Bay 12 Games Forum
Dwarf Fortress => DF Dwarf Mode Discussion => Topic started by: Mort Stroodle on January 28, 2020, 07:44:03 pm
-
I've always heard that the higher density of silver warhammers makes them better choices than lighter metals. This makes sense, but I didn't see any testing on it, except for one forum post from seven years ago. So I did a little test. (https://dwarffortresswiki.org/index.php/DF2014:Military_testing#DF2018_.280.42.12.29_tests) Two dwarves face off, one with a steel warhammer, one with a silver warhammer. 100 trials armored, 100 trails unarmored.
Unarmored test: Silver wins in 65% of matchups
Armored test: Silver wins in 68% of matchups
Those are very similar results, which both suggest that silver is meaningfully more effective (silver wins almost twice as often as steel!), and that armor doesn't seem to really change this result. This is nice to hear, as it means silver has a relevant niche in combat.
I'm interested in trying more weapons testing in the future, with much bigger sample sizes. What tests would you like to see?
-
Very interesting.
Will the heaviness of silver tire out Dorfs faster?
How about:
Steel Warhammer Vs 100 zombies
Vs
Silver Warhammer Vs 100 zombies
Since zombies are mainly what we'll be putting our hammer dwarf squads against. And they don't tend to be one on one fights.
-
I'm interested in trying more weapons testing in the future, with much bigger sample sizes. What tests would you like to see?
May as well revisit a summing up of the traditional mainstays. Requesting whatever suitable tests you can cook up involving a:
1. Silver whip.
2. Cave spider silk sock.
-
Does arena allow item wear? Can you see how long it takes the silver to become worn out vs the steel? I suppose wear doesn't matter as much for metal items, since they can be melted down to reuse the material...
-
Does arena allow item wear? Can you see how long it takes the silver to become worn out vs the steel? I suppose wear doesn't matter as much for metal items, since they can be melted down to reuse the material...
Matters if you're an Adventurer...
-
Two versions of the test...
Silver maces verse steel maces, to see if the results are close to the hammer results,
Silver maces verse steel hammers, to see how the results differ verse the others.
-
My understanding of weapon materials:
For edged weapons, hardness matters more than density.
For blunt weapons, density matters more than hardness.
Seems to hold up here, which is at least good to know updates haven't changed.
The differences are when a creature is made of a harder material, IIRC. I'd suggest tests against bronze titans and whatever other hard-material beasties are available.
-
Two versions of the test...
Silver maces verse steel maces, to see if the results are close to the hammer results,
Silver maces verse steel hammers, to see how the results differ verse the others.
I did some more testing last night, with about 315 trials (figured out how to automate better), which suggested silver hammers vs silver maces weren't meaningfully different in terms of success against armored dwarf opponents. I'll post results here later.
-
How about gold hammers? And can you make rock hammers from the heaviest rocks? I remember seeing I could make rock weapons at some point.
Edit: I tested silver vs platinum and platinum won the 2 fights I did seemingly handily.
-
You can only make artifact weapons out of gold or platinum, so neither is available for squads in practice in a fortress. There is the "rock" sword which consists of a wooden piece with obsidian shards sticking out of it, but only obsidian is eligible for those weapons.
-
Bronze pickaxe vs iron pickaxe wouldn't be a bad test. Just for those really early fights, so you know which one is better to bring along for your initial tools.
-
Very interesting.
Will the heaviness of silver tire out Dorfs faster?
How about:
Steel Warhammer Vs 100 zombies
Vs
Silver Warhammer Vs 100 zombies
Since zombies are mainly what we'll be putting our hammer dwarf squads against. And they don't tend to be one on one fights.
I'm pretty sure that tiredness is only increased if they start going above their strength limit. Which early on if you're giving your new recruits heavier armor than they can handle, will be an issue, but once they're trained or if they're kept lightly equipped it shouldn't matter.
-
I recall reading previously that a silver warhammer could outdo steel axes, but don't recall seeing actual science behind it. This thread has gotten me interested in testing that out again. I also vaguely recall seeing a test where someone tested axes vs swords and came to the conclusion that battle axes were better.
-
stuff
I would just like to say: tyvm for this science.
I would request tests on the relevance of body size in contests between armed, highly-skilled sentient opponents.
-
I recall reading previously that a silver warhammer could outdo steel axes, but don't recall seeing actual science behind it. This thread has gotten me interested in testing that out again. I also vaguely recall seeing a test where someone tested axes vs swords and came to the conclusion that battle axes were better.
Silver hammers do great vs armor.
Steel axes can quickly end lightly armored combatants
I think axes penetrate better than swords on slashes, but long edges of swords do well vs big creatures with limbs.
Is what I've seen and read. I swear by silver hammers myself, because they're good and plentiful.
-
I somehow always end up with silver warhammers in majority of military because I often find some ore of silver and can craft lots of warhammers for better qualities and use the other ores for armor.
Only later craft some complementary slash weapons when the armor is done.
Axes and swords seem to be good slashing weapons but what about spears how do they perform in comparison?
-
Never got around the the mace vs hammer shenanigans I was planning on posting. This is still for 0.44.12.
Setup:
1 Armored Hammerdwarf and 1 armored Macedwarf were placed in opposite ends of a 5x5 room. 105 5x5 rooms were created, and macros were used to populate them. The game was saved to allow for loading the scene again to do three total replications. Scores were logged by checking the unit list for the total number of survivors for hammerdwarves and macedwarves. Percentages were calculated based on s/105, where s = number of survivors.
Skills and Equipment:
Skilled Macedwarf, Shield user, Armor user, Fighter, Dodger
Silver Mace, Steel Armor, Wooden Shield
Skilled Hammerdwarf, Shield user, Armor user, Fighter, Dodger
Silver Hammer, Steel Armor, Wooden Shield
Trials per replication: 105
# of Replications: 3
Results
Scores are tallied in terms of surviving dwarves.
Replication 1
Hammer: 53 (50.5%)
Mace: 49 (46.7%)
No survivor: 3 (2.9%)
Replication 2
Hammer: 52 (47.6)
Mace: 52 (47.6)
No survivor: 1 (1.0%)
Replication 3
Hammer: 50 (47.6%)
Mace: 55 (52.4%)
No survivor: 0 (0%)
In total, 155 hammerdwarves suvived out of 315 trials (49.2%). 156 macedwarves survived (49.5%).
Findings:
No significant difference was observed between the survival rate of armored hammerdwarves and macedwarves wielding silver weapons when fighting each other. Both weapons seem equally effective against armored dwarves.
These findings seems to contradict the oft-held notion that warhammers are better weapons than maces due to the hammer's lower contact area. If other factors make one weapon more effective than the other in other contexts (less-ideal materials, enemy size, unarmored opponents, danger of falling, etc), more testing would be required to determine this.
-
All the old tests were literally done under different laws of physics than now, so a different result in new tests isn't too surprising.
-
My question now is: do maces still pulp stuff better than hammers? If so, I'd say maces have become the new king of blunt-force weaponry.
-
I am surprised that maces and hammers matched up so well. I expected maces to do a bit better, just due to the ability to deliver more mass at a higher momentum, which would result in a higher effect on the opponent over time.
Maybe the mace verse hammer test is running into a situation where "the first good hit puts the opponent into crippling pain, so the amount of applied force becomes secondary".
-
instead of pitting a hammerdwarf and a macedwarf against each other and seeing who wins, it might be informative to pit each against a variety of other opponents?
e.g. against a single zombie, is a hammer better, or a mace (or are they about the same)? what about against half a dozen zombies? a cyclops? a titan? a dragon?
-
I am surprised that maces and hammers matched up so well. I expected maces to do a bit better, just due to the ability to deliver more mass at a higher momentum, which would result in a higher effect on the opponent over time.
Maybe the mace verse hammer test is running into a situation where "the first good hit puts the opponent into crippling pain, so the amount of applied force becomes secondary".
I suspect this is true too, is there any way of dumping the logs for all these? It might be possible to look at, say, average hits taken before dying? The science here is awesome, but I'm not quite sure it models the normal "200 goblins vs 10 dwarves" scenario that players see in fortress mode
-
Bronze pickaxe vs iron pickaxe wouldn't be a bad test. Just for those really early fights, so you know which one is better to bring along for your initial tools.
Pick Axes are one of if not the most powerful native dwarf weapon on average or at least they were last time I checked. Back in the day I did a load of weapons testing and I remember some of the standouts being pickaxes and morning stars. It also really comes down to what you are fighting, but a legendary miner with a steel pick is one of the worst dwarves you can ever have go berserk. You can easily lose even highly skilled military dwarves in full gear to them.
Iron pickaxes are better than bronze if I remember correctly, but the difference is slight. Pickaxes are odd in that they use mining skill instead of weapon skill for combat calculations and although they are edged weapons their tiny contact area and high force multiplier make even picks made of soft metals somewhat effective against armored foes. Generally slashing weapons need to be made of harder metals and bludgeon weapons do better with denser ones. This still holds true with pickaxes, but its less of an issue due to their small contact area and heavy impact.
I don't have any old testing logs around anymore but there were a lot of interesting results. For instance I remember testing a bunch of equally highly skilled dwarves in full addy with battle axes against an equal number of naked dwarves with bronze or copper morning stars and the naked dwarves would win.
What prompted my testing originally was I witnessed a berserk legendary miner with a steel pick and civilian clothes kill like a dozen other dwarves including several of my fully geared military stand outs. That along with some other quirky observations such as just how lethal whip and morning star wielding goblins seemed to be compared to all the other gob trash made me decide to do a load of controlled tests in the arena.
Some general observations from back then were small contact areas on bludgeons and stabbing weapons are paramount. The above example in this thread really illustrates this well where we see the hammer actually performing similarly the mace which has almost twice the mass and should otherwise beat it hands down.
On slashing weapons it depends on the foe but generally your slashing weapons are there to remove heads and limbs so you want a large enough surface area and penetration to be able to cut through the targeted body parts. Against armored foes slashing weapons with large surface areas must absolutely be made with a harder metal than the opponent's armor.
Weapon weight and velocity multiplier are important stats that should not be ignored and penetration is extremely important on stabbing weapons vs large foes. A stabbing weapon is garbage if it can't reach deep enough to hit internal organs. Similarly stabbing weapons are trash vs anything that either lacks or doesn't need its internal organs like undead. The one exception to this rule is morning stars which are technically edged stabbing weapons with terrible penetration, but in practice most of the morning star's damage is done by the impact force like a blunt weapon. Although I haven't tested anything recently, what made morning stars lethal to dwarves and other armored foes was they are essentially heavier war hammers that do additional slashing damage so they cause all the same sorts of blunt trauma injuries as the hammer but they also open up arteries and sever nerves on top of it.
-
It's also likely worth doing similar tests with various armor materials.
-
In all of my experience, version-agnostic, maces have been pointless when compared to warhammers. They break bones decently, but if you need to actually punch through armour? Forget it. Especially if your enemy doesn't HAVE bones to break, or a pain threshold.
-
Maces should be able to deliver twice the momentum to the target. That should matter, even if the contact area is twice as large.
-
Judging by the formula, it would not be a flat out 2x difference. More like:
Mace_momentum = Hammer_momentum * 2 * (Hammer_weight + Individual_creature_size) / (2 * Hammer_weight + Individual_creature_size)
...so it will always be below 2x, although nearing it if the user is large enough. Note that Hammer_weight =Material_density * Hammer_size, which is a value of 4196000 for a silver war hammer. For an average dwarf (size 60k) this means there's almost zero difference in momentum between a silver war hammer and a silver mace - a noticeable difference only starts to appear for very big creatures. It appears earlier if using iron instead of silver, but nowhere near dwarf or human size.
This seems to confirm my testing observations that the SIZE value of a weapon matters very little for the general player usage.
Edit: That said, my tests also confirm the superiority of silver over steel in blunt combat, which implies there's likely something going on that is not reflected in the formula, but it has to do with density rather than weight.
-
Would it be possible to test the effect of a dwarf's strength on the weapon's performance? Say, weak-competent macedwarves vs strong-competent macedwarves, weak-competent hammerdwarves vs strong-competent hammerdwarves etc.
Anectdotal as it is, my experience with maces is that unless the user has some really significant muscles their performance is toward the poorer side, requiring more hits on the same target than hammers, before inflicting an incapacitating wound.
-
Judging by the formula, it would not be a flat out 2x difference. More like:
Mace_momentum = Hammer_momentum * 2 * (Hammer_weight + Individual_creature_size) / (2 * Hammer_weight + Individual_creature_size)
...so it will always be below 2x, although nearing it if the user is large enough. Note that Hammer_weight =Material_density * Hammer_size, which is a value of 4196000 for a silver war hammer. For an average dwarf (size 60k) this means there's almost zero difference in momentum between a silver war hammer and a silver mace - a noticeable difference only starts to appear for very big creatures. It appears earlier if using iron instead of silver, but nowhere near dwarf or human size.
This seems to confirm my testing observations that the SIZE value of a weapon matters very little for the general player usage.
Edit: That said, my tests also confirm the superiority of silver over steel in blunt combat, which implies there's likely something going on that is not reflected in the formula, but it has to do with density rather than weight.
Okay. I followed that formula, and went through the information in this (http://dwarffortresswiki.org/index.php/DF2014:Material_science#Attack_Momentum) page.
I do not agree with the wiki's statement "Since momentum = velocity * mass, and lighter items can be swung faster, attack momentum is largely independent from weapon weight." That seems to be the equivalent of stating that acceleration will remain constant during the attack and that the weapon can achieve greater velocity than the physical motion that is propelling it.
When you are using tools against a hard target (axes and shovels for spitting wood and for breaking ground), the heavy tools are always preferred, because it is much better to imbue a heavy tool with velocity than to put your body behind the mass of the tool. The second option is painful, damaging, and exhausting, because of the shock that goes into your hands and arms. There is no alternative... you either imbue the mass of the tool with velocity, or you hurt yourself.
I am not attempting to make any conclusive statement, just to point out what seems to be missing in the formula and the wiki information and why I disagree with it.
-
This seems to confirm my testing observations that the SIZE value of a weapon matters very little for the general player usage.
Edit: That said, my tests also confirm the superiority of silver over steel in blunt combat, which implies there's likely something going on that is not reflected in the formula, but it has to do with density rather than weight.
this is confusing, and contradictory, and i'm not sure if you were trying to express something different than what i'm reading
cause "density" is simply "weight per size". so if the formula is not quite right, and maybe density is the correct term, then we're back around to size mattering again...
-
If talking real-life physics, sure, both of you are correct, but we're talking about DF abstractions here. And yes, the formula on the wiki might be incorrect - I could never find the source for it.
Regarding the weight question specifically, I was talking about how there is little difference in performance of two absolutely equivalent weapons of the same material, but of different SIZE values.
On the second note, it is possible that density doesn't matter either, as is originally implied by the formula. There were some tests earlier, which suggested a much lower deviation between silver and steel hammers, and with the sheer amount of factors to take into account (random individual creature size that impacts charges and wrestling, semi-random target bodyparts from a huge selection, skill gain during the fight, effective and ineffective usage of wrestling moves, punches, kicks and bites, successful and unsuccessful stuns, snowball effect etc.), on such a small sample size the difference might be well within the margin of error.
And then it is possible that the formula is correct, but actually deals down the zeroes on some of the values from the raws (be it density, or weapon size, or creature size) behind the scenes, so they can have a greater or lesser impact than what is seen.
Edit 1:
To clarify further on the weapon SIZE question, what I meant is that it doesn't have much impact within the values used for vanilla-ish weapons, as most of them are beyond a threshold, so to speak, where weight is already too big to matter. Assuming the wiki formula is correct, if you mod in two weapons with extremely low SIZE values (say, 10 and 20), it should amplify the difference in momentum between them.
So, for vanilla silver mace and hammer comparision it goes like this:
2 * (10490[silver density] * 400[vanilla war hammer size] + 60000[average dwarf size]) / (10490 * 800[vanilla mace size] + 60000) = 2 * 4256000 / 8452000 = 2 * 0,503~ = 1,007~
Now compare the same, but assume the hammer has [SIZE:10] and mace has [SIZE:20].
2 * (10490 * 10 + 60000) / (10490 * 20 + 60000) = 2 * 164900 / 269800 = 2 * 0,611~ = 1,222~
This is basically going from a 1% mometum gain from 200 volume-points to a 22% gain from just 10 volume-points.
...Now that it is layed out like that, one thing surely seems odd - for the weapon, the formula takes both volume and density into account, but for the user creature it only counts the volume :-\
Edit 2:
Assuming the momentum table on the wiki is correct, the numbers suggest that raw density value is divided roughly by 10^5 when used in the formula. If so, then the same calculation would go like this:
2 * (10490 * 400 / 10^5 + 60000) / (10490 * 800 / 10^5 + 60000) = 2 * 60041,96/60083,92 = 2 * 0,999~ = 2~
In that case anewaname is right.
Edit 3:
Out of interest, I've run a test with maces and hammers both modded to have the same contact area of 12. Their velocity (2k for both) and size value (800 for maces, 400 for hammers) remained the same.
1v1 fights, x14 per test (individual cells)
Skilled Mace/Hammer, Proficient Armor User
full steel armor (helm, mail shirt, breastplace, greaves, 2 gauntlets, 2 low boots), 2x steel mace/war hammer (for each hand to reduce the frequency of wrestling)
Results over 10 tests:
50 mace victories
85 hammer victories
For a total of 62-63% hammer advantage
This would lead to assume that having a lighter weapon is somehow actually beneficial in armored blunt combat, until you look at the creature descriptions to highlight the problem with this method:
7 Hammerdwarves are larger than average
2 Macedwarves are larger than average
4 Macedwarves are smaller than average
To conclude, in lab conditions where combatants and weapon properties are exactly the same, higher weapon weight may or may not give an edge to one weapon, but, regardless, the difference is rendered inconsequential by the "heavier" factors.