Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Schmaven]

It is well known that higher skill levels produce better quality weapons.  However, once legendary +5 is reached, does further experience further improve the odds of crafting a masterwork?  I almost have 2 legendary +5 weaponsmiths to do some testing with.  One is at about 27,000 exp (legendary +4), the other is a bit over 50,000 exp.  But the one with higher exp is likely to go insane soon, so I'm not sure how much testing will be accomplished.

[Urist9876]

Even if there is a difference after +5 (probably) it will be offset by focus.

A dwarf so stressed out he's on the verge of insanity probably has a very bad focus and as such produces slower and lower quality items.

Testing with only 2 would require a long time. Even with full focus and +5 legendary skill not every item is a masterwork. To find a small difference in these random results wouldn't be easy.

Testing without similar focus, attributes and preferences would not produce very useable info about skills beyond +5.

For testing purposes you could use dfhack fillneeds while the unit is selected. Messing with preferences in dfhack crashed my recent game - use with caution or make sure it does not affect the test.

[Schmaven]

I forgot about the other variables effecting weapon quality.  They would definitely have to be normalized to test for the effect of profession skill levels.  Mostly I was thinking about whether it was worth training up another weaponsmith as a backup, or just pouring the extra copper into more exp for the 1 weaponsmith.  But given that this is Dwarf Fortress, there is a very high probability of any dwarf dying at any time, so I'm going with multiple weaponsmiths whenever resources make that possible.

[Superdorf]

I've found that once you've got your smith to Legendary+whatever, the best thing you can do is assign said smith to military crosstraining. The extra strength and agility and suchlike from constant sparring makes a tremendous difference for the smithwork-- and makes your smith a little tougher for when things go south!

[anewaname]

If you do want to test if 27k or 50k experience makes a difference, you could use the same dwarf to test, by save scumming and boosting the xps with dfhack the 2nd time.

[kaijyuu]

My understanding is that levels do continue to benefit crafting past level 20, but with rather extreme diminishing returns.

[Schmaven]

Well, in the name of bad science, I simply logged the current 2 dwarves in my fort as the armory was being stocked.  I tried to split the work evenly between the two weaponsmiths I had. 
Key: Weapon Forged (Masterwork-Exceptional-Superior-Fine-Well-Ordinary) (Star = * + -   )

Legendary +5 Weaponsmith (29,900 exp) 61 year old dwarf: Stress: -63 [Likes Battle Axes]
After crafting 50 (32,900 exp) Stress: -527
After crafting 100 (35,900 exp) Stress: -771
Overall unfocused by unmet needs

20 Short swords (5-13-2-0-0-0)
60 Battle axes (20-37-1-1-1-0)
20 Spears (4-11-4-0-1-0)

29% Masterworks
61% Exceptional
 7% Superior
 1% Fine
 2% Well Crafted

Legendary +5 Weaponsmith (58,400 exp) 87 year old dwarf, Stress: 4,200
After crafting 50 (61,400 exp) Stress: 2,800
After crafting 100 (64,400 exp) Stress: 388
Overall unfocused by unmet needs

50 Short swords (8-40-1-1-0-0)
 0 Battle axes (0-0-0-0-0-0)
50 Spears (14-36-0-0-0-0)

22% Masterworks
76% Exceptional
 1% Superior
 1% Fine

There were too many variables left uncontrolled for to draw any meaningful conclusions from this data.  However I am now left with a hypothesis to test:  Perhaps crafting masterworks is one way to reduce stress.  If so, it is fairly easy to do if you have a modest pigtail farm.  However, it could have to do with the value they created.  It worked 100% of the time in these 2 cases, but could use some further testing on more advanced cases of stress, and a much larger sample size.  These 2 dwarves were using steel, and so each piece is much much more valuable (and rare) than cloth.  So in that case you would have to choose which dwarves to prioritize de-stressing unless you have a lot of gold or other valuable metal to pour into making hundreds upon hundreds of items.  But I don't want to read too far into it. 

Some item values for comparison:
Masterwork steel battle axe: 12,240 Urists
Exceptional steel battle axe: 5,100 Urists
Masterwork steel short sword: 5,040 Urists
Exceptional steel short sword: 2,100 Urists
Masterwork steel spear: 6,480 Urists
Exceptiona steel spear: 2,700 Urists

Masterwork pigtail robe: 1,052 Urists
Exceptional pigtail robe: 450 Urists
Masterwork pigtail sock: 404 Urists

[Blue_Dwarf]

I'm not sure how focus factors in, but the fact that your first dwarf likes Battle Axes is skewing the results. Item preference is huge.

[Fleeting Frames]

From 47.01 file changes.txt:

   new reaction tags:
      [SKILL_ROLL_RANGE:<basic range>:<multiplier>]
         The skill roll is: random(basic range) + random((skill level * multipler)/2 + 1) + random((skill level * multipler)/2 + 1)
         random(x) returns a number between 0 and x-1, so basic range is always 1 or more.  The default is 11.
         The default multiplier is 5.

What this means is that skill impact to quality "score" roll is linear. Looking at the link in Blue_Dwarf's sig, I then guess the threshold for ⛭ is roughly 69, back-of-the-enveloping it. As zero-skill dabbling dwarf has 1,16% chance for masterwork with 2k attrs, either attributes add skill beyond the cap or they add their own roll (or something more complex, as they're intended to mainly matter at low skill levels).

It also means the L+5 cap is hardcap (otherwise it's easy to see how L+10 would have over 50% chance of masterwork, and so on).

Further note: Badly hit by statuses, such as Distracted! people don't really produce masterworks as normal. Focus is up to ±50% to skill roll, which means the cap is applied before focus, as otherwise L+35 dwarves could still maintain the production rate.

[HrumpfOfDoom]

Urist9876, were you using pref-adjust when it crashed your game? There's a bug with it. I got turned onto the non-bugged version in another post, myself.

[Sarmatian123]

I had lots of psycho Dwarves to deal with. This crafting-cure works only temporarily.

I think generally the crafting workshop with 1 auto-job per month for stressed Dwarves is based on it. I guess expensive steel makes better impact then cheap bone. Producing large quantities gives better stress reduction though. Plus for valuable material or material from Dwarf's preferences. Plus for preferred item.

Though easily stressed Dwarves will go fast back to stressed. I had too many legendary craftsmen put in suicide squad, because how temporarily this cure works. Even when producing large quantities of items. Then I have a fragile to stress Dwarves (He is always tense and jittery due rain), who does fine, because they are getting a pretty decent drinks (their favorite), eating a legendary meal (their favorite) and being somewhat focused with satisfied needs. FOCUSED! Focus is the bogeymen of stress in DF. Focus will not rise stress on its own, but without focus, your stressed Dwarves will never ever relax from stress on their own. It just works that way. Only way to fix stress in manageable way is to fix stressed Dwarf focus. Not all Dwarves can be fixed.

[Urist9876]

Yes I was using pref adjust. I wasn't even sure if the crash was because of a buggy script or passing invalid arguments. I'll try an updated version sometime later.