Print

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

[Zivilin]

Historical note: Bows, cross and otherwise, were extremely deadly back before guns let you kill the bowman first.

Point taken. I am merely observing that the suppressive ability of ranged weapons necessitates conservative military tactics. That is, if You value the lives of Your dwarfs, or the time spent on their training.

Then again, there is a... tragic (or is it epic?) aesthetic in a decimated squad reaching the enemy archer division and mowing it down. Worthy of a few engravings, at the very least.

[Pirate Bob]

Historical note: Bows, cross and otherwise, were extremely deadly back before guns let you kill the bowman first.

True, but do you really think it's realistic that steel, or even candy, plate armor does literally nothing against crossbow bolts? 

The internetz and modern testing are somewhat mixed on how effective medieval armor was against arrows, but from what I have read it seems that plate armor would probably deflect arrows sometimes, but not direct hits at close range, and the protection from chainmail depends on the type of mail (riveted vs. non-riveted, thickness of weave) and the type of arrow used (broadhead vs. bodkin), but was generally worse than plate armor.

In the current version of Dwarf Fortess:
1) Plate armor does nothing at all.  It NEVER deflects bolts, and it does nothing to improve survival or reduce damage in any way (except against addy, wood, and bone bolts, which probably should be useless against armor...).
2) Chain armor seems to convert a large fraction (or maybe all? needs more testing...) of hits to blunt damage.  This does have a significant impact on survival.
3) Range has no impact on effectiveness.  Lines of marksdwarves 20 tiles away kill dwarves with the same number of hits as when they are 2 tiles away.  They of course hit less often, but that's not what we are talking about here.  More testing is required to confirm that range has no effect, as in my tests the range was not well constrained (the targets could move and the shooters could shoot at large angles).

I DO think that there is a serious problem with the way ranged weapons work right now, which needs to be addressed.  I am currently trying to see if modifying the raws can produce a more balanced result.  I am hoping to find a solution where armor does something without making dwarves invulnerable, while unarmored opponents are killed just as quickly, but I'm not sure if this is possible without changing the underlying game mechanics.  So far I have just been adjusting the force of crossbows.  Setting Force=42 (chosen totally at random  ) approximately doubles the amount of hits required to kill steel armored dwarves, and a bit more than triples it for candy, while doing nothing for lesser/no armor.  Decreasing the force further to 26 renders candy/steel armored dwarves nearly invincible, and provides some protection for iron armored dwarves.  I will add the full tables to this post later, but I have to go to work now.

Bottom line - I wasn't able to get exactly what I wanted.  I'd like all armors provide some degree of protection, with the degree depending on the armor and bolt materials.  I definitely agree that no armor, other than maybe candy, should provide complete protection.  I will try adjusting bolt mass, contact area, and maybe others (suggestions?) to see what happens.

Also, to be clear, I'm not saying what I want is necessarily the "correct" way that armor should work.  I'd like to figure out if it is possible to tune the raws to get a situation where all armors do something (but don't convey invulnerability), because I suspect that from there it will be relatively easy to tune the effectiveness up and down with small tweaks of the raws.  This will allow wiser individuals than I to create balanced and realistic projectiles, which in my humble opinion are currently neither

[Edit]
One other very important, !!SCIENCE!!, reason for finding raws where bolts are less than 100% effective - this will then allow us to much more easily determine what factors impact effectiveness.  For example, does armor user/marksdwarf level matter?  It's impossible to say if armor never does anything, but if it's say 50% effective, then it's easy to see small changes.

[PaleBlueHammer]

Hm.  Seeing as bolts are just small, flying spears, I wonder if these results apply to spears as well?  Obviously without the benefit of being ranged, of course.

[Pirate Bob]

Hm.  Seeing as bolts are just small, flying spears, I wonder if these results apply to spears as well?  Obviously without the benefit of being ranged, of course.

I do hope to eventually test other weapons in the area for comparison, and spears are a great choice to start with for the reasons you state.  I do not think that spears will prove as deadly as bolts, because the speed of a spear attack is much slower, and the the contact area is significantly larger (at least in the DF raws). 

One problem I have is how to have the attackers not be damaged by the targets, as injured dwarves likely won't attack as well and could skew results.  Maybe have attackers in fully candy, and unarmed defenders?

[MaximumZero]

That would probably work.

Also, RE:Armor:

They were making bulletproof cuirasses in the 16-17th centuries. It was closer to scale than plate, though.

[Blakmane]

Any interest in looking at shield effectiveness vs bolts? Certainly in adventure mode shields give the impression of extending lifespan... but how significant is the effect in reality? As far as I know all shield materials have the same defensive properties (although perhaps this is a myth!) so you would only have to run an iron bolts VS wooden shields cohort.

Actually, thinking about it, I wonder if metal shields would actually be worse than wooden shields, assuming all else is truly equal, because they weigh more and thus reduce speed. Might be interesting.

[Zivilin]

There is quite a lot of interest in bolt vs shield study. I have heard it said that shields are a zero-or-one protection: they either completely block the attack or completely let it through. If that were the case, then blocking would be a function of shielding skill alone, or possibly the defender's shielding vs the attackers marksdwarfship. However, I know of no "official" experiments were shields of different material were tested, so the issue is not certain. So there might be some very interesting results there.

In my previous study (in which I was still experimenting with performance criteria) shields proved to significantly increase survivability, in particular when combined with armor. They have a synergistic effect - armor helps dwarfs survive longer, giving them more chances to block with shields. But that study was iron vs iron only, and only had 100 samples.

Edit: fixed link

[Pirate Bob]

Any interest in looking at shield effectiveness vs bolts? Certainly in adventure mode shields give the impression of extending lifespan... but how significant is the effect in reality? As far as I know all shield materials have the same defensive properties (although perhaps this is a myth!) so you would only have to run an iron bolts VS wooden shields cohort.

Actually, thinking about it, I wonder if metal shields would actually be worse than wooden shields, assuming all else is truly equal, because they weigh more and thus reduce speed. Might be interesting.

I would like to eventually run tests of various shield vs. bolt materials, unless someone else does it .  It is my belief that you are correct and shield material has no impact on blocking so the lightest material is the best choice.  To confirm this, I would see if the number of *unblocked* hits results in the same rate of death, unconciousness, serious wounds etc as for naked dwarves.  If there is a difference, then either shields would have to act as armor on top of blocking, or blocked shots would have to sometimes do damage (which I don't think is true, as it's never been reported).

Along these lines, does anyone with modding experience know how valid Zivilin's "fake bone" is, and if it would be OK for me to use similar "fake wood" and "fake leather"?  It is easiest for me to run my scripts if I can treat all materials as inorganic "metals", such that if I swap out raw files the new materials will show up in the same spot within the menus and I can select them using the same macro.

[Joben]

*snip*
Regarding the bolt momentum/energy, as I indicated above, I estimate that DF bolts weight about 500 grams by comparing the weights of weapons/armors with their real world counterparts, however the weight wiki page suggests that weight in Urist is equal to weight in kg, and I have confirmed that steel bolts weight 1.17 urists (117 for a stack of 100).  Does anyone know if it is true that urists=kg?  In either case this is VERY heavy for a bolt/arrow.  *snip* Anyway, I am curious if anyone has any idea what a velocity of 1000 means in DF? 

I think it's safe to assume that urists are kilograms. Density for materials in the raws are given values that match with their real world density in grams/cubic centimeter. If you do the math for a given SIZE and MATERIAL you end up with a Urist count that is the same as a real world chunk of that material in kilograms. Steel bolts weigh 1.17 urists because they are SIZE 150 and IRL 150 cubic centimeters of steel weights ~1.17 kilograms.

The units used in SHOOT_FORCE and MAX_VEL appear to be unknown. I've asked about this a few times on the forum and never received an answer. I wonder if Toady would answer if we emailed him.

btw, have you seen the settings I use for projectile weapons? link is in the sig.

[Pirate Bob]

I think it's safe to assume that urists are kilograms. Density for materials in the raws are given values that match with their real world density in grams/cubic centimeter. If you do the math for a given SIZE and MATERIAL you end up with a Urist count that is the same as a real world chunk of that material in kilograms. Steel bolts weigh 1.17 urists because they are SIZE 150 and IRL 150 cubic centimeters of steel weights ~1.17 kilograms.

The units used in SHOOT_FORCE and MAX_VEL appear to be unknown. I've asked about this a few times on the forum and never received an answer. I wonder if Toady would answer if we emailed him.

btw, have you seen the settings I use for projectile weapons? link is in the sig.

Thanks very much for the info about units!

So, I started this project round about Aug 15, and at that time did a pretty exhaustive forum search to see what information already existed (which was none).  It seems that at least two other people have been working on the same thing simultaneously since then, lol. 

I will make sure to read your results carefully this evening, although a quick glance suggests you are using similar numbers to where my testing has been taking me.  I definitely like your use of real-world data!  Would you like me to run full sets of tests using your bow/ammo parameters?

I see you never adjusted the bolt mass?  1.17 kg is clearly way too much for a steel bolt/arrow, and lowering the bolt mass would both allow adventurers to carry a more reasonable number (currently carrying more that one stack gets really heavy) and make them penetrate armor less.  I suspect that armor penetration is based on momentum (which is what it should physically be, as a deflection is a largely elastic collision), so if we reduce the mass and keep the same energy, it should decrease armor penetration without decreasing damage to unarmored targets, which should mostly depend on energy.

[GreatWyrmGold]

Historical note: Bows, cross and otherwise, were extremely deadly back before guns let you kill the bowman first.

True, but do you really think it's realistic that steel, or even candy, plate armor does literally nothing against crossbow bolts? 

The internetz and modern testing are somewhat mixed on how effective medieval armor was against arrows, but from what I have read it seems that plate armor would probably deflect arrows sometimes, but not direct hits at close range, and the protection from chainmail depends on the type of mail (riveted vs. non-riveted, thickness of weave) and the type of arrow used (broadhead vs. bodkin), but was generally worse than plate armor.

In the current version of Dwarf Fortess:
1) Plate armor does nothing at all.  It NEVER deflects bolts, and it does nothing to improve survival or reduce damage in any way (except against addy, wood, and bone bolts, which probably should be useless against armor...).
2) Chain armor seems to convert a large fraction (or maybe all? needs more testing...) of hits to blunt damage.  This does have a significant impact on survival.
3) Range has no impact on effectiveness.  Lines of marksdwarves 20 tiles away kill dwarves with the same number of hits as when they are 2 tiles away.  They of course hit less often, but that's not what we are talking about here.  More testing is required to confirm that range has no effect, as in my tests the range was not well constrained (the targets could move and the shooters could shoot at large angles).

I DO think that there is a serious problem with the way ranged weapons work right now, which needs to be addressed.  I am currently trying to see if modifying the raws can produce a more balanced result.  I am hoping to find a solution where armor does something without making dwarves invulnerable, while unarmored opponents are killed just as quickly, but I'm not sure if this is possible without changing the underlying game mechanics.  So far I have just been adjusting the force of crossbows.  Setting Force=42 (chosen totally at random  ) approximately doubles the amount of hits required to kill steel armored dwarves, and a bit more than triples it for candy, while doing nothing for lesser/no armor.  Decreasing the force further to 26 renders candy/steel armored dwarves nearly invincible, and provides some protection for iron armored dwarves.  I will add the full tables to this post later, but I have to go to work now.

Bottom line - I wasn't able to get exactly what I wanted.  I'd like all armors provide some degree of protection, with the degree depending on the armor and bolt materials.  I definitely agree that no armor, other than maybe candy, should provide complete protection.  I will try adjusting bolt mass, contact area, and maybe others (suggestions?) to see what happens.

Also, to be clear, I'm not saying what I want is necessarily the "correct" way that armor should work.  I'd like to figure out if it is possible to tune the raws to get a situation where all armors do something (but don't convey invulnerability), because I suspect that from there it will be relatively easy to tune the effectiveness up and down with small tweaks of the raws.  This will allow wiser individuals than I to create balanced and realistic projectiles, which in my humble opinion are currently neither

[Edit]
One other very important, !!SCIENCE!!, reason for finding raws where bolts are less than 100% effective - this will then allow us to much more easily determine what factors impact effectiveness.  For example, does armor user/marksdwarf level matter?  It's impossible to say if armor never does anything, but if it's say 50% effective, then it's easy to see small changes.

If I recall correctly, arrows could allegedly pierce plate armor if fired correctly, ie using terrain to your best advantage. Until bolt heads and better physics get added, I intend to imagine that overpowering projectiles is a way to counter the inability of the AI and engine to use arcs to maximize projectile deadliness.

[Qwernt]

Great !!Science!! (even with the lack of Magma).

On the historical nature of xbows - war science is a series of creating a weapon to be the current armor and updating armor to beat the current weapon.  xbows were designed with two main purposes: low skill requirement (vs longbow) and high power armor piercing point.  Basically, allow a bunch of untrained peasants decimate the charging knights.  And they worked beautifully for that - as long as they had the farthest range.  Problems arose when they went up against profession archers and horse archers (shorter range, but much fast reload and movement).

If you think about later armor, the bulk of the protection was from deflection, not from stopping power. IE, transfer the energy in a new direction using angles rather than diffusing energy on impact.  This approach worked weel up until the musket and early rifles projected enough energy that the metal couldn't disperse.  So we moved from deflection to flack jackets (with a "there is nothing that works anyway" period in between - IE US revolutionary war through WW1).

Very long way to say: Great Stuff and fairly realistic (except for the whole softness of the bolt material show allow less impactful energy transfer... of course we do still talk about silver bullets, don't we).

[misko27]

If you think about later armor, the bulk of the protection was from deflection, not from stopping power. IE, transfer the energy in a new direction using angles rather than diffusing energy on impact.  This approach worked weel up until the musket and early rifles projected enough energy that the metal couldn't disperse.  So we moved from deflection to flack jackets (with a "there is nothing that works anyway" period in between - IE US revolutionary war through WW1).

note, A bullet proof vest was developed before WWI by a priest who wanted to defeat bullets and used spider-silk. I konow because arch-duke ferdinand was wearing one when he died. He was shot in the neck.

[Miuramir]

I think it's safe to assume that urists are kilograms. ...

...
I see you never adjusted the bolt mass?  1.17 kg is clearly way too much for a steel bolt/arrow...

Based on some quick research into historical crossbow bolts, I'm coming up with a plausible range of volumes between 4cc and 28cc.  The lower end of that range involves *very* old (Chinese, 5th Century BC) solid bronze bolts that due to corrosion are probably thinner than they started; and the upper end of that range seems only to be used for wooden bolts. 

I'd probably give wooden bolts a SIZE of 25, bolts made from a cutting-edge-capable metal a SIZE of 15, and perhaps softer metals an intermediate SIZE of 20 (this last is more handwavy as not backed much by historical evidence).  Calculate the MASS from the MATERIAL; assuming steel at around 7.8 g/cc this would give you a steel bolt at around 117g (0.12 kg, or Urists). 

Once you've got the size & mass more correct, it should be much easier to get desired results by adjusting other variables.  In particular, in an ideal world this would hopefully generate less bruising despite armor, yet still do serious puncture damage to lightly or unarmored targets. 

Realistically, you'd also want to adjust the contact area / sharpness based on the material (MAX_EDGE ?) as well; all-wooden bolts in reality are mostly designed to kill small critters by blunt trauma, and obviously even with a fire-hardened point can't be nearly as sharp as a steel tip.  I'm not clear on whether this is already taken into account elsewhere. 

[Oaktree]

Hm.  Seeing as bolts are just small, flying spears, I wonder if these results apply to spears as well?  Obviously without the benefit of being ranged, of course.

Load a bunch of spears into a mine cart railgun and they become a ranged weapon...