Does anyone knows , preferably from a valid official source , if the chances to enchant from +15 to +16 is 66% as the previous ones ( and is this true anymore as well ) ?
there was some theorizing (backed with some statistical data from a korean pts) that enchant rates dropped after +9 and again after +15 ... but "official sources" don't exist, NCSoft doesn't like divulging details like that so ...
there was some theorizing (backed with some statistical data from a korean pts) that enchant rates dropped after +9 and again after +15 ... but "official sources" don't exist, NCSoft doesn't like divulging details like that so ...
Yeah thats what I am looking for . Thought someone could had something in handy from korean powerwikian or the russion official version . . .
Does anyone knows , preferably from a valid official source , if the chances to enchant from +15 to +16 is 66% as the previous ones ( and is this true anymore as well ) ?
In my lifetime I have gotten a decent amount (20+) of weapons to +15 ranging from C to S grade, but none have ever gotten to +16.
This doesn't necessarily mean the success rate is lower, but you get the idea. Good luck!
In my lifetime I have gotten a decent amount (20+) of weapons to +15 ranging from C to S grade, but none have ever gotten to +16.
This doesn't necessarily mean the success rate is lower, but you get the idea. Good luck!
Thanks ! Just for informative purposes and to make some remarks on the binomial distribution the pseudo-random generator this game uses, did you tried +15 to +16 right after the +14to+15 enchant or after some days have passed ?
As for my case , I still need that weapon so I ll hold on to it for as long as I need it or for when I get a EWS of destruction ...
4 to 15 is about 60% or 66% and 15 to 16 is like 10% or 30%, i am not sure
Next time when u gonna use ur crystall ball, check the weather for next week and if i become rich kk?
From 15 to 16 is same success rate than from 3 to 4. Ppl who got math on decent lvl in school should knew that if u got a chance to get something with 66% chance then get it 13 times in a row will be about 1% chance, thats why mass enchants fail oftenly before 16++. But theoretically if u buy +15 weapon, u got 66% chance of success ri get it +1. Anyway 16 is just a magic number that makes u'r weapon shiny red and means u got expensive **** in ur hand. Its just a bit better from +14 weapon, and little more better from +12 weapon etc...
Once i had huge load of D enchants from event. I enchanted few weapons to +15 (from over hundreds), and i enchanted few of them to +16. From 4 weapons 2 successfully enchanted to +16 from +15. I got about 100m profit by selling ssd/cry d overall :)
P.S. : Each private server got his own enchant rates on each weapon stage(Like many other things tbh). For example mine own i played long time ago got much different then official one so dont be suggested by some private servers, mostly pvp ones.
I meant each step is 66%, not 4 to 15 is 66%, sorry if that was not clear.
as about 15 to 16 thing, well you may believe me or not, but this is a known fact. people tested that. do try yourself if you feel lucky, not that hard to throw 3 out of 10 but if you would ever have 100 +15 items and would try to oe them all to +16 you'd end up with less than 1/3 of the original stock and lots of crystals.
Next time when u gonna use ur crystall ball, check the weather for next week and if i become rich kk?
From 15 to 16 is same success rate than from 3 to 4. Ppl who got math on decent lvl in school should knew that if u got a chance to get something with 66% chance then get it 13 times in a row will be about 1% chance, thats why mass enchants fail oftenly before 16++. But theoretically if u buy +15 weapon, u got 66% chance of success ri get it +1. Anyway 16 is just a magic number that makes u'r weapon shiny red and means u got expensive **** in ur hand. Its just a bit better from +14 weapon, and little more better from +12 weapon etc...
For 13 consequent trials in a row ( each individual enchant try ), assuming noone else enchanted items on the server at the time but you , the chances of at least one failure,be it +3->+4 or +15->+16(assuming that comes with the same chances of enchanting) that coulld be anyone due to the combinatory logic nature of binomial distribution this game uses, would be 99.55% . So you can easily see why enchanting in a row for a more than 4 times is not a very good idea.(even for 4 trials your chances for all successes are down to 18.97%)
I am good with probability theory , I just wanted to know EXACTLY how the chance system of this game works so I can make more educated guesses and the like.
As long as each draw is statistically independent what difference does it make wether you enchant 16 times in a row or 16 times in the course of a few days? If you throw a fair 6-sided dice with 2 red and 4 green faces what does it matter for your probability to have at least one red throw in n trials if you pause while doing the test and someone else is throwing the dice in between?
I'm not a computer scientist, so if it has to do with the supposed implementation on the server that's another thing (which I have no clue about).
As long as each draw is statistically independent what difference does it make wether you enchant 16 times in a row or 16 times in the course of a few days? If you throw a fair 6-sided dice with 2 red and 4 green faces what does it matter for your probability to have at least one red throw in n trials if you pause while doing the test and someone else is throwing the dice in between?
I'm not a computer scientist, so if it has to do with the supposed implementation on the server that's another thing (which I have no clue about).
Because due to the nature of binomial distribution and the pseudo-random implementation this game has , getting 15 out of 15 trials ( ie each enchant ) as successes has a chance of 0.45% . But imagine your 15 trials are a part of a bigger total of lets say 1000 trials (in this example the amount of enchant trials over time since you started your own enchants ) , planting your own 15 trials throughout will have better results due to the nature of the combinatory logic the binomial distribution uses for determining results.
tl;dr it doesnt have to do with the specific algorithmic implementation that is used here ( I would love to see the programming code of that to draw some conclusions!) but with the general implementation of the binomial distribution .
Hope I made that clear , scientific explanations in English is not my strong point.
Because due to the nature of binomial distribution and the pseudo-random implementation this game has , getting 15 out of 15 trials ( ie each enchant ) as successes has a chance of 0.45% . But imagine your 15 trials are a part of a bigger total of lets say 1000 trials (in this example the amount of enchant trials over time since you started your own enchants ) , planting your own 15 trials throughout will have better results due to the nature of the combinatory logic the binomial distribution uses for determining results.
tl;dr it doesnt have to do with the specific algorithmic implementation that is used here ( I would love to see the programming code of that to draw some conclusions!) but with the general implementation of the binomial distribution .
Hope I made that clear , scientific explanations in English is not my strong point.
So it's only a problem of the implementation using pseudo-random variables, right? Thus violating the assumption that all tries are statistically independant?
Cause in a mathematically ideal system (independent tries) I still wouldn't see what influence the total amount of tries would have.
So to sum up my question: In an ideal case (mathemcatically) it would not matter whether you enchant 15 times in a row or whether 201,923,141 people enchant stuff in between, right?
On topic: I don't know if you've seen that, but himred collected data on OEs on l2wh. He recorded ~2900 OE attempts and listed all of them on the site, just look at tools->OE tools->OE success rates. Unfortunately only 3 tries are recorded where someone went from 15->16.
So to sum up my question: In an ideal case (mathemcatically) it would not matter whether you enchant 15 times in a row or whether 201,923,141 people enchant stuff in between, right?
On topic: I don't know if you've seen that, but himred collected data on OEs on l2wh. He recorded ~2900 OE attempts and listed all of them on the site, just look at tools->OE tools->OE success rates. Unfortunately only 3 tries are recorded where someone went from 15->16.
No , because in binomial distribution , there is no lack of memory propert as in other distributions . So while each of the independent trials has a 66% chance to be a success in that ideal world, being independent bernulli variables , all together follow the binomial distribution ( which is just the groupping of all the independent trials) .
Using binomial ,
for 12 out of 12 trials
getting all successes is only going to happen(theoretically since implementation derives due to lack of precision)
only at a 0.6832 % rate
Now , if you expand the number of total trials to 1000 lets say ,
you can see that the number of succeeded in that total will be a lot bigger than the first total of just 12(obvious) , making the chances of your trials being the successful ones better as a part of the bigger total .
So basically what I am trying to say is that in the first case ,
all 12 trials are for your enchants , making even 1 trial failed to make all your experiment/enchanting fail .
In the second case , you share successes and failures of enchantment with others as well , so that 1 failure could belong to a trial by someone else which wont affect your results .
I hope I gave you the basic idea of what I am saying.
tl;dr
Each trial is independent but the success of it only calculated as part of a binomial distribution , that calculates results taking into account previous trials as well.
Add in the lack of true random numbers generator and you got some faulty implementation of that , on top of everything.
Mixing numbers from various sources enchanting success rates from +4 -> +15 is about 70% for fighter weapons and 40% for mage weapons.
For +15 -> +16 there is only one experiment from Korean Labs showing a fixed 1/3 chance for both classes of weapons.
That's pre-GoD numbers for a range of C-Grade -> S84-Grade Weapons.
As correctly stated by some ppl b4 are some factors that may skew the numbers above (type of distribution, lack of "real" random number generator, etc, etc).
For the sake of convince and from a practical standpoint of view you could rely on the 70% / 40% success rates
On topic: I don't know if you've seen that, but himred collected data on OEs on l2wh. He recorded ~2900 OE attempts and listed all of them on the site, just look at tools->OE tools->OE success rates. Unfortunately only 3 tries are recorded where someone went from 15->16.
That's an old story, I have made a post on that while aggregating numbers. There is an ERROR on Himred numbers.
L2wh drops all weapons in the same basket when doing math. IF you split weapons to mage / fighter (done that) you will end up to a completely different conclusion. By coincidence, l2wh mixing of weapons did end up to an "expected" rate about 66$ that's why nobody argued for so long.
The analogy of fighter weapons with ~ 70% and mage weapons with ~40% provided an "expected" 2/3 flat rate.
There is also another problem with l2wh data depending on how you calculate the overall success rate. Many mage weapons were overenchanted as fighter weapons and converted back to mage from mamon.