What will crafting quality products look like?

Because there are too many crafting/profession skills for each to have its own forum.
Post Reply
User avatar
nobody
Posts: 501
Joined: Wed Jun 17, 2020 7:43 am
Contact:

What will crafting quality products look like?

Post by nobody »

From past conversations and chats, and my own speculation has doubtless crept in also, I've thrown together some ideas of what the 'algorithm' of crafting high quality goods might look like. TLDR; this is a weird post in which I speculate on a good way to set up crafting higher quality items within the parameters vaguely discussed in the past.
  • Reimplment quality variation: With the recipe point removal, all quality levels became flat (1-399 always gives average, 400+ gives fine) and need to be tiered again based on skill.
  • Add random: Crafted item quality levels become slightly more random and based on skill, with a minimum of terrible (-3) and a maximum of fine (+1) or average (+0)).
  • Add tracking: The quality at creation is stored as an item property.
  • Add improvment: Crafters should then be able to improve the quality of items passively (with a multi-hour cool down like study or scavenge), and possibly an active method. The maximum improvement on an item should be capped at +2 or +3, depending on whether the "add random" cap is fine (+1) or average (+0) (a bottom tier creation at terrible (-3) item could become cheap (-1) or average (+0), and a top tier creation (either fine (+1) or average (+0)) item could become exquisite (+3)), with other limitations based on skill.
  • Add wear: Have the improvement level drop whenever an item gets very damaged and when the repair is botched.
Reimplement quality variation
This would be a fix to a known bug, I don't have a bunch of extra thoughts on it, other than that the add improvment idea (details below) will make this less painful for starting crafters.

Add random
There are a lot of ways to add randomization. Taking a minimum (or maximum) of the rolls produces order statistics, and fits well with the already established rolling system in the game because having positive rerolls is essentially the maximum version and having negative rerolls is the minimum version. Summing the rolls produces will either produce a t-distribution or a normal distribution depending on the number of rolls. There could also be a counting-successes approach (each roll is against an individual target number and then the result is a count of successes), which would effectively create a binomial distribution. Any of these distributions could be made to work, but the order statistic approach is probably the best for integration with existing game mechanics and entirely avoids (1) the necessary question for all the other distributions: how many times do you roll? and (2) the issue of morale (or similar mechanics) adding rerolls in non- order-statistic distributions would then make them behave in ways that are... complicated. So, as much as I like the idea of an additive approach, I have to recommend the order statistic approach. Rerolls should come from morale, the relevant craft skill, and possibly from tools (maybe a 5% chance per +1 of a positive reroll, or per -1 for a negative reroll, so using a terrible tool would cause one to get a negative reroll 15% of the time?), and materials (similar bonus to tools, but most materials are going to be average with possible exception of intermediate materials like small blocks in stoneworking, wooden blocks in woodworking, and tempered clay in pottery). I think I favor the idea of this capping at average to keep the most common market quality at average.

Add tracking
Whatever method gets used for adding randomness to the item creation process, the result should get stored as an item property. I'm imagining this will be the underlying number that was rolled, from which the initial quality is derived. That serves as the item's permanent baseline, with the item's current quality based on its initial value plus its current state, which is increased by improvement and decreased by wear.

Add improvement
The "passive" bit here is perhaps misleading, it is still an active choice, but might be a 30-second roundtime and an 8-hour cooldown (like study) to increase the post-production quality of an item. Probably multiple applications will be required per tier of quality improvement, possibly with increasing costs/diminishing returns on the same item (eg, if it is 1 use to move an item from terrible(-3) to crude(-2), it might be 2 uses to move from crude(-2) to cheap(-1)). It might be reasonable to give the merchant guild a "bigger tank" where they have 2 uses that each have an 8-hour cooldown, with maybe 3 uses for the crafter-specific class of the merchant guild as a class perk. The additive benefit could also be random, with rerolls as for the initial creation, and the improvement amounts all getting stored in a secondary variable apart from the initial quality. This should cap at +2 or +3 total improvement, in general a terrible(-3) item shouldn't be able to make its way to exquisite(+3), but that is definitely just my opinion. I imagine some kind of cap on additional improvement tied to skill would also be good.

This quality improvement mechanism allows starting crafters to turn their terrible(-3) creations into cheap(-1) or average(+0) versions that can actually be sold to market, and solves the problem of having way too many bad starter pieces. At least, kind of - it would be very easy to make a quite large back-log of bad pieces in need of fixing (especially if you end up running out of skill points and rely on using that craft for exp to get the next level) and it would probably make more sense to throw them away and focus one's limited improvement resources on making better-than-average goods rather than making terrible ones salable. And that is why I also would love to see an active version of the improvement feature.

The active version of this basically mirrors the initial crafting process but has maybe double the steps, and functions like the normal craft activity (costs roundtime, energy, wear on tools, grants practice and exp, etc.), and has a lower improvement cap than the passive approach above. So you might be able to improve a terrible(-3) item to cheap(-1) more quickly with this, but to make it average(+0) when it started terrible(-3) would require the last bit to be done with the passive ability. For items created at cheap(-1) it should probably cap at average(+0), and for items created at average(+0) it should cap at fine(+1). This would make it so that skilled crafters can always quickly produce an above-average piece, but need to really invest time to get something better than fine(+1). It would allow master crafters to still sell average quality gear to newbies that are broke, and also allow them to invest a good amount of time to make a superior(+2) or exquisite(+3) item for a character that actually has the riln to pay them for their time. The added roundtime will mean that crafters that have easy access to materials will probably be better off trading time for riln by producing more average goods rather than flooding the market with higher quality goods.

Some considration should be given to mechanics of multiple crafters working on the same item and possibly-skill based caps there.

Add wear
The idea is that using an item beyond scuffed would lower its post-production quality attribute, until eventually it decreased in quality. This could go negative, so an item created average and worn down could become cheap(-1) or crude(-2) after repeated wearing down. Because the item could be improved again though, people wouldn't need to be overly hesitant with their rare special items (at least, the ones made from repairable materials). Having a bad repair job might also make sense for this.

Examples
Suppose this hypothetical table initial quality rolls and corresponding finished item initial quality levels:
1-50: terrible(-3)
51-100: crude(-2)
101-150: cheap(-1)
151-800: average(+0)

And suppose this table with values from min(initial_quality,300) + post_quality:
1-50: terrible(-3)
51-100: crude(-2)
101-150: cheap(-1)
151-300: average(+0)
301-500: fine(+1)
501-750: superior(+2)
751-1000: exquisite(+3)

If the addition per passive use is 10% of skill ranks, it'd be a short climb for high skilled crafters, but a much longer one for low-skill crafters. And maybe a 10% of craft roll result for an active use that caps lower. Here are some examples with my not-at-all-optimized tables.

Leo the Leet put 1 skill point in all the crafts and then bought exquisite tools and keeps their morale boosted. Most everything they produce is going to be crude initially and they'd have a really difficult time getting any where with additions since that's adding .1 per 8 hours for passive and between .1 and 10.1 per active attempt. If the post-production quality is capped at crafter skill, they wouldn't benefit from active or passive improvement at all.

Hector the Hobbyist has 150 skill points in woodworking and uses average tools, 40% of their goods are average, 20% each to cheap, crude, and terrible (unless they get some rerolls, I've assumed a default of 0 rerolls, but who knows, maybe 2 or -7 is more optimal). Each passive use will add 15, so if they take an average piece, they could probably make it fine in 3.5 days using only passive work. Or with active work they could improve that average to fine after six active sessions of improvement if they got pretty lucky with the rolls, but it could only be improved passively after hitting fine. If post-production quality is capped at crafter skill, Hector could produce fine with 150 points and some time investment, but they couldn't ever reach superior (they cap at initial quality of 250 and added quality of 150 for 400 total).

And, because I'm better at tables than stories, below is a table of skill ranks needed to achieve a given quality under the haphazard assumptions above
1-25, cheap(-1)
26-100, average(+0)
101-200, fine(+1)
201-450, superior(+2)
451-700, exquisite(+3)

So, that's probably too generous compared to the current set up, but it isn't bad for a number example built on the fly. Probably. Lets look at another example to test it: Marty the Minimalist wants to put 201 points in to squeeze into superior. They could roll a 301 on the crafting roll and that'd go down to 300 on the initial and set it to average. Then with active improvement they could roll perfect again and add 30.1 to get to 330.1 which is now fine(+1). Since it is fine(+1), now it can only be improved with passive improvement, at 20.1 per it'd take 3 days if they hit every 8 hours or 4.5 days if they only hit 2 per day. Yeah, that's probably too generous. I may come up with more number examples later, but I think it is a good enough outline to ask, 'does this line up with the vision, or does it miss the mark?'
Post Reply