Crafting RNG

I think that a less RNG based crafting system would not only make crafting better, but it would also make the market much better for consumers. As it is now, almost all crafted gear on merchants is just the failures of people trying to make particular things for themselves.

RNG also really has nothing to do with expense or attainability. A 1% chance to craft an item at an expense of x has the same expectation as a 100% chance to craft the same item at a cost of 100x. Removing or reducing randomness doesn't necessarily have to make anything cheaper or easier.

I like a lot of different ideas that people have had for changing crafting, but reducing randomness doesn't even require any creativity. Instead of scaling success probability with skill, you could just scale material quality with skill and scale crafting results with material quality and quantity. Then people could make the kind of things that they want, but there would still be (diminishing) rewards for investing more time, experience, and materials.

 

Another advantage of a non-random crafting system is that it gives you more useful player-preference data that can be used to help make adjustments. For example, if there is a particular masterwork bonus that no one ever voluntarily chooses over other options, it might be adjusted or replaced with something that is at least situationally useful. It's a lot easier to make that determination without a bunch of dice rolling involved in the crafting process.

 

It's an interesting proposal and one that fits well with the ethos of Shroud, letting us work a lot harder for a slightly better result at the top end, if we so choose. Similar to running dynamic decks, taking skills to silly high numbers, etc.

My personal vote would be against tying it to cotos, but I'm also trying to work out in my head exactly how it would function.

 

I don't agree with all the particulars (chaos is not at all the word I'd use) but at least this seems to mostly tided itself up.

Ammusingly, this is one of the reasons I so love the current system. You have to be clever to minimize your risk and loss.

I really do protest the view of "jackpot pieces" (which incidentally not only pay for themselves, but effectively subsidize lower end gear, making better pieces cheaper for others) and the idea that folks are stumbling into high end gear. I have put a lot of practice, effort, and calculation into maximizing my output. For example, one of the best ways to get a top end piece of gear isn't to set out trying to make it, it's to let it develop naturally. If a Warlock Chain rolls a Major Strength, then I might take it in the direction of a death mage chain over the air mage change that a Major Dex would prompt. 90% of the time I enchant pieces first. What do I start Quivers of Jealousy on the masterwork side? Because if I proc a Major Int over there that's no longer a dex quiver it's a mage quiver. (Mage quiver is perhaps an odd thought, but Quiver of Jealousy gives draw speed and is a best-in-slot back piece for many builds.)

Granted this is a style that is well suited to, well, crafters, as opposed to adventurers who are just looking to make their own gear or someone with high levels trying to make a piece or two for friends. But I sometimes feel like the people who enjoy the crafting system are rolling with the punches, getting their jabs in, and making great sport of it. Then I get the impression that some folks treat it like this:

 

I won't write a book to contest it, but a lot of what is being said here is completely contrary to my experience as a crafter. I just want to throw that in since I've already replied to other bits and don't want it to seem like I agree with what I've passed over (as is often the case). I suppose if anyone wants to hear me blather on about this subject there's several hours of video of me talking about it.

Just to voice what's probably my biggest concern, it's made most clear in the case of epic artifacts. What is the cost mechanism going to be that lets people get perfectly worked epic artifacts, something that is indeed very difficult (but not impossible, often magnitudes easier than the pants I mentioned). What's going to be the input cost that makes something that previously was impossible on a large scale doable without it being pay-to-win or power creep?

And to toss in a side note (see, I really do blather), a lot of the 100-tries pants are 17str 28dex... they're better than what I use as a reasonably high level adventurer. Others are only partially worked and may still come out better than the top end pair I already made.

To end, if the idea is to make the non-RNG route cost more, but you already feel "When I work out single pieces that are costing in excess of 500k in materials at market rates, it is unfathomable that someone is going to actually pay that." then what is the point? If you already think paying what the gear costs to make is too much, what good is making it available at a higher price?

 

Just as a reminder:
This wishlist request is 100% about avoiding frustration from RNG. Even at higher material costs.

It's not about shifting market prices. If it did, that's a side effect.

It's not about creating better opportunities to sell gear.

It's not about making something concretely better than RNG crafting (quite the opposite, pay more for more control and less XP transferred).

It's not about ditching RNG crafting.

 

All fair points. I don't think its a simple black or white either way, I just think the current system is pretty hateful and if we change it, it must be done thoughtfully with regards to resource economy and other factors.

Regarding this cost point though, the reason why the cost is absurd is because it is so very unreliable and relies on pure luck. If you had made those pants on the first try instead of the 100th, the cost you are right, is probably too cheap and trivializes end-game gear. I'm not suggesting we go that extreme, either. However, if we establish what a good cost input is (probably somewhere in between the two), we can cost the recipe accordingly. Then those who wish to go the cheaper RNG loot don't actually lose anything, you just create an alternate route to a result for the people who despise the RNG and would rather pay a bit more to guarantee the outcome.

Its kinda along the lines of when you salvage an uncommon artifact vs. a rare to get a gem. Law of averages ,crushing uncommon ones you're probably better off, but if you want to guarantee it, you can increase the cost of your inputs. This works.

 

I'm not sure I am following. If we determine the average cost (which is what you seem to suggest when you say "if we establish what a good cost input is (probably somewhere in between the two)") of a piece under the current system and make a way to achieve the same result by paying more, the cost of going that route will be more "absurd".

I am likely missing some piece of logic here, because I cannot find a way to read this that doesn't boil down to "RNG makes crafting absurdly expensive so let's make a second system that gives the same end result but costs a lot more".

Are you also proposing just making gear much cheaper/obtainable overall, regardless of the method used to make it?

 

Sorry if this is unclear.

Under the current system, the cost of creating a specific item is completely determined by luck. It could cost player A 5000 gold to create the item (1st try success, very lucky) or player B 500,000 gold (or 5 million gold depending on what we're talking about) to make the exact same item (100th try)

For most people, it will be somewhere in between these two extremes. And in an increasing number of cases, the player will give up completely without ever succeeding. They'll invest enough to say, do 20 goes, and take the best result, even if its not what they wanted (accept mediocrity, move on, go gather materials for another month and have another go).

The problem is the inconsistency and possibility of not getting the outcome you want at all.

The suggestion is to provide an alternative.
Taking your chances with the RNG and trying your luck would be unchanged as it is now.

Going back to the examples, I agree that Player A got off cheap, they got very lucky, and they got an end game quality item for way too cheap. Removing the RNG with no cost increase would make us all Player A. And I agree, that's probably too easy. But I'm NOT suggesting that we jack the cost all the way up to the absurd other extreme either. 10 or so attempts is probably a reasonable benchmark for the removal of random stat/skill choices. Bear in mind I'm not suggesting that we eliminate the possibility of failure/durability loss. For my part, I just want to be able to say "Right, I want to put str, major str, and minor str" on this thing. If I fail this a few times / trash the durability - I'll make another. I might have to make 4 or 5. If these are all 10x base cost, I'm still spending half of what I was making 100 of something under the current schema. Over time, the costs will likely even out, since with RNG sometimes you'll get lucky and sometimes you wont. The notion is to make individual crafts more predictably and consistently costable.

To do this, I would suggest an additional component, either another resource, or some sort of crafted/prepared catalyst be added to the enchant/mw recipe which allows the crafter to choose the bonus they want (full list selectable on success). That's probably the simplest way. That component should increase the cost of the enchant/mw action significantly enough to address resource usage impact but would still be a better case scenario than the expenditure of 100 attempts. I'd suggest just making a special sort of ingot, like, I don't know, infused gold ingots/infused silver ingots, and to infuse an ingot you need to provide additional materials. Would make an excellent and much needed cash sink, too, if there was an NPC buyable component here. For those who were happy to take their chances and do it the cheap way, (perhaps they are working on lower end gear, or don't know what they want, or simply are on a budget), nothing would change. Using infused ingots, you'd get the full whack of choices but you'd pay more for the consistency.
Likely, people would only do this after the first couple of tries doing it the old way, mix and match, and use the expensive things when its important to get it right. Overall, the economic impact should be negligible but this offers the possibilities for crafting as a service, for the market, and for mitigating an extremely frustrating/not fun system of punishment while still making gear creation require considerable effort/cost.

It isn't about making gear cheaper.
It is about:

• Eliminating thousands upon thousands of botched, unwanted, failed junk cluttering up vendors that nobody wants, and seeing more deliberate/intentional pieces built with purpose. (There's likely still be -some-, but you'd eliminate a lot of leftovers by making it so you don't have to create all the extra crap, you can simply spend what you would have spent to create crap, to create your actual item you want. )
• Being able to price things in a consistent manner and understand the cost ceiling using a non RNG methodology
• Allowing players to plan, source materials, and build the gear they want incrementally instead of playing gamble-box with the crafting table until they run out of resources

The overall cost for crafted gear should not significantly change. This is just about flattening the failure curve to something more palatable whilst simultaneously cleaning up a lot of clutter.

Given the choice between "I can go to work today, and I'll make anywhere between $1 or $100 per hour depending on RNG" or, "I can go to work every day and make $50 per hour" personally, I'd go for the steady salary. If others want to gamble - that's their business - its just tiresome that its the only option.

 

I'm standing at a slurpee machine that has 5 flavors.
A slurpee costs $1

I want a blue one, I know I only want a blue one, and only a blue one will do.
The catch is, the machine will randomly give me 1 of the 5 flavors.

I might get lucky and get blue for $1. Awesome.
If I'm unlucky, I could in a truly random system, spend $100 and still not get blue. (There's no adaptive system where, I've gotten a red and I can't get red again.) Just like its possible to flip a coin 10 times and never get heads. Each flip is unrelated to other flips.

However realistically, assuming an even chance of each color, I should expect to buy 5 or so. (20% chance for each color so do it five times). Yes, I know this isn't a perfect application of statistics. This is just an example involving slurpees.

What I'm suggesting is, if we want everyone to have to spend $5 to get the type of slurpee we want, just make Slurpees cost $5, or add the option to pay $5 and choose your own slurpee color.

If you are happy with a random slurpee for $1 you lose nothing. You can still do this.
If you want to try your luck trying to get your favorite slurpee color for less than $5, you're welcome to try this. You can also stop at any point and drink all the slurpees you've bought so far, even if you will not enjoy them.

The only difference is nobody gets screwed into paying $100 for a slurpee because at some point its easier to just pay the $5 with confidence that this give you what you want. Without having to stand on the side of the road trying to sell second-hand slurpees for 10 cents each to get some of your money back.

And all the people stood over there thinking about slurpees, but not buying them AT ALL because they don't want to gamble on getting a yucky flavor they don't want, well, maybe they'll come buy some too if there's a different pricing model. And that's the key here.


------------
The example about "letting gear pieces develop naturally" and going where the RNG takes you to maximize output is valid, if your ultimate goal is to create a massive amount of gear and try to market it very widely hoping you'll find a lid for every pot in that each item you've made has someone who wants that combination of effects. And there is some merit to that, if your end goal is being a bulk producer and seller of items. For most people however, the objective is not to become a marketing mogul, but rather, to simply outfit themselves.

I've just spent several weeks assembling my epic warlock chain. If I don't get the stats I want, sure I could -try- to sell it to someone who does want those stats, but its a very small market and I'll likely not get anywhere near what it cost me to build the thing in the first place. Especially since with many artifacts, the builds they are suited for are pretty limited and most people will want similar sorts of bonuses. And most players who want high end gear can make their own. Really, really small market.

So my choices are:

• Start again, spend more countless hours/days/weeks repeating the process trying to get lucky
• Spend hours shopping around trying to find someone selling one even though it can't be found- (basically, don't bother with crafting)
• Or accept mediocrity and wear the chain with strength on it. It isn't SO bad, it will give me a bit more health.

Those are all bad choices.
I see most people going with #3 - accept that its close enough because the cost differential between "okay" and "good" can be measured sometimes in months of effort. It simply isn't worth it. And if something isn't worth doing, people will eventually, stop doing it.

And this is where the system falls apart. If you were buying slurpees to resell them, then sure, you'd want a bunch of all colors so it doesn't theoretically matter. Most of us aren't doing this.

I realized fairly early on that doing this was dicey at best, since far more pieces never sold than did, and the ones that did, didn't ever come close to covering the input costs. This is because demand for botch jobs is pretty low, and there's no other output for them. This sort of RNG system would be a lot more acceptable if there was some other consumption of all the crap it produces (such as reasonable salvage recovery of materials). If we aren't going to fix the output, we need to fix the input - stop/limit the junk from being created to begin with.

I used to do exactly what you describe - oh, first enchant was Major Strength, this is obviously a STRENGTH cloak. I'd finish putting it together, and set it aside. Ok, this one went more dexy, this is a dex cloak, finish it, set it aside. After an hour, I (might) have the Int cloak I was trying to make, or something close to it, and I have a pile of "these might be ok for someone" things. Trying to sell these, even for less than the base cost of materials used for each individual cloak, usually is impossible. There's just too many people doing the same thing. I set them down on a guild chapterhouse "yard sale lot" for FREE, and they don't even get picked up.

Nowadays, if the first couple of enchants/mws aren't what Iwant, its not even worth spending the additional ingots to complete the item. Just trash it and move on. And watch your pile of resources slowly dwindle as you beat your head against the RNG wall, again and again.

I see no benefits or upsides to this system over one that allows active decision making.
The two systems aren't even mutually exclusive so I'm a bit perplexed why anyone would be against it - you'd still be able to leverage "see where RNG takes you" to get cheaper end results for mass marketing purposes. The only real issue I can identify is that the mass market crafter would potentially lose those customers who are buying the things because the system is too frustrating for them to make it themselves. Then it just becomes a question of what is better for the many vs. the few, imo. We're a small community here and if we want to be bigger we have got to make things more "fun, progress, achievement" and less "grind, waste, frustration."

And as mentioned by others above, letting players actually make choices would help identify desirable vs. undesirable effects and help guide the possibilities of future implementations/additions.
------------

The issue of crafting RNG and its detriment to gameplay and player retention is very well-established. There's easily hundreds of posts touching on the problems with this system dating back to before persistence. Hopefully, someday, if people keep complaining about it maybe it will get addressed.

Anyway, in light of the fact that certain people have decided to take this argument off the forum, tracked me down to my guild discord and resorted to personal attacks insulting my intelligence, I'm going to stop replying here.

 

Yikes.

 

That should not be happening at all. No one in the game should be harassing you in that way at all. Period. Harassing anyone in Our community is completely and utterly unacceptable behavior.

Please feel free to express your options here on the forum. We are a community, and we have different, and various opinions. You are ALWAYS free to express them here.

 

I would like to suggest for devs to consider the following:

When the player chooses to bypass RNG, the game will calculate the overall probability for the selected MW/E sequence. This can be easily made with analytical equations, such as for example 4 MWs without fail: 100% * 95% * 56% * 18% = 9,6%. Monte Carlo approaches could be used as well.

Considering the overall prabability - let's assume our example at 9,6% - determine the average number of tries to craft it. In this case I think, for the sake of this thread, 1/9,6% would do the trick: average 10,44 tries to get it done.
(Note that I'm not being statistically robust here, approaches considering standard deviation and level of confidence would yield scientifically acceptable results)

I'm completely aware that this level of control and convenience could be perceived as cheating the game's systems, that's why I suggest from the very beginning to add "extra" cost, so the RNG free isn't right on the average of the regular RNG crafting.

The devs should decide what kind of extra cost it'll charge players who use the system.

Let's assume they add a +25% material cost, so 10,44 * 1,25 = 13,05. Round it up to 14, and that's it, you'll need to provide materials (including the silver/gold ingots) enough to make 14 pieces, so you get only ONE with the desired specs.

Then there we are, one option is a gamble in which you're more likely to spend less materials and a RNG free option that will charge extra materials for the convenience.

 

Translation - I think this solution would be a reasonable compromise between coswald and everyone else.

 

You are one of the most intelligent thoughtful people on the forum, in my opinion. You've posted so many great ideas over the years, that the devs could significantly improve SotA just by prioritizing your requests on the development roadmap.

Thank you very much for all the excellent feedback you post. I, for one, appreciate it.

----------