Print

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

[jokermatt999]

I just wanted to say thanks for running these tests. It's nice to see an accurate, well tested breakdown of all the weaponry/armor in the new version. Personally, I think this is the most interesting topic on the forum at the moment.

[Vastin]

One other odd note about DF weapons currently is that in the real world if someone had an ultra-strong material like Adamantine that didn't have enough actual weight to carry a strike (apparently a problem in DF3.x), they would simply weight the core of the weapon with an inexpensive, dense material such as lead to overcome this deficiency.

As for the other discussion, the way that DF3.x apparently defines the impact of a strike based mainly on the weight of the weapon seems quite unrealistic. In most cases the overall kinetic impetus of a strike is going to be based on the strength or skill of the wielder more than the weight of the weapon in question. (and lets ignore jointed weapons such as flails and nunchaku for now)

One could treat these factors as being additive, with strength and skill being the dominant factors, and weapon weight being a secondary one - but treating weapon weight as a multiplier or otherwise dominant factor doesn't seem to jive with real results.

Skill does feed into impact velocity pretty substantially btw, as there are swing techniques that will allow even a fairly weak fighter to get their weapon up to brutal speed and put them in a good leverage position for follow through.

I suppose if you have someone so massively strong that they can throw a heavy maul around at the limit of their twitch muscle speed then they'd be best off with a very heavy weapon, but in reality it is hard for strong fighters to reach that speed even with the lighter weapons.

Anyway, my experience with impact factors in live combat are that (Skill and/or Strength) > Weapon mass. High weapon masses always result in much slower swings (thus little improvement in impact) and of course a greatly reduced ability to connect with the intended target.

The one place where strength is the absolutely dominant factor in impact results is in Archery. A skilled swordsman can use various techniques to impart momentum on their weapon, but for a bow the primary factor is the weight of your draw - raw muscle power (and the technique to draw it efficiently). The missile will travel faster or slower based on its mass, and there are some optimal weights vs. air resistance that you want to find, but fundamentally its all defined by the drawstring pull, not weight or dexterity or any of that.

[zagibu]

Hehe, yeah, those DEX 18 STR 8 longbowmen in certain RPGs always hit me as kinda unrealistic.

Soooo...as I've mentioned before, I'm currently simulating the spear fights, and it seems I have to come up with a new formula that displays weapon efficiency, because the current one applied to iron spear vs. silver armor results in 12624%, which is hard to compare to the <30% of hammers and maces. Too bad I suck at math. Well, I guess something additive might work, like successful blows - (deflected + glances) or something.

I'm also currently writing a bash script that will do some more interesting analysis of the produced combat logs, like exact statistics of deflected blows, glanced away blows and blows that hit through armor per bodypart. And I'd also like to extract the effect of a successful blow (bruising skin, fat, muscle, bone, tearing, severing, etc.).

[Proteus]

I agree with Vestin vonverning archery.
Another thing whcih can be impacted by draw strength of the  bow is the accuracy at long distances.

The stronger your bow, the less high you have to aim to hit targets at long distances (like 90m for example),
therefore the less curved is the trajectory of your arrow.
This and the fact that with higher draw strenght more kintetic energy is  transferred to the arrow
make the arrow also more susceptible to deviations due  to crosswinds.

(same goes btw. for crossbows, heavy crossbows [which have to be drawn by winches] should profit from similar factors vs. lighter crossbows [which on the other hand should get reloaded faster, as you can pull the string by hand])

[zagibu]

I am so dumb. My formula for calculating the hit ratio was wrong. You want to know successful over total blows, not successful over blocked...fixed it, which dropped percentages for hammers and maces even more. I've also included test data for spears as far as I have it, so you can see the binary situation of armour in this version of DF, and why I've decided to work on another script that calculates some more detailed statistics...

[puke]

is there any way to gather data on how much damage from each weapon gets through on a successfull hit, or penetrates tissues? 

I would guess that hammers are less likely to be deflected, but that spears should deliver more of their dammage when they do connect.

[o_O[WTFace]]

I think this combat system is designed for armor damage, even though that isn't in yet.  A steel armor is so good at eating damage from a steel weapon because its in perfect condition.  After taking a few swings from a mace or whatever in battle, that steel plate is going to have problems and stuff is going to be more likely go get through.  Its like a pinata, the first few people just dent the thing up and punch a little hole or two, but the thing starts loosing structural integrity and pretty soon someone just destroys it and gets at the juicy insides.  That should increase the difference between materials too, your wooden shield will work once, your iron shield might work a few times and a steel shield is going take a beating before it looses all effectiveness. 

As for the weight = damage thing, I think it might be that all weapons are assumed to be traveling at the same speed, modified by the weapon speed modifier in the raws. 

[Dwarfoloid]

One potentially interesting solution to the Adamantine weapons (and some other issues) would be adding the weight of the swinging bodypart(s) to the weight of the swung object.

This doesn't seem to currently happen, for example try to pit an ogre with a dwarf sized toothpick-mace against an unarmed one. In my observation, the unarmed one does better, presumably because he is using his own high weight bodyparts to attack. Heck, the unarmed one was doing way too well even when I made a size 20000 two handed club for ogre size creatures. He was causing heavy damage by scraching the armed one.

Rather unrelated, but does anyone else think that some of weapons have rather inane stats? Most especially, why does the axe have more contact area than swords? Shouldn't axe have lowish contact area and good penetration, and sword have big contact area and low penetration. I modded my axe to be 2000:2000 and short sword to be 4000:500, based on my arbitary assumption that ever 100 is 1 cm of real life.

As for blunt weapons, one big problem at the moment is that protection of armor is too absolute. The game mechanics should deal more damage with non-penetrating hits, even with the edge type attacks. Even if every hit is glanced off, two mail clad dwarfs armed with axes should end up at least heavily bruised after half a dozen pages of combat. Additionally, the head should be very vulnerable to various kinds of currently unrepresented damage; unconciouness via blunt trauma, internal bleeding and death by the same, and possibly fatal damage to the upper spine from pretty much any kind of hit on the head.

Excuses if I went past the topic too much.

[Vastin]

Well, Toady set himself up for a tough slog in trying to make combat resolution so physics defined. A little abstraction can go a looong way.

The big problem right now sounds like swing speed is fixed, which makes mass the dominant factor for impact calculation. Also pretty unrealistic. Swing speed is, at its simplest, a function of (strength / mass), which if you just translate back to (velocity*mass = impact) means you basically end up at (strength = impact), and lighter weapons are swinging a lot faster, meaning they should be harder to avoid, and thus better overall. Obviously not very realistic, but it is more realistic than (mass = impact)!

Now, real life calculations against armor quickly get absurdly complex. Rigidity, absorption, depth, ablation. yada yada yada. Now a lot of these values are fairly fixed in any contest between Weapon A and Armor X - but the variables that change most on a swing to swing basis are usually Thickness and Deflection Angle.

First off, no suit of armor is uniformly thick. Design and Weight considerations demand that there be gaps, overlaps, and different thicknesses of armor in various sections. Helms tended to be the thickest and are often well rounded or angled for deflection (but require gaps for sight and breathing), Breastplates tend to be fairly thick, but are often flatter and difficult to design for deflection without adding a lot of weight. Legs and arms tend to be thinner in general, but a skilled wearer may be able to quickly angle them for excellent deflection. Etc.

Assuming a warrior could repeatedly deliver a series of blows of similar strength, the results against a given suit of armor would be primarily decided by the exact location and deflection angle of a given stroke. Many strikes would bounce off harmlessly against good plate mail, causing no injury or ablation whatsoever, a few might cause minor cuts or bruising with the bulk of the blow stopped or deflected, whereas a single lucky or well placed strike might still immediately kill or dismember the wearer.

Ablation was not generally a major component of ancient armors. They could be damaged and their efficiency reduced certainly, but more often than not an injurious stroke was simply delivered through an existing gap or via a well delivered crushing/piercing blow, rather than a rain of consecutive attacks intended to gradually degrade the armor. Piercing weapons would tend to deliver slightly injurious hits every time, killing through cumulative wounds rather than meaningfully degrading the armor. Only blunt weapons would be likely to have a useful ablative effect on rigid armor. Conversely, blunt weapons would repeatedly injure when used against non-rigid armor, such as chain w/padding, but would cause very little damage to the armor itself.

My real point is that all of this would be REMARKABLY difficult to model via physics, even for so ambitious a game as DF. Abstracting the combat system to a moderate set of strike locations, angles and piercing/cutting/crushing values would produce a set of remarkably detailed and realistic looking results without the overwhelmingly difficult to resolve set of mathematics that it looks like it is setting out to solve.

But it will be interesting to see how it goes. It is technically solvable, but it could take a very long time.

On the flip side, the interface is now a good 60% more obtuse than it was before, which I didn't really think was possible. Burrows are still giving me fits. Especially that one keystroke deletion terror... ;P

[zagibu]

is there any way to gather data on how much damage from each weapon gets through on a successfull hit, or penetrates tissues? 

I would guess that hammers are less likely to be deflected, but that spears should deliver more of their dammage when they do connect.

Yeah, I'm trying to classify the wounds into light, medium and serious classes based on what the combat log says (bruising = light, tearing = medium, severing/shattering/jamming = serious), then count them and display them along the hit ratios. I'm probably going to remove the blows to the face and fingers/toes, as those are currently unprotected, so expect the hit ratios to go down a bit more.

First off, no suit of armor is uniformly thick. Design and Weight considerations demand that there be gaps, overlaps, and different thicknesses of armor in various sections. Helms tended to be the thickest and are often well rounded or angled for deflection (but require gaps for sight and breathing), Breastplates tend to be fairly thick, but are often flatter and difficult to design for deflection without adding a lot of weight. Legs and arms tend to be thinner in general, but a skilled wearer may be able to quickly angle them for excellent deflection. Etc.

Don't forget we're talking about the armor of a race of craft-gods. I think dwarven armor would be much thicker and with a lot fewer gaps than human armor. The problem is that ingame, humans seem to use the same armor, so...

[Geti]

The whole strength = impact thing is silly. Think of it in terms of tools. get a wooden hammer and one with a steel head, and see how they compare. you tire faster with a heavier hammer, but you can get more force into it. The length of the handle is also important, but I'm assuming that's uniform in DF anyway. Yes, you can make smaller hammers go faster, but you end up with less momentum at the end of the day (especially on downward swings). There is a reason we use dense metals for the heads of sledge hammers and the like, guys. I might talk about the physics of it later, but I've just got back home and don't want to think numbers for a while.

I just think fatigue needs to be taken into account more, along with making heavier weapons increase strength and toughness faster. variation of swing direction and speed would also be nice. Better armour effects would be nice too (the whole rigidity thing vastin has gone into in more eloquence than I could muster), but mostly I want cumulative wounding fixed.

[Vastin]

There's a lot of physics behind the weight, length and balance of a melee weapon, and a number of different methods to swing them, all with situational trade offs.

At the end of the day my experience is that heavier weapons are capable of higher MAXIMUM impacts, but only when the wielder has the luxury of a massive wind up. This is a rarity, and thus the standard impact level of a weapon is much more reliant on strength and skill than weight.

Most fighting styles are build around short, fast strikes that rely much more on the strength and speed of the fighter than on the weight of their weapon. Most weapons are constructed out of the lightest, strongest materials available to support this approach.

Oddly, because I am a very light weight fighter, my personal combat style does in fact involve using heavier blades with complex wind up maneuvers that allow me to strike with force - but it does make it it very dicey dealing with stronger fighters who prefer lighter weapons and very short, sharp striking styles.

I could write a long treatise on this stuff, as I do it a lot and enjoy the applied physics aspect of it all.

Anyway, for standard striking techniques, strength + skill are going to affect the impact force a lot more than weight. If you goal is to strike an immobile, massively armored target with the most force dwarvishly possible, THEN you may want that 15 pound maul, though a good rockfall trap or bottomless chasm is a much better method in my book.

[Geti]

What style of fighting do you do? I'm currently a student of aikido, but we only do so much weaponry training.

[zagibu]

Again, it would be misleading to compare real fighting techniques with those in DF. A) they are Dwarves, a fantasy race which is almost as wide as tall and drinks alcohol like water B) in Aikido, Kendo, whatever, you don't fight to kill, which obviously has a lot of influence on your striking technique. Also, in real combat training, you never fight opponents behind thick layers of steel plate and mail, maybe even with a shield in front. I don't think you can apply what you learn in Kendo to this sort of fighting.

On a side note...do goblins wear steel? I don't think so, and if I'm right, steel armor and weapons are no longer that necessary, since bronze pierces iron easily, and withstands it well. Course, if you CAN produce steel without hassle, it's still a good idea to use it at least for armor.

[Rask]

Those are very interesting results for fortress mode. Looks as if bronze or better spears would be the best choice for fighting goblins.