Crafting System IS Flawed

I am going to be short and sweet about this.

GM+ tailoring = 95% chance success rate on most items.
Craft 4 items in a row with 95% chance success rate = 4 failures in a row. Probability = .00000625% chance
Craft 4 items in a row on a different day = .00000625% chance.

How can any developer sit here and tell me to my face this is happening so frequently with so many players and that your system is 100% flawless?

@Chris

We all know you guys do a damn good job at Portalarium and I never argue that, however, the crafting system is one of the most vital aspects of the game and should be one of the primary focuses of getting 100% right.
I'm not saying your system is broken but it's definitely not flawless and the odds of this continuing to happen are as great as you or I winning a 500 billion dollar powerball ticket. In this case every crafting in-game would be a winner of the powerball because it's not just 1 or 2 people this is happening to, it's happening to nearly all crafters who are GM+ and have crafted enough items to see this.

My thoughts are your RNG system works great, but I'm not 100% convinced your random number doesn't reset after each succeeded crafted item or failed item. Rather to me it seems the number doesn't change but every so often(10-15 seconds for example) or once a condition is met. Whether or not that is leaving and resetting the crafting table(not likely) OR whether your RNG works like an authentication device and resets that random number every so many seconds.
Crafting should and needs polished and these 95% chance to succeed with 4 fails in a row need to stop. 80 beetles is not easy to come back, especially when we are consistently blowing them up at the crafting station on these .00000625% rates...

I would also go through make sure the RNG for crafting an exceptional item and the RNG for the regular item aren't somehow mingled together throwing the %'s way off because that could be another screw up causing your overall success chance much lower then the intended value.

In other words if you want to keep a good player base in this game, on a bad day of crafting meaning the system gives the worst numbers possible in the RNG system, you causing players 3 days of mining/gathering beetles to create 1 set of decent armor that gives any durability whatsoever, for which 3 pieces blow up will not do it. I would suggest you guys start investigating the flaw.

Thank you in advance,
Cairo Hayden

 

@Cairo Hayden - Thank you for reporting, but even though this post is more "feedback" than a bug report, I can tell you that this is a known issue and there are plans in the works to correct this.

I'm not certain, but I believe currently each and every attempt at crafting or harvesting something is calculated individually, and not as a string of attempts. The "fix" will likely involve an adjusting formula of sorts that takes into consideration strings or sessions of doing the same action instead of taking every case as a separate event. Sorry I don't have more details, but that is what I heard is in the works.

 

I'd like to put in my $0.02 here. I do believe its a bug, but without the ability to hunt for it I'm going to have to just use programmer's intuition:

Test case conditions: Grandmastered all Enchants / Masterworks. +5 Crafting Tables. Tools of Prosperity.

Probabilities math says to determine my chance of success to do a 2 enchant/2 masterwork, all I need to do is multiply all the chances together. So 95% * 82% * 95% * 83% = ~60%. However, I get maybe 1 in 10 items reach this point.

I believe we're seeing incorrect stacking, that is, the success resolution code is somehow counting the 2nd enchant and/or masterwork as the 3rd or 4th, to reach the 3x% percent chance. After that, there is a 14% or so percent for the fourth enchant/masterwork, which is almost exactly the same as the rates I see.

To state this another way, the math really is something like this (estimated): 95% * 82% * 34% * 14% = ~3%. Which explains why it took me 20 attempts to create a 2/2 Halberd.

I've been talking with other extremely grandmastered players and they are reporting similar rates, that they are burning through over 100 items to produce just 1 3/3 item "if they are lucky". This would jive with incorrect stacking as well, as two additional 14% chances would get down to a number closer to the 1 in 100 value which I'm hearing about.

While most have done what @Cairo Hayden have and blame the random number generator, @Chris posted code earlier showing how interruptions work and it included a snipped of code showing the use of the standard random() library function, so I know gazing in a sea of random isn't going to help. However, this smacks to me of code that isn't properly accounting for the fact that Enchantment and Masterwork are considered different.

I'd suggest Portalarium look at the code that is used to produce the success chance on the SotA client and compare it to the code on the server and see if the two produce the same values, because I have a strong feeling that the code that determines the outcome is calculating the percentage chance differently.

Portalarium, please keep in mind that resources are ratcheted extremely tight, so if you want my to produce enough information to prove this statistical occurrence, it's going to take me days of solid work to get enough to prove it beyond a RNG level of reasonable doubt without any expectation of return so please refer to your ELK deployment and run a report of broken/success events from players that are known to be producers for resell and see what their rates of success are. More to the point, I'd review the code and give it a solid rethink because "We're all just unlucky" isn't jiving with me.

Thank you for looking into this matter.

Engineer

 

First off, I did not mean to portray this issue of "incorrect success rate chances" was not a bug. It certainly is!

What I meant was that the OP by @Cairo Hayden was posted in the form of feedback and not as a bug report.

@Datendrache - Your theory does make sense (at least to me) as a possible cause of the issue, and I agree that investigation by @Chris is a great suggestion. As I said in my first post, they do know that the current method of determining success is "off" but I don't think they had enough information to determine the cause and how to fix it. Perhaps with your theory laid out here, they will figure it all out soon!

 

I personally think this bug should be towards the top of the list on getting it worked out. If any developer had actually spent the time it actually takes for one person to go out, mine up all the resources, farm up all the beetles, just to craft one full set of carapacian cloth armor for example, you would know how discouraging it was to take those raw resources to the crafting stations, spend hours refining it, then crafting it and having a high chance to break the armor vs the 95% chance to succeed on crafting it or masterworking/enchanting it.

Sure, I don't expect it to be a walk in the park or everyone would do it, however, this isn't about making it easy, it's time vs reward in this case but the reward in this case is pure disappointment over 80% of the time. Realistically you may spend up to 1 full week, 6-8 hours a day farming up all the mats to make a decent set of cloth armor, assuming it doesn't break 3-4 times in a row that it does quite frequently with a 95% success rate.

Thanks for checking into this bug.

 

Just to add a bit more, I also notice when I have a 95% chance to succeed on gathering a resource, I have also failed 3-4 times in a row. I can't help but think there is a code in gathering that is also shared in the crafting / masterworking / enchanting code. Perhaps this can help in searching for the proper code and find the error.