Please click here to read the announcement.
2 Likes
We would also like to add that enhancement attempts performed one hit at a time or three hits at once are a mere visual difference and do not affect the actual chances of success.
Can you add some detail to this? There has always been a debate whether success is calculated for each hit or per anvil (that’s where the the one hit / three hit myth comes from).
This is the calculation for chance of success:
This is the determiner if it succeeds or not:
This is chance of what hit the anvil will show being failed on:
As soon as the anvil is placed, success or failure is determined. The hits on the anvil when it breaks are random and don’t actually mean anything. Each step is done separately from the rest.
20 Likes
Thanks, that clarifies it once and for all.
1 Like
Mindless whiteknights proven wrong yet again.
The most important part of the Q&A is of course this:
Some players have suggested that the error can be reproduced through specific addons, but we have yet to find any connection between them.
1 Like
IMC can tell you with a straight face that success rate will always be greater than failure rate. 
3 Likes
I can feel the same thing too, hitting on anvil, just an animation.
Can you please explain this to me lol
I wasn’t expecting such a thoroughly detailed announcement about this issue and I’m surprised in a positive way. However, the hot debate we had yesterday could’ve been avoided if the staff members said they were preparing this announcement.
I can’t say I’m at ease yet, specially after reading what @Queue posted. Some people were exploiting that “guess the number” event by inputting a command to show what was the correct number of the day and that makes me think the way enhancement works may be easier to exploit than one might think. What if there’s an input command or another way to check the result of your enhancement attempts before hitting the anvil too? You could even reach high enhancement levels with just regular anvils by putting them, checking if the enhancement will be successful or not and get away from the anvil in case it’s a failed attempt for it to disappear, then put another anvil, check if it’s a successful attempt, rinse and repeat until you get only successful results, even if you need to retry a million times with regular anvils.
Sometimes I think the best way to handle this would be giving the enhancement function to the Blacksmith NPC instead of placing the anvil on the ground for you to hit it, i.e., talk to the NPC, a window will appear for you to select the item you want to enhance and the anvil you want to use for it. That would make people not feel doubtful about the enhancement results being manipulated or not by the player.
7 Likes
what part of it exactly did you not understand? all of it?
@Pokart That’s actually very possible considering what queue posted and your solution could fix the issue indeed. Probably why most games use npcs for that huh.
After +5, your success rate will decrease the higher your enhancement level.
Armors have a slightly higher success rate compared to weapons.
Your success rate for both weapons and armors will never be less than 51.2%
1 Like
This might be what “that addon” does. 
After all these clarifications, I can now safely say that I have absolute sh!t luck in this game while others are the opposite. Guess I’ll just buy TP and roll gach—oh! 
This assumes the “success/failure” cases are sent to the client which they most likely aren’t. And yes, some of the lua code never runs on client-side.
So I think the basic pre-requirement for this case to exist might not be met.
And given that they could track how many times the anvil tried to destroy an item, that means that the old assumption that as soon as you destroy the anvil it gets saved on the DB is true, unless something else happens there.
Assuming this, unless you can find the seed and how many numbers were generated prior to your enhancement attempt or read the memory of the server processes it’s highly unlikely that you can do anything of the sort.
Also, this is the ktos announcement on the same topic:
http://tos.nexon.com/news/tosnotice/view.aspx?n4ArticleSN=1110
It’s actually funny how the highest number of fails are attempts to +8 ~ +10. It reflects what I see in-game when trying to enhance things. lol
1 Like
wait am I reading that table right? Is it showing number of tries vs success/failures? Meaning 3 tries at 35 , then 2 at 36 then 3 at 37? Did someone there manage to go from 35 to 37 skipping 36?
edit: oh never mind . Forgot diamond anvils were a thing for a second there. Guy probably used diamond at 37 , failed and tried again.
do you have this IMCRandom logic? Like, is it true random or there is some logic shenanigans behind this?
it has to be otherwise they couldn’t claim the minimum chance to be 51.2% , meaning any number has the same chances of coming out of that IMCRandom.
I guess it depends on how the logic behind it works.
If IMCRandom sets a variable to define a “dificulty” level, that could use the actual enhance level number (as the successRate use, even though this will not be set lower than 51.2), then increasing the the value of IMCRandom based on that dificulty level, making (IMCRandom > successRate) to more commonly return true based on this internal “dificulty” level it created.
It would be easy to do, if you think that any enhance level >= 8 will end up always using 51.2 (for weapons and shields), so this number will be the only successRate compared most of the time, as +8 is quite easy to achieve, and greater values than this are always target for discussions on how the enhance mechanics work, or rather how fair it is.
EDIT: but IMCRandom have attributes (1, 100), so I guess it really pick a number in between it. But who knows if it is really random
I’m willing to bet all IMCRandom(int a , int b ) does is get a random number between a and b ( with a and b included) .
They probably did it like this because some languanges don’t have an easy to read random function ready to use, so they just made their own. It probably just calls the Math lib RANDOM and asks it to pick a number. No reason to get worked up by it.
But is funny to see that every attempt higher than +8 (weapons and shield-type subweapons) will always be, virtually, 50/50.
It doesn’t feel like this, specially if you look at the charts here:
and here:
Only if anvils play a role in the process other than Potential consumption that we don’t know.
I understand that a 50/50 should seem like upgrade should be easier, but think of it like this, how hard it is to throw a coin 16 times and get the same face every time? heck say you can get a different face 5 times but you have to rethrow it again. Do the test in your home and every time you do a throw add the cost , if you fail 5 throws start again and add the cost. You will notice you will get similar results.
1 Like