Alright .. I was keeping track... It took about 10 hours or so to do all these but here we go. No real bias just the numbers. We (my wife and I) used apprx 6500-7500 scrolls, these numbers are based on 5,350 scrolls and 811 total Yogi Staffs. It does appear that you have about a 70% chance of success on everything up to +15, over +15 however it seems to be about 25% chance of success. Again .. we did not turn in any of these staffs, each one blew up trying to get to +19. Here are the stats:
As you can see after +15 there is a significant drop in success rate. The approximate constatant success rate +4 thru +15 would indicate that the success rate over +15 is not an anomaly and a real change in success rate. This would indicate that NCSoft has been a bit misleading in relaying this type of information.
Had we know this before this event we would certainly have not tried for the +19 as it is statistically extremely low and certainly more profitable to sell the S Jewelry and buy a S weapon with the proceeds than to attempt further than the +15.
Anyway... I thought that our time would be helpful to others. Of course we have never tried enchanting real weapons over +15 so are not sure if this is event related or a real drop in all weapon enchanting, but if you do get a +15 real weapon ... consider yourself warned.
I would tend to agree with these findings. The max I have gotten to was +15 on three staffs and they all blew up. Could be luck, but lower success rate would explain it better.
RumPunch ... your post was based on the theory that all enchants are a 2/3 success rate. I started with this theory as well and then kept track with a study to try to prove this theory right/wrong. What I have found is that the theory itself is flawed in that over +15 the % success changes. If you have documentation (ie. enchant study info) that would point to a different success rate in any of the categories I would be interested to see it as it could help us all understand better what we get ourselves into.
Technically a statistical sample size is what you start with ... not what you end up with. The 811 staffs that I started with is my sample size, it does not bear any weight how many were attempted going from +15 to +16, but rather the statistical % success rate.
As I stated in my first post, it is reasonable to assume (because of the consistent average % success rate up to +15) that over +15 the % success rate is significantly lower as is indicated by the results.
Does this mean it is absolutely correct ... no. What is says is that per this sample these are the results, I would have loved to see someone else put in the time and effort we did to come up with different/other results. We kept track of each individual staff. Without a different sample size, of even size or larger, I have no reason or basis to believe that the success rate would change by any substantial amount.
I bought 5k enchants (125kk) and at least 1k from my 2 accounts. My plan was to go +19 for one S grade weapon. I dont have stats up to +14 but i record what happened from +15 and on.
A lot of people think this event was part test for the 10% oe purchase bonus available in korea. I think most people would agree that the +4 from this event was much easier than a standard +4 and this research definately reinforces the extra 10% bonus.
Also, 2/3rds enchant rate is not a theory. It is blatantly stated in one of the lineage 2 tips shown in the lineage 2 splash/load screen, get your info straight.
haha! You're the ignorant one bro, you are going off information that has never been provided. Using rumors as your factual evidence is not suggested.
well "ignorant" is someone who ignores information right? when lots of people suggest one fact and you don't believe them, and once you try it yourself you see you where wrong .. does that not make you ignorant?
also how can you call me ignorant, do you really know I didn't have statistics myself? well I did! I have done lots of OE on the PTS servers, I've seen lots of them and on some there were enough free adena to gather some statistics.
sure if you don't want to gather statistics yourself, you can believe all you want to the only official source which is known long time for its tendency to give wrong and incomplete information .. oh well if you choose to believe that source no matter what, that speaks volumes about you.
Olgen, if you look at Geraki's results, he had a 45.45% success rate going from +15->+16 with a sample size of 11 while you had a 26.67% success rate going from +15->+16 with a sample size of 15. The fact that geraki's rate is 1.7x yours shows how inconclusive it is to do a statistical test with such a small sample, and to come to any conclusions is nothing other than silly. Hell, geraki had a 100% success rate going +17->+18 with a sample size of 2, guess that means you're guaranteed to get +18 if you have a +17 staff by your logic regarding small samples.
but yes NCsoft claims not backed with any statistics is heavier than lots of people claiming otherwise with some statistics. keep living in your imaginary world.
but yes NCsoft claims not backed with any statistics is heavier than lots of people claiming otherwise with some statistics. keep living in your imaginary world.
lmao, by far I would rather live in a world where I can say "I don't know, all conclusions are inaccurate at best" than one where you only believe inaccurate conclusions with a high chance for a large degree of variation (thus making it useless). I'm not saying ncsoft is right (even though I'd more readily use the 2/3 probability for any oe), i'm saying your conclusions are premature and inaccurate at best.
Edit: Lets see these numbers from "lots of people claiming otherwise with some statistics", I'd be more than happy to be proven wrong if some people would post their results after +15 with a significant sample size.
lmao, by far I would rather live in a world where I can say "I don't know, all conclusions are inaccurate at best" than one where you only believe inaccurate conclusions with a high chance for a large degree of variation (thus making it useless). I'm not saying ncsoft is right (even though I'd more readily use the 2/3 probability for any oe), i'm saying your conclusions are premature and inaccurate at best.
Edit: Lets see these numbers from "lots of people claiming otherwise with some statistics", I'd be more than happy to be proven wrong if some people would post their results after +15 with a significant sample size.
not many people care to gather statistics, of them not many read this forum and of those few, not many will bother to post it here.
and btw enchant chances are not only from statistics, they are also taken from certain code disassembly but i wont be too specific here
bit confusing as to how that's laid out, but from how it looks that's with 6 staves being enchanted to +16 with 2 succeeding? If it's the 16 staves that's still an absurdly small number of attempts. If we could pool statistics to a sample size of *at least* 100 attempts going +15->+16 it'd be better, but that'll still be fairly inaccurate.
Just by observing the results ppl posted here, it seems like there's a rough consistant of +15; with some luckier than others.
Barbados - 15 staffs (5350 scrolls)
Geraki - 11 staffs (5000 scrolls)
olgen - 12 staffs (5760 scrolls)
I think to be more accurate, someone needs to do like 20k or more scrolls from +0-23 XD
bit confusing as to how that's laid out, but from how it looks that's with 6 staves being enchanted to +16 with 2 succeeding? If it's the 16 staves that's still an absurdly small number of attempts. If we could pool statistics to a sample size of *at least* 100 attempts going +15->+16 it'd be better, but that'll still be fairly inaccurate.
check that post again, I have edited it to provide fails/total numbers instead of only total.
I agree the sample is small, but there are no other kind of data yet. to beat existing sample, you have to find really lucky player, who gets like no fails for about 12 attempts ... but all we have now are 1/8 reports and stats like those above.
you have a sample size of over 40 enchants for 15-16 which does give a rough idea of what the enchant rate is like (about 25% off the top of my head)
either way NC should have just made the top prize a controlling share in microsoft for the odds they gave us but i do always enjoy their epic events with nice rewards like D and C enchants, crap roleplay hats and lowbie gear that would have been useful to me a year ago
I just noticed something with the original numbers after a few checks. Assuming a 2/3 probability per overenchant (where one overenchant has a 2/3 chance, two overenchants in a row have a 4/9 chance (2/3rds of 2/3rds so 2/3*2/3 giving 4/9 etc.) and using your numbers:
however, +4 -> +15 aren't in question, so I'll jump ahead to +15:
+15 - (2/3)^12 * 811 = 6.25065, you got 15 to +15, 9 more than the average for 811 staves
+16 - (2/3)^13 * 811 = 4.1671, you got 4 to +16, basically the average
+17 - (2/3)^14 * 811 = 2.77807, you only got 1 to +17, a little under the average
Overall, with barbados' 811 staves, they almost perfectly follow what's to be expected with that many trials all the way to +16, he just happened to be a little lucky getting up to +15 (as shown by the %'s written next to each oe) until he starting going for +16, where it averaged out, then a lower success rate going for +17 (with 4 staves, a small sample).
Edit: I decided to check overall probabilities to each oe as that allows us to use a much larger sample (such as those 811 staves) as we use the starting number and compare how many you got to each oe to the overall chance of each oe. This is much better compared to checking one oe level (in this case getting +16) as that drastically cuts down on our sample, as shown, staves in the hundreds are quickly cut down to a handful by the time you get high enough.
edited a second time as i somehow thought olgen was the OP, now fixed.
Just by observing the results ppl posted here, it seems like there's a rough consistant of +15; with some luckier than others.
Barbados - 15 staffs (5350 scrolls)
Geraki - 11 staffs (5000 scrolls)
olgen - 12 staffs (5760 scrolls)
I think to be more accurate, someone needs to do like 20k or more scrolls from +0-23 XD
no, 12 is the number of +15 shafts. I have not tried all of them to +16, only 6 of which 2 succeeded.
no, 12 is the number of +15 shafts. I have not tried all of them to +16, only 6 of which 2 succeeded.
So it's yes. What I posted was the numbers of +15 and yes I noticed you have turned in some +15, so the number is no longer accurate to be compared with the other 2 datas.
What I find amusing is that everyone is saying that the +15-->+16 doesnt have enough sample size, however in that case .. neither does the +14-->+15 but noone is arueing that that one is incorrect.
One thing many people do not understand is that when you have as consistent of numbers from one point to another and then using the same methodology the immediatly following point is consistently inconsistent then it is not a matter of statistical anomoly but of statistical indifference.
I appreciate others adding the information regarding their enchant results I think everything here would prove very well that the enchant rate over +15 is between 25%-33% on average. That is awesome, I am glad you were able to help me out with additional info!
BTW, the 2/3 up to +15 counting quantities based on probabilities you cannot judge the quantities as being too high because the previous enchant probabilites were not 2/3 but higher (and in this case, rather considerably higher being at 3-4%) therefore raising each and every following statistic. I can see your arguement and the basis you used is sound but the process is incorrect.
I used 1443 scrolls, 244 staves, got to +16 3 times.
The first one succeded to +17 and I traded it for a EWS (stopped right in time, because my next staff blew up after +3), the next one blew up going +17 and the last one blew up going +18.
An interesting detail: I made the first 2 +16 staves out of my first 200-300 scrolls, then a thousand more scrolls brought nothing but failures going above +8 or so (behold the Korean Random at its best!), and in the very end of the session I got another +17 (which then failed). Also, during the end of the session I got 3 or 4 staves to +10 (out of 6 in total).
I did not record how many staves broke at +3, +4, +5; the rest of results are:
+6 = 24 (i.e. broke going +7)
+7 = 17
+8 = 12
+9 = 7
+10 = 6
+11 = 2
+12 = 0
+13=0
+14=3
+15=0
+16=1
+17=2 (1 turned in, 1 blown up).
What I find amusing is that everyone is saying that the +15-->+16 doesnt have enough sample size, however in that case .. neither does the +14-->+15 but noone is arueing that that one is incorrect.
However, the probability of success going from +14->+15 is not in question as brought up by yourself, it's the probability of going from +15->+16
I appreciate others adding the information regarding their enchant results I think everything here would prove very well that the enchant rate over +15 is between 25%-33% on average. That is awesome, I am glad you were able to help me out with additional info!
BTW, the 2/3 up to +15 counting quantities based on probabilities you cannot judge the quantities as being too high because the previous enchant probabilites were not 2/3 but higher (and in this case, rather considerably higher being at 3-4%) therefore raising each and every following statistic. I can see your arguement and the basis you used is sound but the process is incorrect.
having a variation in the actual probability is to be expected, olgen's result going from +10->+11 with 56/74 succeeding (75.676% success rate) has a specific chance of 2.58%, given every other possible result for that round that had a significant chance of occuring with just above a 1/25 chance. (for the record that formula is (nCr)x((2/3)^r)x((1/3)^(n-r)) with n being the total number of staves, r being the successes, it gives the chance of only that specific combination) the chance will also be considerably higher if you include the chance of getting 56 to 74 successes out of 74 attempts. Actually, if you had 60 staves, the chance of having 41 successes (1.66% higher success rate than 2/3rds) is 10.607%, and the chance of 39 successes is 10.3547%, so you actually have a slightly higher chance of the deviation resulting in more successes.
Regardless, with a 2/3 chance (which is what you're claiming +15->+16 is deviating from), (2/3)^13 (13 being the number of overenchants to get to +16) you get .005139231, or just a little over 1/200 chance, you attempted 811 staves and 4 of which got to +16, sounds correct to me.
So far with the results of getting +16 corresponding with the expected probabilities, the only way to prove a change in probability is with several hundred staves going from +15->+16, not 40 or so.