Are the upgrade percentages for +7 and more real?

    The upgrade chances for +7 Kaia's Soul gear supposedly start off at 10% and increase by 0.6% at every failure, so getting an upgrade off might not seem very likely at first. However with some quick maffs, you can determine that on average, you should only get around 6 failures before having your gear upgrade.


    I have tried with my weapon and gloves and have failed respectively 13 times (and still counting) and 17 times. That means they had a 14% and 6% of chances to happen, and surprisingly, as I was asking around, most people I know seemed to have had to do that many tries. So either EVERYONE I know is very unlucky, or those displayed chances are rigged in some way.


    To get a better sample size, I'm turning to you. Have you guys been lucky with your upgrades? How many tries did it take you to get to your +7,+8,+9?

    Short suggestion : They shouldn't display numbers that have absolutelly nothing to do with upgrades since everything is RNG...

    If you wanna show numbers, show only one : 100%, everything below that is pointless ^^

    Quote from MasterMico

    Short answer: 10 % is 10% you can get it first try you can get it after 100 tires. RNG is RNG.

    Yes, but they don't have the same chance of happening. Getting it first try has 10% chance while getting it after 100 tries has a 0.002% chance. RNG is RNG, what I'm questioning is whether the 10% displayed is the real chance or not, because from my experience, everyone I know seems to be unlucky, which might not just be a coincidence.

    The average wants it to be 6 fails into success. I'm just wondering if that's what most people are getting on average....

    it works much better when you view it differently.

    a 10% chance to succeed is the same as a 90% chance to fail.

    (yes i am a glass half empty man, but you can never convince me 10% is half full.)


    the best advice has already been given: it is RNG. might get it in 1, might not get it in 100 tries. just luck.

    Rara avis in TERA

    Quote from Somnambula

    it works much better when you view it differently.

    a 10% chance to succeed is the same as a 90% chance to fail.

    (yes i am a glass half empty man, but you can never convince me 10% is half full.)


    the best advice has already been given: it is RNG. might get it in 1, might not get it in 100 tries. just luck.

    It's hard discussing this with people that haven't studied any probabilities...

    The chance to get it in 1 is the same chance as getting it on exactly the 100th try. But once you have it, it's done, so you shouldn't count the chance to get it on exactly the 100th try. For the 100th try example, you also have to take into account the fact you could have gotten it on the 1st, 2nd, 3rd... 98th,99th or 100th try.

    So actually, the chance to get it before your 100th try is 0.9^99 = 99.997%. IT IS ALMOST CERTAIN, IT'S NOT JUST RNG.

    In the case of getting +7, the 50% mark is at the 7th try. So you SHOULD get it around the 7th try, sometimes a little more, sometimes a little less. (You can test this with the formula "product of 0.9-0.006*k for k=0:5")

    This is actually very comparable with a dice roll (16% chance instead of 10% but it doesn't increase with fails so fairly equivalent). If you bet on one number and start throwing the dice, you will usually reach that number in 4 to 5 tries. Sometimes it will be less, sometimes more, but on average, over a large number of tests, you will get it around your 4th or 5th try. With reaching +7, it should be around the 6th or 7th try.


    It's like saying yeah, getting heads or tails on a coin toss is 50% so getting heads 100 times in a row is 50%. HELL NO, that's actually almost impossible statistically.

    Quote from victoryzeta

    In the case of getting +7, the 50% mark is at the 7th try. So you SHOULD get it around the 7th try, sometimes a little more, sometimes a little less.

    I believe this is not how statistics works, there's no "SHOULD" in statistics. Statistics don't describe what should or will happen to you as one specific person, because as you said,


    Quote from victoryzeta

    over a large number of tests

    You and your few friends are not a large number of tests.

    I believe psychology is at play here. People remember their failures more than victories. So they will always remember it when their enchant fails more than "the expected number of times" but they might have forgotten the number of times they have managed to enchant faster than expected (possibly on lower enchantment levels). So in conclusion, people will always feel more unlucky than they really are.


    If you want to actually prove that the enchant rate is lower than stated, you need to collect the exact number of tries from a large number of players (and make sure they report the number for all their gear, not only those pieces of gear that took too long to enchant.)


    All of this aside, I do believe 10 % is a frustratingly low chance so your anger at the devs is justified.

    Like I feel there's no need to prove that the chance is lower than stated because the chance is too low ALREADY.


    EDIT: Now I see that you actually asked people to share their enchant succes in the original post, before other people misdirected the discussion to argue about the meaning of probability. I'm sorry I don't actually have any data for you because I never bothered enchanting my gear (too much grind for too low upgrade chance).

    Edited 2 times, last by Sewin ().

    Quote from Sewin

    EDIT: Now I see that you actually asked people to share their enchant succes in the original post, before other people misdirected the discussion to argue about the meaning of probability. I'm sorry I don't actually have any data for you because I never bothered enchanting my gear (too much grind for too low upgrade chance).

    This was indeed the point of my post. I realize how small my sample size is, so I wanted to expand it. Still surprising every single example I know of has had more than 10 failures when in reality, 10 failures only have a 25% chance of happening. I'm personally at 15 and 17 failures with each having below a 10% chance of happening. It just seems very fishy so I wanted to collect more data. If it were only around 6 failures every time, the grind wouldn't be too bad and I might keep at it.


    Sadly, not only are people not answering the question, but they are actively showing they don't understand how probabilities work by crying "RNG" out loud as if a 10% chance were the same as a 50% and the same as a 0.002%.

    Quote from victoryzeta

    This was indeed the point of my post. I realize how small my sample size is, so I wanted to expand it. Still surprising every single example I know of has had more than 10 failures when in reality, 10 failures only have a 25% chance of happening. I'm personally at 15 and 17 failures with each having below a 10% chance of happening.

    Mathematician here: If you really want to validate your claim that the failure rate displayed is too low you will need a much larger testing size and should check for p value (statistical significance) of the question is the real upgrading probability 10%. It is nothing unusual in real world probabilitys to have an event happening with 10% chance many times in a row for example rolling a 10 sided dice and getting 4x times 10 in a row has a probability of 1/10^4= 0.0001 and yet someone will get it in 1 try on this planet every day.

    If p< 0.01 you should reject the hypothesis above with 99% significance which should be enough in this case D:

    Sadly I dont have time for this so if you want to give it a try and calculate it with an algorithm have fun

    And also empirical data always has errors dont believe everybody who writes: Oh I got 100 fails and run a simulation to proof the validity of you claim.

    Quote from Sewin

    I believe this is not how statistics works, there's no "SHOULD" in statistics.

    Probability, not statistics. The probability of getting +7 after 7 tries is around 50% (I think it's actually the 6th try because of the 0.6% increase on each fail) so on average 1 out of every 2 players SHOULD get +7 within the first 7 (or 6) tries. But that also means 1 out of every 2 players will take more tries.


    But I agree, I think psychology is at work here. People see 10% and think that means they'll be guaranteed to succeed in 10 tries and expect to get it sooner, while in reality it's a lot more complicated.

    ^^ all very interesting.(no sarcasm)

    unfortunately in any calculation there will be missing facts of the kind that businesses will not publish for various reasons.

    for example they might have a "hidden" factor that increases chance after having spent an x amount of cash so as to have a balance between "we will take as much money as we can from you" and the player's toleration upon accepting failures untill getting the desired result versus the part of ragequitting and never spending a dime on the game anymore.

    this leaves us with incomplete calculations because it lacks certain values/factors.

    especially when programmed, an RNG's randomness can be manipulated in many ways.


    so in conclusion it still seems to me the best way to approach this is to get what you can, enhance chances by shopping and/or farming, and, mainly: dont worry about it.

    Rara avis in TERA

    Thats the thing behind my post. If you really want to check if the true underlying percentage is 10 % then you can do it with enough data + checking p values in hypothesis testing and arguing about possible errors like measurement error or omitted variable bias, which should be the case in this situation.

    Just ask equinox :^) i think he is the living proof that you can be lucky there aswell. Think he got like everything 1st or 2nd try form 7-8 / 8-9

    I tried so hard, and got so far... but in the end, it doesn't even matter...

    as someone who spent 18 months suffering being undergeared and handicapping every shorehold team i was in going from +6 to +9 pvp weapon with odds many times higher than yours (only counting 8-9 upgrades, mine passed on attempt 168) ... i can tell you its pure rng, the gearing system is shit and wont change cause its a great p2w system and it works for drawing easy money to the devs