Print

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

[Shinziril]

I did some research on weapons versus armor.  The main conclusions:

TL;DR version: For unmodified Fortress Mode, if you want weapons that will work against any enemy, stick with (steel/adamantine) (battle axes/short swords/spears) and non-adamantine war hammers.  Bronze or iron armor is sufficient to protect your dwarves from enemy slashing weapons, but steel or adamantine provide better protection against certain other weapons.  Nothing can fully protect your soldiers from morningstars, scourges, large daggers, whips, or war hammers. 

Testing was done with completely unskilled human subjects in Arena Mode.  Note that in cases where weapons could not penetrate armor, they generally ended gaining weapon skill fairly rapidly, and it generally didn't seem to affect the (lack of) penetration. 

There are three major categories of weapon attack- broad edged attacks, narrow edged attacks, and blunt attacks.  For convenience, I will refer to these as "slashing", "piercing", and "blunt" attacks. 

Slashing weapons are almost entirely incapable of penetrating armor of their "level" or below.  Any injuries sustained are generally minor bruises (or, more likely, various damage from lucky bites and punches).  The levels go Silver<Copper<(Bronze=Iron)<Steel<Adamantine.  Thus, a silver battle axe cannot penetrate any armor, and neither can a copper battle axe.  Both bronze and iron battle axes will penetrate copper armor, but they bounce off bronze and iron armor.  Steel battle axes can penetrare copper, bronze, and iron but bounce off steel and adamantine, and adamantine battle axes can penetrate any armor but adamantine. 

Piercing weapons can penetrate armor made of better material, but the penetration ability varies with their contact area.  Of course, if the piercing weapon is made of stronger material than the armor, they are guaranteed to penetrate anyway. 

Blunt weapons can again penetrate armor.  Their penetration ability seems to vary with impact velocity, contact area, and possibly size.

Conclusions about individual weapon types:

-Edged weapons:

All edged attacks with a CONTACT_AREA of 50 or greater seem to be slashing attacks, incapable of penetrating armor of their level or higher.  Note that this includes the "stab" attacks of all swords.    Steel or adamantine slashing weapons still work just fine in Fortress Mode, as no vanilla enemy of the dwarves has access to steel armor in significant quantities.  If dwarves ever start exporting steel armor in quantity during worldgen, stick with adamantine for guaranteed-to-work weaponry. 

Spears and pikes seem to be the weapons with the largest CONTACT_AREA (20) that perform as piercing weapons.  However, their penetration ability is not necessarily that great.  Adamantine spears can pierce all armor but adamantine armor- in other words, exactly the same as adamantine battle axes.  Steel spears cannot pierce steel or higher armor.  Armor made of iron and bronze is the point where we see a noticeable difference between spears and battle axes, with silver/copper/bronze/iron spears penetrating iron armor about 5% of the time and bronze armor about 20% of the time (one of the few times we see a noticeable performance difference between bronze and iron).  Silver and copper spears can penetrate copper armor, again about 20% of the time.  Note that "penetration" in this case could mean, say, fracturing bones through the armor without cutting through it- battle axes can't even do that.  Spears also seemed to cause major injuries more of the time (compared to battle axes of the same material) if their material was stronger than the armor material, but be warned that this could be more "tearing the muscle!" hits and fewer "the severed part sails off in an arc!" hits. 

The next step is weapons with a CONTACT_AREA of 10, in this case morningstars and scourges (both of which also have an IMPACT_VELOCITY of 2000, twice normal, which doesn't seem to hurt).  Any non-adamantine morningstar or scourge can penetrate any armor (yes, that includes adamantine).  Steel and adamantine armor both reduce the incidence of major injuries slightly, but even adamantine armor lets major hits through about 20% of the time. 

Large daggers can also penetrate any armor regardless of material, but their penetration is lessened by the presence of a slashing attack in addition to their (CONTACT_AREA 5) piercing attack.  Watch out for those goblin thieves, they have a nasty habit of headshotting your soldiers right when you weren't expecting it. 

-Blunt weapons:

Weapon material is almost entirely irrelevant for blunt weapons as long as they're not adamantine.  Copper and silver work well since they're much less useful for edged weapons and armor, especially in Fortress Mode where you can't make silver morningstars. 

Whips are the best, as is now common knowledge.  Whips seem to flat-out ignore armor, with even adamantine armor providing no significant protection.  They only cause serious injuries about 33% of the time, but that's not particularly reassuring when a single serious injury can be crippling (and often has fatal follow-up). 

Warhammers are second best.  Warhammers again deal serious injuries about 33% of the time through armor, dropping to 22% against steel armor and ~15% against adamantine armor.

Weirdly, other blunt weapons are significantly worse.  Maces penetrate armor up to bronze and iron about 20% of the time, and can barely penetrate steel or adamantine armor at all.  Mauls and flails seem incapable of penetrating any armor at all. 

[MarcAFK]

interesting. seems reletively similar to the results from the table someone posted in 31 except for the whip tests,but i think the numbers have changed a bit. i havent the patience to test this stuff myself,going through combat logs determining the hit percentages,minor,moderate or major injuries etc. but i think the old table is probable still accurate enough. we can generally agree that large daggers and whips are way overpowered. if i wasn, on my kindle i would search for that table and link it here.

[o_O[WTFace]]

Interesting.  I had always assumed spears had better piercing abilities then that. 

[Shinziril]

i havent the patience to test this stuff myself,going through combat logs determining the hit percentages,minor,moderate or major injuries etc.

Actually, that was made vastly easier by a little Python script I managed to write (after teaching myself rudimentary Python).  It takes a text file as input (the "gamelog.txt" file includes everything you'd find in every combat report, although you'll need to clear it after every test) and spits out various combat numbers.  Specifically, it outputs the number of dodged hits, parried hits, shield-blocked hits, and hits that actually made contact, and then the number of hits that were deflected by armor, the number that caused minor wounds, the number that caused major edged wounds, and the number that caused major blunt wounds. 

I suppose I could upload it if anyone wants it, although it's a bit specific- among other things, at the moment it's designed to look specifically for weapon attacks and ignore the inevitable scattering of bites/punches/kicks that get thrown in.

[Carnes]

I suppose I could upload it if anyone wants it, although it's a bit specific- among other things, at the moment it's designed to look specifically for weapon attacks and ignore the inevitable scattering of bites/punches/kicks that get thrown in.

I think that sounds great!  Could you put it on git or sourceforge?  So you can retain control over it but still let others fork/modify?

[xeniorn]

It would be great if you'd upload the script. It'll be much easier for other people (and me) to modify the script to suit their needs than write the whole script themselves.

[Shinziril]

I just stuck it up on DFFD.  Feel free to do whatever you like with it, it's maybe 60 lines of text search parameters at most.  The hard part is working out exactly what search parameters will get it to detect what you want it to detect in the first place. 

Notes for its current version:

You need Python to run it, obviously.  It has to be run from the command line at the moment to work properly, since it outputs there.  It finds a text file named "gamelog" in the folder the program is located in and reads it.  It lumps all weapon attacks together without distinguishing between weapon types, and it'll read the entire gamelog, so be careful.  What I found best was to keep it in a seperate folder along with a dummy gamelog.txt file, clear the DF gamelog, then run an arena test using only the weapon and armor I'm testing at the moment.  Once that's finished, cut the text from the DF gamelog, paste it into the dummy gamelog (select-all first to clear any previous text), and run the evaluator. 

For the arena test, I modified the testing arena slightly to include a 9x5 grid of 5x5 tile cells with walls.  I had one DF macro that would move the selector to "human" once I opened the creature creation window, a set of macros that would add full armor of the various metals (helm/mail shirt/breastplate/greaves/gauntlets/boots), and a macro that would add one creature to the same location in each cell when pre-positioned (that last one I would play twice to get two people in each cell).  Macros and arena are uploaded here if you're having trouble following this. 

[FrisianDude]

Interesting. I guess the few militia I have who liked maces, and as such have been assigned maces, won't be killers but rather speedbumps. Good to know, I suppose. I also did notice the lack of effect mauls seem to have; a goblin took quite some time killing an animal with one.

[Shinziril]

Maces are . . . halfway decent?  Given that enemies generally wear bronze/iron armor, they'll be maybe 2/3 as effective as warhammers.  Try to loot some morningstars if you can, those still use the Mace skill and are crazy good. 

[FrisianDude]

Yeah sorry, for some reason I had taken 'not as good as warhammers' as meaning 'shit.'

Question though, perhaps you've noticed it; is there anything changed in how layering of armour works? Because my melee-militia seem to refuse wearing two coats of mail and a breastplate as ordered.

[Shinziril]

Hmm.  I hadn't ever given them anything besides a basic set of everything (helm/single mail shirt/breastplate/greaves/gauntlets/boots).  You could test layering in Arena Mode, but I suspect it works differently if you only apply equipment during creature creation.  Spawning an enormous amount of equipment on something easily killable and then assuming control of your test subject so you can try to manually put stuff on is probably your best bet. 

[Radiant_Phoenix]

I have an explanation for whips penetrating armor: They have a ratio of size to contact area of 100, and a velocity multiplier of 5; this gives them 1250 Urists of kinetic energy per Urist of area, and 500 Urists of momentum per Urist of area. The runner-up weapons are:

Morningstar (bash) has 100 units of momentum and energy
Scourge has 60 units of momentum and energy
Mace has 80 units of momentum and energy
Warhammer has 80 units of momentum and energy

Pikes and large daggers have 40 units of momentum, and 20 units of energy

(p = mv, KE = 0.5*mv^2)

Increased mass may, however, have a less-than-linear benefit to damage, because it becomes harder to swing the weapon, which would suggest that lighter weapons with the same ratio are better, all else being equal.

Note that this is assuming that all weapons fail to penetrate, and are thus bludgeoning.

[Carnes]

Thanks!  Downloaded and i'll play with it tomorrow.

[MarcAFK]

Awesome, i might play around with that, once i read up on wtf to do with python

[Jacko13]

Thanks Shinziril for the OP. I found this very helpful ad I was somewhat overwhelmed with all the data on the wiki for weapon and armour types especially as some of it seems outdated. Did anybody do any recent tests on obsidian? It sounds like this makes rubbish (if trade profitable) swords in current versions?  My current fort is in a terrifying biome and we only appear to have copper and silver... War hammers all round right now but I have a legendary swordsdwarf just arrived and wanted to take advantage...