Print

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

[franti]

I spent 4 hours testing all in-game metals for combat effectiveness. I started by randomly picking two metals, pitting them against each other, and then picking another and pitting it against both to find out where it stood, and so on and so forth down the line. I'd pick a new metal, and pit it against the best so far, and if it failed I'd keep moving it back until it won.

The tests were two groups of male dwarves arranged in to two 5x5 squares with a 5 space gap between them. Very standardized.
Every dwarf was a Proficient Fighter, Proficient Swordsdwarf, Skilled Armour User, Skilled Shield User, Adequate Dodger
Every dwarf was armed with a Short Sword, Shield, and Mail Shirt of the metal type being tested.

Why the mail shirt? The test was done primarily to test the effectiveness of the weapon, but certain material attributes are more valuable to armour than weapons, and I also wished to prevent a lucky insta-kill heart/spine stab. I wanted the fight to go on for at least 2-3 pages in the combat logs so I could determine the limb-slicing capabilities of the material.

After the fight, I took the proportion of dwarves that had won with a metal and ran it through a significance test. If a side won, but so narrowly it isn't statistically significant, I ran it again and again until it was.

A dwarf that suffered an injury that would result in a bleed-out or that would not be repaired to functionality after a reasonable bout in a hospital is counted as a loss even if the dwarf did not die in the fight: Spinal Injury, Heart Injury, a lost body part bigger than a finger or two. My way of determining it: if the dwarf spent the year in the best dwarven hospital imaginable, but either died anyway or survived but without a reasonably similar functionality, he's a casualty.

Margins:
In a 25 VS 25 fight, a victory of 10 or more is a Steep Margin, 10-5 is a Medium Margin, and >5 is a Small margin. If the margin is very small, I may re-do the test until I'm satisfied.

I mainly ran this test so you know what metals to keep your weaponsmith away from when he goes into a mood.

Results, from least effective to most effective:
Lay Pewter (2 Tin, 1 Copper, 1 Lead)
Fine Pewter (3 Tin, 1 Copper)
Trifle Pewter (2 Tin, 1 Copper)
Nickel Silver (2 Nickel, 1 Zinc, 1 Copper)
Lead
Zinc
Nickel
Silver
Sterling Silver (3 Silver, 1 Copper)
Tin
Copper
Gold
Aluminum
Bismuth
Electrum (1 Silver, 1 Gold)
Rose Gold (3 Gold, 1 Copper)
Billon (1 Silver, 1 Copper)
Black Bronze (2 Copper, 1 Silver, 1 Gold)
Brass (1 Copper, 1 Zinc)
Platinum
Bismuth Bronze (2 Copper, 1 Tin, 1 Bismuth)*
Bronze (1 Copper, 1 Tin)*
Iron
Pig Iron
Steel
Adamantine

 
*It's worth noting that for the Bronze VS Bismuth Bronze, I had to run the test SEVEN times before I could get a reasonably accurate outcome. I don't want anybody to see this and decide that they're not going to use that Bismuth they just mined out because they think it's not worth it: The difference is so small that the extra armour you'd get would make up for it: Enough Copper and Tin bars to make 6 Bronze Bars could make 8 Bismuth Bronze bars if you had Bismuth available. That swings the tide in favor of Bismuth Bronze in the real game.

Another thing worth noting: the Black Bronze VS Bismuth Bronze and Black Bronze VS Brass tests were fairly close, but Platinum absolutely destroyed Black Bronze.

It's also, I think, worth adding [ITEM_WEAPON] to the Brass entry, because it's much better than copper and nearly as good as bronze: I had to run the test twice to get an outcome.

If anybody wants to know the margin of a test, or thinks a re-test is in order, let me know.
I didn't see any instances of A > B and B > C but C > A, but I think you might if you test Blunt weapons.

[CriticallyAshamed]

Did you only run one test of matchups where you considered there to be a decisive victory? Because decisive victories may be a fluke compounded by the snowball effect, once one dwarf dies that is one less heavy chunk of DPS for that team.

More importantly, it should be noted that many of those "bad" metals like silver and lead actually make great blunt weapons... Adamantine on the other hand.

[o_O[WTFace]]

On some of my tests between two identically armed, armored and skilled groups, I found that a one would win with as many as 27/30 surviving.  1v1 tests are alot more work, but I suspect they are more accurate.  Those were full armored humans though.  I suspect full armor tests are way wayyyy more random because most of the combat is decided by random criticals and loosing too many teeth. 

[franti]

Did you only run one test of matchups where you considered there to be a decisive victory? Because decisive victories may be a fluke compounded by the snowball effect, once one dwarf dies that is one less heavy chunk of DPS for that team.

More importantly, it should be noted that many of those "bad" metals like silver and lead actually make great blunt weapons... Adamantine on the other hand.

The fights are 25 VS 25 to prevent a single, lucky death from completely turning it, and they're put in a square so they don't all arrive at once and gang up on/get ganged up on.
If a fight was close enough that I'm not sure, even after a significance test, I re-ran it.

And most crappy edged weapon materials are so because they are soft, and therefore deform on impact regardless of what shape they're in. The best all-around blunt material, you could argue, is Bismuth Bronze because it's denser than Iron or Steel but holds it's shape better than Silver or Lead. Although, Bronze is nearly as good and much easier to get (Bismuthine being a rare ore)
Still, I'm not even going to suggest that these results would be even remotely similar in a Blunt test. That's a whole nother post.

[franti]

On some of my tests between two identically armed, armored and skilled dwarves, I found that a group would win with as many as 27/30 surviving.  1v1 tests are alot more work, but I suspect they are more accurate.

At first I would do a bunch of 1 VS 1 to see if the results differed, but the differences were never statistically significant, and I found that lucky strikes are more damming in 1 VS 1, as far as an accurate outcome goes.
However, it's important to know that the higher your subject count, the more accurate your results: A single 1 VS 1 is less accurate than five 5 VS 5s or one-hundred 100 VS 100s.

If a test is close, I'd run it again several times to make sure. That's why it took me 4 hours. I do think these tests accurately reflect the weapon-capabilities of the materials.

[Darth_Karl]

And most crappy edged weapon materials are so because they are soft, and therefore deform on impact regardless of what shape they're in.

JUst checking, but you are aware weapon deformation is not modelled in game. As such you can hit soone with a feather pillow and it's treated as a solid non-deformable object. Hense density is all that matters for blunt weapons. As such platinum blunt weapons are god on legs.

[Corneria]

Not complete by any means.

[franti]

Not complete by any means.

Is there a metal I missed?

[IT 000]

At first I would do a bunch of 1 VS 1 to see if the results differed, but the differences were never statistically significant, and I found that lucky strikes are more damming in 1 VS 1, as far as an accurate outcome goes.
However, it's important to know that the higher your subject count, the more accurate your results: A single 1 VS 1 is less accurate than five 5 VS 5s or one-hundred 100 VS 100s.

Personally I disagree. Yes, you need to run multiple 1v1 tests in order to get accurate results. But if you run 5v5 it's prone to dwarves ganging up on each other and skewing the results.

Still, considering they are armored, getting through it is either hit-or-miss so I would still consider the results accurate.

[franti]

At first I would do a bunch of 1 VS 1 to see if the results differed, but the differences were never statistically significant, and I found that lucky strikes are more damming in 1 VS 1, as far as an accurate outcome goes.
However, it's important to know that the higher your subject count, the more accurate your results: A single 1 VS 1 is less accurate than five 5 VS 5s or one-hundred 100 VS 100s.

Personally I disagree. Yes, you need to run multiple 1v1 tests in order to get accurate results. But if you run 5v5 it's prone to dwarves ganging up on each other and skewing the results.

Still, considering they are armored, getting through it is either hit-or-miss so I would still consider the results accurate.

If a dwarf in a 1vs1 suffers a hand injury, it's over for him. He's done. Even if it's Trifle Pewter VS Adamantine, if he can't hold his sword/shield, he will die quickly.
But, being as they're armoured somewhat and the n (sample size) for the tests was usually well over 20 (the required minimum), I would say these results are pretty good.
There isn't any one metal that looks out of place, although I was surprised Zinc lost to Nickel by such a large margin.

[IT 000]

If a dwarf in a 1vs1 suffers a hand injury, it's over for him. He's done. Even if it's Trifle Pewter VS Adamantine, if he can't hold his sword/shield, he will die quickly.

This is why you test it multiple times to reduce the odds of this happening. And since they're armored Trifle Pewter will never penetrate adamantine except if it's a blunt weapon.

[franti]

And most crappy edged weapon materials are so because they are soft, and therefore deform on impact regardless of what shape they're in.

JUst checking, but you are aware weapon deformation is not modelled in game. As such you can hit soone with a feather pillow and it's treated as a solid non-deformable object. Hense density is all that matters for blunt weapons. As such platinum blunt weapons are god on legs.

I don't use blunt a lot, so I'm no expert, but I'm fairly sure that hardness also matters somewhat: Lead is dense, but not that hard. Platinum is both dense and hard. Slade, if you add it to the game, is incredibly dense, and is either the 2nd or the 3rd hardest material in the game. Silver is more dense than Bismuth Bronze, but Bismuth Bronze is considered by some to be the best blunt material because of it's decent mix of density and hardness.
I think some science is in order.

[franti]

If a dwarf in a 1vs1 suffers a hand injury, it's over for him. He's done. Even if it's Trifle Pewter VS Adamantine, if he can't hold his sword/shield, he will die quickly.

This is why you test it multiple times to reduce the odds of this happening. And since they're armored Trifle Pewter will never penetrate adamantine except if it's a blunt weapon.

If you fully armour the dwarves, I find that the losing metal never penetrates the winning metal, and you end up with a large margin and you don't get an accurate representation of the capabilities of the weapon material vs organic tissue.
It's simple statistics: a higher sample = better accuracy. More tests, and tests with larger groups, help to sort out the "flukes" from the "freaking good".
Still, I doubt any of these results would be different if it was 25 1vs1 matches rather than 1 25vs25 match. The better weapon material will shine through.

[Vattic]

Nice research but I'm concerned about the impact on the results from dwarves bashing with their shield or pommel striking and flat slapping with their sword. You could always mod out the extra sword attacks to avoid some of this.

[franti]

Nice research but I'm concerned about the impact on the results from dwarves bashing with their shield or pommel striking and flat slapping with their sword. You could always mod out the extra sword attacks to avoid some of this.

I actually thought that may be an issue too, so I would read some of the combat reports to see.
Result: insignificant. The only weapons heavy/hard enough to do damage with a shield/pommel strike were also really sharp (Steel, Iron, Bronze, Bismuth Bronze, Black Bronze, Platinum, Brass), so the difference was negligible. If anything, shield striking with one of those materials would hurt the striker more than the victim, as simply swinging with the sword is much more efficient.