Starman complain about "Portalarium Crafting Math" #6 with data.

Ive posted at least half a dozen of these complaints about the math of the crafting system. Across the board ... blacksmith or enchanting / masterworks your math is bad in this game. I think Im actually getting kind of angry about this now since Its a catastrophic waste of time.

19 Bronze hammers made now at 24% exceptional chance and I have 1 mastercraft.

At least 25 bronze chest chain pieces (and 100 others) with 2 matercrafts. Supposed to see 25% and Im lucky to see 10%.

hundreds of enchanting and Masterworks (silver / gold) with massive failure even with 95% chances.

This is pretty bad. Think about, it... So I finally make 1 exceptional item... go to enchant it with a 95% chance and poof.. its dead. That's a good 10 hours of farming and tens of thousands of gold spent with a result of nothing.

 

Are you talking about the exceptional +1 for the durability on the hammers and chest?

 

aye! Exceptional would give those +1 and they would have 150 dur.

 

I believe if they sat down and mapped out thes complaints that all lead back to a bad math calculation. Hmmmm.... And Fix that, maybe archers would not CRIT everyother hit other weeping would CRIT a reasonable amount of time. People could craft with a certain degree of knowledge about there chances. It is just math and that should not be this complicated to fix. If they can't maybe they should open the hood and let some of the testers take a look at the code, someone may see something port is missing.

 

I don't see where you have offered statistical proof the math is incorrect. You do understand that it's statistically possible to have a 25% chance to do something, but to fail a lot more than 1 out of 4 times, right?

 

Maths is irrelevant if it's not fun. If it isn't fun it doesn't matter if the maths is right.

They need a pity timer to keep statistical variance within reasonable parameters when there are so many steps where you can destroy hours of investment

 

As was pointed out here, the number works for n -> inf. The RNG is ok, I tested it on mining, I was doing some n = 120. You need to realise that every sample (attempt) is independent on the results of preceeding samples (no memory of previous results). So if you take a set of many samples, the probability of successes being within this sample goes close to your 0.95. But if you use a tiny sample window and move it across this set of samples, probabilities within those subsets WILL BE SPREAD in a huge range. I don't want to go too deep into the statistics, I just want to emphasise that you CANNOT expect 0.95 on tiny set of samples in system of independent results (no memory of previous results). So the randomness basically says you can have more fails in a row as long as the probability of the case in > 0. Please do not expect that when you got 0.9 probability of success, you will reach 9 successes of every your windows of 10 attemps. What people complaining about failures rate usually mean, they want to have assured the probability value to get aplied to tiny sets of attempts. But that only can be achieved with a semi-random system having memory of previous results (feedback system), which is easily exploitable.

 

Each try is independent. Like how I fail harvesting sometimes at 91% 3 times in a row.

I understand the Math but mannnn, it's still irritating!

 

This actually might not be a bad idea - it's true that the accuracy of the tooltips is less important than player perception, and the larger the player base the more cases there will be of people failing many many times in a row.

I hope that with all the data available to the devs behind the scenes, they can get a much clearer picture than us of whether the math is working correctly on a global scale, because individual players' feedback will almost certainly suffer from bias (mostly because people won't bother collecting and reporting results of extended studies unless they believe something is wrong) so peripheral cases will be highlighted disproportionate to their actual rate of occurrence. The best way to get reliable test results is sampling the broader population without their conscious involvement. But people will still be unhappy if they personally feel they're failing too often, so maybe accurate maths is not such a good thing to have in a game.

 

The key here is "over a large enough sample size, these percentages should be reflected."
However, due to the exorbitant costs of materials, a player being able to do this sort of testing is not really feasible.
I would imagine that these sort of stats are tracked by Portalarium and they could see if it is working correctly.

And you know what? Over 10,000 crafts, maybe there really are 25% exceptionals.
That doesn't help the average Joe crafter who has just burned 2000 ore trying to make a single chestpiece which then blew up on the first enchant.
Nope, he's out several hours of work with nothing to show for it.

As I've said elsewhere, the math may be right and it may be wrong, but the crafting system is too expensive, too punitive, and you are too likely to be left with nothing but a handful of metal scraps for significant time/resource investment. In the cases where you aren't left with nothing, you have something with stats/buffs you didn't want and now have to start again. This is a really crappy design.

If it took 10 ingots to make a full set of armor, the breakage and randomness would be fine.

Either:
Costs need to be toned down by about 90% across the board,
or
Breakage rates should be scaled up to zero failure on 1st/2nd go at grandmaster level
Exceptional chance should be 100% at grandmaster level
or
Randomness factor should be eliminated to allow successful ench/mw to actually give you what you want.

If you have breakage AND random effects AND crappy chance of exceptional item AND ridiculous resource costs the whole system just becomes awful and the only reason people do it is there is no other way to get gear.

 

Heh the average joe like me that scored 4/5 bows being exceptional on my last craft? Lol.

Wouldnt making it any easier butcher the economy? I mean once u MAKE that armor, or hammer, you ll be using for weeks and weeks. I got lucky with 1 bow and been using it over a month.

 

My data set for this post with 19 War hammers with a success rate of .052%

I also established that in previous posts my findings are actually very similar. For example, I have made over 25 Bronze chain chests and only made 2 exceptional. That is a .8% chance. I have made 100+ more various other chain pieces and I would also estimate a 5% = 10% success rate on those.

I have dumped 1000's of gold and silver ore into enchanting and master crafting with significantly lower chances then the ones they list. I cant even count the number of 95% failures I have had.

This is just the items on my vendor (I have dozens more in chests and many more on guild vendors). When it comes to enchanting and mastercrafting, it feels like 25% of the things I make dont even get this far without being broken. So each item say listed below represents actually 3 more crafted and destroyed.

 

The part that you're sure about is a very small sample size, and more importantly, you can't decide to cite data after the fact. That just leads to biased samples because people are much more likely to cite data when they run below expectation than they are when they don't.

 

I only know 2 things about this random crafting system:

1. The percentages shown at the tables don't appear to be anywhere near what they actually are. I know each attempt is a separate roll of the dice. But, there should be some type of failure counter that can auto-adjust the calculation so that multiple attempts average the stated percentage. Losing hard earned materials to a loaded set of dice is not fun!

2. I left another game and stumbled across SOTA because their crafting was random too and I got tired of losing things I had already crafted because of a random number generator. When I found SOTA and invested in my accounts, the crafting system didn't have any items being destroyed on a failure and that was a big deciding factor in my investment of this game.

I would much rather have finite, 100% chances to create something specific... and then have 1 specific ingredient to each possibility be a rare or hard to get item. That way, once you finally get that 1 thing you need, you can start making that personalized piece of gear without fear of losing it to the RNG Gods. Make the ingredients hard to get, not the crafting combine itself!

 

The good news is that if the numbers are really that far off, an enterprising avatar could use that information to make a lot of money. If someone opens up a crafting casino that has 4:1 payouts on exceptionals and similar bets, I would give my business until I was broke.

 

Do you craft?

 

math is easier than your question.
if you do something 100 times, and has 25% of posivilidades to get something, it simply means that 25 times out of 100, will happen.
Mathematics is never erroneous, we who were we wrong.
Nothing to do with 1/4. ??¿¿

 

There is a problem. Maybe it's not the RNG. Maybe it's not the percentage math. However, with this many people complaining, there is a problem.

The problem is that it take a lot of effort to gather the resources required to craft anything, and we are often left with nothing. If you do both sides of it--gathering & crafting--then you sink a lot of time into it and it is frustrating to have little to show for it.

The crafting system needs an overhaul, not just for plus weapons and armor, but for regular, every day items. It should rely less on a RNG, or at least less on the RNG for complete and utter failure. Perhaps a failed item isn't destroyed, but takes a small step backward and cannot be upgraded any more. There needs to be a better balanced tiered system for items that built upon previously crafted items, requiring higher levels as it goes up and increasing the difficulty appropriately.

 

I said this elsewhere:

When I play D&D, I don't roll a 20 5% of the time, every time. When I roll the die 20 times, one of those rolls is not guaranteed to be a 20.

If I were to roll the die 1000 times, perhaps it will be around 5%.

The point being, every time I roll the die, there is a 5% chance. When I roll the die a second time, there is a 5% chance it will be a 20. The percent doesn't increase if I didn't roll a 20, or lower if I did.

That is why, some of us end up with three masterworks when we do 5 crafts, or 0 masterworks when we do 20.

Like LoneStranger said, the issue isn't the math (which I believe is still sound), its the effort we put in. If it didn't cost us hours upon hours to gather thise materials and have them removed from us, maybe it wouldn't hurt so bad.

But, then, doesn't that make the masterworks we do produce worth that much more?

I dunno. I just don't like that people are complaining about the percentages.

 

@2112Starman and others who feel the RNG is broken, I really think the best thing you can do about it is record *absolutely everything* you craft, all % chances displayed, every success and every fail, and post the complete data in the bug forum. Anecdotal evidence isn't strong enough to show there is a problem with the mechanics, just that the system can cause frustration.

I'm totally in favour of changing the system to one which is not purely random, for the reasons already given in this thread. However complaining that crafting 'does not appear to be random' is about as useful as simply stating that crafting isn't fun. It doesn't actually demonstrate that something is wrong. If you have solid data that indicates there might be a problem - report it as a bug and the devs will better know what to look for!