Masterwork / Enchanting exceptional Items - - - -

Ok is appears that items that are Exceptional (+50) durability seem to have an affect on the RNG when Master working or Enchanting they seem to be break much more. As you know I have made ALOT of items in testing and I cannot put actual numbers on this (I will try) but it appears that items that where made exceptionally break on the first try more often than item not exceptional. All I can figure it that somehow the higher durability is affecting the math??

I am sorry but since I cannot see the code I cannot dig to deep to actually see what is going on but it appears pretty consistently.

Loc: (428.5, 77.0, -320.6) Area: POT_island_metropolis_01_templateSotA.EarlyAccess.Win.64.442.Date.07.11.16GPU: NVIDIA GeForce GTX 980M GPU RAM: 8093CPU: Intel(R) Core(TM) i7-6820HK CPU @ 2.70GHz (8) System RAM: 32636OS: Windows 10 (10.0.0) 64bitUser Specs:Steps to Reproduce: Details: Blocker? Reproduction Rate: Title: 7/13/2016 11:58 PM

 

I have noticed that, also. The exceptional items seem to have difficulty making it past the first enhancement. When I have 90% chance on the first attempt there will be breaks but most of those first attempt breaks happen to be on the exceptional items. I thought it was just me being extremely unlucky.

 

It still might be you being extremely unlucky.

Consider the following:

1. Exceptional Items are far more likely to be enchanted than non-exceptional items. Do you have any Exceptional Items that you're not enchanting? Probably not, which really skews the numbers in a way that could leads us to falsely believe that exceptional items are more likely to break than non-exceptional items.
2. Statistically we don't have a sample size large enough to tell us if the randomness of the system is working correctly. So as unlucky as we might be, there's no way for us to know if it's broken or not. I don't know the exact sample size we'd need, but I'm guessing 100 would be a good start. Someone that did better in their statistics class than I did would have to explain all the nuance involved in determining what the sample size would be, and how many should be exceptional or non-exceptional before we have a clear understanding of what the actual probability of each one breaking is, and it the randomness is statistically random or not.

 

I just spent about 2-3 weeks crafting my butt off, and I have noticed this as well. I can get about 1 in 5 to break getting double masterworking at 95% then 82%. I can get about 1 in 3 exceptionals to break getting double masterworking. This COULD be a case of luck and conformation bias. I am not going to say I am 100% right.