Another Bloom/Slag experiment

Ask, answer and discuss any and all topics about the hows, whys, wheres and whens of playing Haven & Hearth.

Another Bloom/Slag experiment

Postby bwbw » Fri Nov 20, 2015 5:58 pm

Experiment:

A long story short: I want to get some "real" numbers behind wrought iron and metal progressing. I did around 500 cast irons, and got only 8 bars increased their quality by 20% over 3 week period. Now I want to get clean experiment with decent numbers.

500 bloodstone ore with Deep artifice skill become 125 bars of Cast Iron

Stated chances:
1-in-2 chance of becoming slag (50%)
1-in-3 chance of becoming cast iron (33.3%)
1-in-6 chance of becoming wrought iron (16.7%)



There's also a 40% chance for the quality of the iron to go down 20%, and there's also a 10% chance for the quality of the iron to go up 20%.



Results:

23 Bars went DOWN 20% (11 Cast Iron + 12 Wrought Iron) ~ 43%
29 Bars stays the SAME quality (13 Cast Iron + 16 Wrought Iron) ~ 55%
1 Bar went UP 20% (1 Cast Iron) ~ 2%

Total: 53 bars (42%) and 72 slag (58%)

I put that slag back into Smelter amd got 2 cast iron back.

Сonclusions:

Bloodstone has 25% chance become Cast Iron
Cast Iron has 42% chance become Bloom and 58% become Slag (with Deep artifice skill)
Bloom has 43% chance loose qality, 2% increase quality and 55% stay the same
Slag has 3% chance become Cast Iron again

72 Slag is almost 3 full smelters. 2 of them get no Cast Iron at all and 2 bars in last one. This is not the first time I'm getting very low amount cast back. Based on my experience, I do believe: chance to get Cast from Slag is most likely counted for Smelter to have any bars and then modify that by Slag chance (10%). I did get that 10% number from the sky.

Our 5 lvl mine required 250 STR and 100q pickaxe to mine Bloodstone. I've used around 50l of water to get 500 bloodstone. With this numbers, metal progression is unrealistic.
bwbw
 
Posts: 4
Joined: Wed Nov 04, 2015 12:49 am

Re: Another Bloom/Slag experiment

Postby Sevenless » Fri Nov 20, 2015 6:15 pm

250 STR is actually low for mining L5. Furthermore, we have no idea if your node is harder or softer than the average. But even if that's all told, a single spiralled anvil set costs ~3500 blood stone ore, 140 smelter fulls. That does sound a little high on the ore count.

I think the real target here is the bloom/slag ratio. You coulda been unlucky with only 50 bars hammered, but 60% slag sounds a little too punishing.
Lucky: haven is so quirky
Lucky: can be so ugly, can be so heartwarming
Sevenless: it is life

The Art of Herding
W16 Casting Rod Cheatsheet
Explanation of the logic behind the cooking system
User avatar
Sevenless
 
Posts: 7609
Joined: Fri Mar 04, 2011 3:55 am
Location: Canada

Re: Another Bloom/Slag experiment

Postby bdew » Fri Nov 20, 2015 6:32 pm

Assuming i remember my statistics correctly, computing the (cumulative) binomial distribution for the given number of attempts and results:

If the chance to get slag is 50% per bar then the chance of getting 72 or more out of 125 attempts is 5.35%
If the chance to get quality increase is 10% then the chance of getting it 1 time out of 53 attempts is 2.21%

Not entirely impossible, but extremely unlucky. Or the given chances are not correct.
User avatar
bdew
 
Posts: 179
Joined: Mon Mar 22, 2010 3:27 pm

Re: Another Bloom/Slag experiment

Postby MagicManICT » Sat Nov 21, 2015 4:51 am

The accuracy (variance) of the sample is a a more or less trivial calculation, and very well could explain variance in the numbers displayed versus what is expected. This is assuming that the old values are still correct. (I've not seen anyone posting something other than that for the new world, and no reason to assume they actually were changed.)
Opinions expressed in this statement are the authors alone and in no way reflect on the game development values of the actual developers.
User avatar
MagicManICT
 
Posts: 18435
Joined: Tue Aug 17, 2010 1:47 am

Re: Another Bloom/Slag experiment

Postby bwbw » Sat Nov 21, 2015 1:25 pm

MagicManICT wrote:The accuracy (variance) of the sample is a a more or less trivial calculation, and very well could explain variance in the numbers displayed versus what is expected. This is assuming that the old values are still correct. (I've not seen anyone posting something other than that for the new world, and no reason to assume they actually were changed.)


The only problem is, that old values calculated without Deep artifice skill. This skill should improve chance of getting Cast/Bloom. It's possible that I was extremely unlucky, but I did this test multiple times. And numbers more or less the same.

100 bloodstone ore become 20-30 bars of cast iron
100 cast iron become 60-70 slag and 30-40 wrought/cast (only 1 in 100 will go up by 20%)
100 slag become 2-5 bars of cast iron

Deep artifice skill affecting Ore -> Cast, but does not affect Cast -> Bloom/Slag and Slag -> Cast. There's no problem with smelting ore, but forging iron (after patch), chances should be more fair. I can live with 40% bars going quality down, if 10% go up.

Forging Wrought Iron now has a slight chance of reducing or increasing the qualities of the end product, as it was in Legacy.


Slight chance qualities go down is 40%
Slight chance qualities go up is 1%

In my case slight chances looks like, 135 bars of going down and 8 bars going up of 500.
bwbw
 
Posts: 4
Joined: Wed Nov 04, 2015 12:49 am

Re: Another Bloom/Slag experiment

Postby bdew » Sun Nov 22, 2015 11:10 am

Under the old system full industry (which deep artifice is supposed to emulate gave 20% of ore smelting into metal and 2/3 chance to turn bloom -> wrought iron.

Note that it doesn't change the bloom/slag ratio, so you would still get 50% slag but the remainder would be 33.3% wrought 16.6% cast.

In your original experiment you got:

72 (57.6%) slag
25 (20.0%) cast
28 (22.4%) wrought

Getting that number of slag is not likely (as i posted above), cast/wrought ratio doesn't match any option either.

With expected 50/33/16 (no sliders/skills) distribution you get a chance 5.87% to get this outcome.
With expected 50/16/33 (full industry / deep artifice) you get 0.5%

In either case - not entirely impossible, but extremely unlikely.

More samples would be nice, but it think this is enought evidence for the devs to look into this and confirm whether this is intended and the chances have changed from legacy, or its a bug.

Edit: Changed some wrong math and incorrect conclusions. Oops, don't do statistics before the second cup of coffee in the morning :P
User avatar
bdew
 
Posts: 179
Joined: Mon Mar 22, 2010 3:27 pm


Return to How do I?

Who is online

Users browsing this forum: Amazon [Bot], Claude [Bot], Google [Bot] and 81 guests