 60% for a craft or not (success or fail)
We agree so far? Good.
Now if you do this often enough (I would say for something like a 60% craft 100-200 crafts are enough) and take note of the successes and fails, you get a good enough idea if the game really DOES use the 60% as advertised or a different number. By this I mean you can see if it is around 60% or maybe only 40%. The more attempts you do, the closer you should end up to whatever success % the game actually uses.
Agreed? Good.
That's what I mean when I wrote
Yes, but the CHANCE is still nice to know. As you can see e.g. from the official experiments linked above, the OE chance for nuker weapons is lower than stated in game on the enchant scrolls.The Korean NCSoft employees who do the "L2 Laboratory" did a few hundred OE attempts on their PTS and found out that the OE for mage weapons is in fact closer to 40% instead of the 66% the enchant scrolls claim. They did a couple hundred enchants, and if you after all those get a ratio success:fails which is SIGNIFICANTLY different from the claimed ratio, it shows that it cannot be 66%. Even taking into consideration that one enchant attempt has no influence on the next one and that "<beep> happens" and you can get 20 fails in a row.
That's what I meant - that an experiment with a couple hundred crafts or enchant is indeed interesting because obviously (as was proven in Korea) some of the information in game is wrong. The difference is just too big. Even a "real" statistics expert will admit that it is more probable than not that e.g. the success rate for enchanting armor from +5 to +6 is indeed not 66% if out of 300 attempts only 25% succeed.
Please note that I avoided the words "chance" and "rate" as much as I could so that you do not go off on another rant and claim I am an idiot who thinks that if I throw a coin ten times I HAVE to get 5 heads and 5 tails.
hehehe funny as hell when I remember how people were arguing that mage weapons has same OE chance as fighter ones ignoring the fact of l2wh statistics + hundreds of OE from PTS's :)
Btw about 60% crafting thingie... My chance is smth like 90% or smth. Haven't failled Tallum heawy at all, making IC failed only gloves once, crafting weapon 1 FB failed other succeeded as well as DB, making TTS haven't failed anything. Even made masterwork ring.
Now new and improved. However, knowing NCSoft, it is something like this
public boolean enchant(int directionfacing, int typeofweapon, int servertime, string charactersname, boolean promisedfirstbornchild, boolean prayedtogodforsuccess, int currentenchant){
At some point, somewhere in the game, there HAS to be some code which
a.) rolls a random number
and
b.) checks if this number is equal to or lower than 60% for a craft or not (success or fail)
We agree so far? Good.
I'll go a long with that. Let's further this and say an invisible wheel is spun, this wheel has 99 numbers on it, 33 of those numbers are fail, the other 66 are success. Your success chance is 2 in 3. This does not, however, mean you will succeed 2/3 of the time.
Now if you do this often enough (I would say for something like a 60% craft 100-200 crafts are enough) and take note of the successes and fails, you get a good enough idea if the game really DOES use the 60% as advertised or a different number. By this I mean you can see if it is around 60% or maybe only 40%. The more attempts you do, the closer you should end up to whatever success % the game actually uses.
Agreed? Good.
Not agreed. In my wheel you have a 2/3 chance to succeed per try but you may only succeed 1/3 of the time on 1,000 tries. Does this mean not enough tries were made?
The part I emboldened only appears true to those people who have that happen, otherwise it appears false. It is not a constant.
That's what I mean when I wrote
The Korean NCSoft employees who do the "L2 Laboratory" did a few hundred OE attempts on their PTS and found out that the OE for mage weapons is in fact closer to 40% instead of the 66% the enchant scrolls claim.
The success rate is 40% here, the success chance is 66%. Different items being defined here.
They did a couple hundred enchants, and if you after all those get a ratio success:fails which is SIGNIFICANTLY different from the claimed ratio, it shows that it cannot be 66%. Even taking into consideration that one enchant attempt has no influence on the next one and that "<beep> happens" and you can get 20 fails in a row.
That's what I meant - that an experiment with a couple hundred crafts or enchant is indeed interesting because obviously (as was proven in Korea) some of the information in game is wrong. The difference is just too big. Even a "real" statistics expert will admit that it is more probable than not that e.g. the success rate for enchanting armor from +5 to +6 is indeed not 66% if out of 300 attempts only 25% succeed.
A success rate of 40% does not mean the success chance somehow magically changes to 40%.
The Enchant scrolls do not say, "One in three enchants will fail (rate)." They say, "There is a 1 in 3 chance to fail (chance)."
You're still confusing Success Chance with Success Rate.
So you disagree that an experiment with several hundred or maybe thousand enchants/crafts/whatever will show if the 1/3 or 40% in the recipe/enchant description are correct or not? After that many tries, everybody would except the success rate and the success chance to be similar unless something is seriously wrong.
Look e.g. at the other weapons they enchanted. Bows had a 66% chance and a 68% rate. Melee swords 66% chance, 70% success rate. Nuker mace CLAIMED 66% success chance, 43% success rate. Nuker staff 66% success chance, 39% success rate.
Even a drunk clown would see that there is something wrong there with the success rate vs. chance of nuker weapons. This cannot be just bad luck - ca. 25% difference is significant when the melee weapons were just 2-4% off. With a sample size of 300 you SHOULD be within 5% or so of the real chance.
Yes, it is possible to enchant 1000 times and fail 999 times. It really is. But if for the last few years it is common knowledge that nuker weapons are harder to enchant and now even a test by NCSoft employees themselves shows it, it is not simply coincidence but the enchant scroll description is simply wrong.
So, to come back to what I tried to point out in my first post before all this - the OP asked for a test, and that is a good idea. Luckily it was already done, so we now at least know that the craft percentages are correct and the OP was just unlucky. As opposed to the enchant %, where the descriptions in game are wrong.
The success rate is 40% here,
"Success rate" is not a strict term, it can be used to express many different things. The term you would want to use here is "sample mean".
the success chance is 66%.
How would you know?
Yes, it is possible to enchant 1000 times and fail 999 times. It really is. But if for the last few years it is common knowledge that nuker weapons are harder to enchant and now even a test by NCSoft employees themselves shows it, it is not simply coincidence but the enchant scroll description is simply wrong.
I admit, quite readily, that the chance for success on one weapon type may well be different than on another type. The success rate of enchants does not prove this however. All that proves is people have bad luck enchanting one weapon type compared to another type.
Let's say the success chance on a mage blunt is indeed 40% (or even approximately 40%). This does not dictate that your rate of success will be 40%. If you succeed one time on only one try then obviously your rate of success is 100%. If you fail one time on only one try then obviously your success rate is 0%. If you succeed on 36 of 50 tries then your success rate is 72%. If you succeed 42 times of 136 tries then your success rate is ~30%. Success Chance does not determine your Success Rate.
I admit, quite readily, that the chance for success on one weapon type may well be different than on another type. The success rate of enchants does not prove this however. All that proves is people have bad luck enchanting one weapon type compared to another type.
Here are the words to google for you: "binomial proportion confidence interval". This provides as much proof as possible when using statistics. If you don't like it, then you should not like any other statistical proofs as well.
Even if you could reasonably calculate a certain success rate based on a certain success chance (because of your success rate thus far, or w/e), nothing dictates that is what you will get. Just because you succeeded on the last 10 tries there is no guarantee you will succeed on the next, unless your chance is exactly 100%.
Some people have tried 10 times to enchant a skill to +1 with a 98% chance for success. They failed every time, that is a 0% success rate. No matter what kind of formula you can concoct there is no guarantee for success on a chance unless your chance is exactly 100%.
Even if you could reasonably calculate a certain success rate based on a certain success chance (because of your success rate thus far, or w/e), nothing dictates that is what you will get. Just because you succeeded on the last 10 tries there is no guarantee you will succeed on the next, unless your chance is exactly 100%.
Sorry, but it's not what we are speaking about.
Some people have tried 10 times to enchant a skill to +1 with a 98% chance for success. They failed every time, that is a 0% success rate.
They lied to you.
However, every 2500th player (if there are so many players left) will fail their first two 98% enchant attempts.
No matter what kind of formula you can concoct there is no guarantee for success on a chance unless your chance is exactly 100%.
There's no guarantee that NCsoft won't start wiping the servers right at the moment when you start your enchant attempt. So what?
I just checked various recipes on my dwarf. Those recipes do, in fact, say, "The success rate is x%." It should say, "The success chance is X%," unless there is some variable in the coding that forces a server wide success rate. So I wonder then, even if one person succeeds only 2% of the time while another person succeeds 78% of the time, does a server wide variable exist that forces a certain server wide success rate?
I also checked Enchant Scrolls, I did not see anywhere on them the dictation of chance nor rate. The info I'm thinking of, in regards to enchants, is from the "Useful Information" when you log in. I will not log in several times just to figure out exactly what it says, but it does say something to the effect of "When enchanting over +3 there is a 1 in 3 chance of failing."
I stand corrected, given my new understanding of what's being said in this thread, I would definitely want NC to figure out if the success rate in question is correct. However, if the wording is supposed to be "chance" instead of "rate" then it needs to be changed.
NCSoft can (and probably does on a regular basis) run a report detailing exactly what the success rates of various variables were. Enchant scrolls of different levels on different types of targets, win % of different class match ups in Oly at different levels and level disparities, number of sell shops in Giran at different days and times....if you can come up with the question they can answer it exactly.
These reports would not be in the form of a probability distribution but rather facts. They are not based on samples, they are based on 100% of all attempts.
So they probably do the OP's suggested test regularly. They are under no obligation to tell us the results of any of this and it seems to be their policy not to, which is fine with me.
Do you all remember waaaay back, probably C2, they did share some information on the number of red players, number of PKs, number of class changes, &tc?
NCSoft can (and probably does on a regular basis) run a report detailing exactly what the success rates of various variables were. Enchant scrolls of different levels on different types of targets, win % of different class match ups in Oly at different levels and level disparities, number of sell shops in Giran at different days and times....if you can come up with the question they can answer it exactly.
These reports would not be in the form of a probability distribution but rather facts. They are not based on samples, they are based on 100% of all attempts.
So they probably do the OP's suggested test regularly. They are under no obligation to tell us the results of any of this and it seems to be their policy not to, which is fine with me.
Do you all remember waaaay back, probably C2, they did share some information on the number of red players, number of PKs, number of class changes, &tc?
You are assuming that they log that information, and if they do retain it for long periods of time.
They are under no obligation to tell us the results of any of this and it seems to be their policy not to, which is fine with me.
That is correct. However, we do have the right to ask, and the need to know, if the information we are provided is correct if we are to play the game to our fullest potential. There is information on this website that does not match the information in the game, so how do we know what's correct? Also, even the GM's can't agree on what is and isn't correct. I don't think anyone is asking for coding information... but if information that we receive is incorrect then we should ask for, and be provided with, the correct information.
I also checked Enchant Scrolls, I did not see anywhere on them the dictation of chance nor rate. The info I'm thinking of, in regards to enchants, is from the "Useful Information" when you log in. I will not log in several times just to figure out exactly what it says, but it does say something to the effect of "When enchanting over +3 there is a 1 in 3 chance of failing."
Yes, it was in the loading screen info. I thought I had made a screenie of it, but I can't find it in my screenshot collection.
The links that have been already given to you (which you refused to check) had the experiment listed.
The Lineage II Laboratory which is doing quite good tests regarding game mechanics did 100 item crafts (Dynasty Mace, Dynasty Breastplate, Dynasty jewels).
They confirmed that the success with these crafts was very close to 60 out of 100 tries.
I'll go a long with that. Let's further this and say an invisible wheel is spun, this wheel has 99 numbers on it, 33 of those numbers are fail, the other 66 are success. Your success chance is 2 in 3. This does not, however, mean you will succeed 2/3 of the time.
Not agreed. In my wheel you have a 2/3 chance to succeed per try but you may only succeed 1/3 of the time on 1,000 tries. Does this mean not enough tries were made?
The part I emboldened only appears true to those people who have that happen, otherwise it appears false. It is not a constant.
The success rate is 40% here, the success chance is 66%. Different items being defined here.
A success rate of 40% does not mean the success chance somehow magically changes to 40%.
The Enchant scrolls do not say, "One in three enchants will fail (rate)." They say, "There is a 1 in 3 chance to fail (chance)."
You're still confusing Success Chance with Success Rate.
I would like to suggest for you to read some Statistics 101 before commenting smth like this in feature.
I just checked various recipes on my dwarf. Those recipes do, in fact, say, "The success rate is x%." It should say, "The success chance is X%," unless there is some variable in the coding that forces a server wide success rate. So I wonder then, even if one person succeeds only 2% of the time while another person succeeds 78% of the time, does a server wide variable exist that forces a certain server wide success rate?
I also checked Enchant Scrolls, I did not see anywhere on them the dictation of chance nor rate. The info I'm thinking of, in regards to enchants, is from the "Useful Information" when you log in. I will not log in several times just to figure out exactly what it says, but it does say something to the effect of "When enchanting over +3 there is a 1 in 3 chance of failing."
I stand corrected, given my new understanding of what's being said in this thread, I would definitely want NC to figure out if the success rate in question is correct. However, if the wording is supposed to be "chance" instead of "rate" then it needs to be changed.
There's no need to fix as success rate ~=~ success chance. If the success chance is 66% so the rate will always be smth withing 63-69% taking large samples.
However if success rate is ~40% as with magic weapon case conclusion is simple: success chance is ~40% as well. 10% would be significant difference to consider an error however here we have 28%... this is just too astronomical and NEVER happens in reality.
I'm not gonna quote statistics- but from my and various other personal experiences with crafting, it seems VERY screwed up. The 60% is really "off". See post in Player Questions about same subject for more info on that...
Back when I attempted to craft Drac LA- I crafted 1/7 successful crafts. So, 6/7 crafts FAILED.
In attempting to craft Dynasty LA- I crafted 3/10 successful crafts. So, 7/10 crafts FAILED.
This is not exclusive to me, I know others that have failed 9+ dynasty crafts.
The crafting system is broken. End of story- everyone knows it and a change is long since due. I think at least a portion of the base mats used should be returned. I understand the recipe and key mats going -poof- but, it would be a lot more reasonable if a least some base mats were returned. Losing 400+ mith alloy/c. leather/DMP/Enria/Synth. Cokes/VoP and other high-level mats, per SINGLE craft is just absurd.
Throwing away that many base mats from the economy with every failure only indulges the bot/farmer style collection of mats (24/7 afk) and often forces people to purchase mats from the bots, therefore continuing the cycle.
I would like to suggest for you to read some Statistics 101 before commenting smth like this in feature.
There's no need to fix as success rate ~=~ success chance. If the success chance is 66% so the rate will always be smth withing 63-69% taking large samples.
However if success rate is ~40% as with magic weapon case conclusion is simple: success chance is ~40% as well. 10% would be significant difference to consider an error however here we have 28%... this is just too astronomical and NEVER happens in reality.
Finally someone who knows something about statistics. I was beginning to despair.
You do need a rather large sample though. I would go with something like 100+ to get a low margin of error.
Finally someone who knows something about statistics. I was beginning to despair.
You do need a rather large sample though. I would go with something like 100+ to get a low margin of error.
Which, as you will find out, is exactly what they did in the Korean tests mentioned in the first replies to the OP, as I tried to explain in reply to all those "success rate vs. success chance" posts.
I've been wondering if there % also apply on elemental crystals, could it be possible somehow getting a elemental crystal on a weapon be more harder on a nuker weapon then on a melee one?
I've yesterday tried 19 wind-crystals only 5 succeeded. I know this is a very small sample, and could just be bad luck, but still, if other classes would have it easier getting a high element on the weapon I wouldnt mind if this was atleast communicated to the player (and fixed).
Another related question, is there any information on what the succes rate of the crystals should be?