Print

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

[Deon]

Mace:
Blunt deflect if layer weight more than:        2408

Hammer:
Blunt deflect if layer weight more than:        1204

Doesn't it mean that maces are better vs deflection?

[scamtank]

The double mass gives it a better chance to avoid the immediately disqualifying "punching a boulder" effect, yeah.

[Urist Da Vinci]

...
Edit - The velocity is half, even though the modifier is the same? I suppose this is because the maces are twice as heavy?

What I label as "vel" there is a function of: creature size, creature strength, weapon weight, and the weapon's velocity modifier. Basically it is how fast a creature can swing a given weapon. The result is in the same scale or units as ranged weapon SHOOT_MAXVEL. It is nonlinear so that a gigantic creature with a tiny weapon swings at 100 (pre-application of the weapon's velocity modifier). Twice as heavy doesn't exactly lead to half velocity.

Mace:
Blunt deflect if layer weight more than:        2408

Hammer:
Blunt deflect if layer weight more than:        1204

Doesn't it mean that maces are better vs deflection?

This part refers to the ability of soft items to deflect off heavy (usually dense) layers. I forget if the weight here referred to the displayed weight (in kg) or the numbers used in the calculations (in g). This effect causes items with low volume and low IMPACT_YIELD to deflect off an armor layer if the volume under the contact area was over a certain weight, regardless of the attack's momentum. Generally it only matters for wood weapons on metal armor. In 0.34.11, it was responsible for wood bolts deflecting from metal armors except low-density adamantine, despite the railgun velocities.

[Art]

Is there way to save the state of the object testing arena for quicker repeated testing?

[Pirate Bob]

Thanks for pointing me to this thread.  I agree that it would be nice to make maces better, since they are the only weapon among the standard (dwarf-creatable) ones that is significantly worse than the rest. 

To reiterate some of what Urist is saying, you can give maces (nearly) identical armor/tissue penetration properties as warhammers by doubling their velocity multiplier.  But then they will smash twice the area, doing more total damage, and making them better.  So you could try some intermediate value of multiplier.  You could think of this as being due to them having a longer handle, making the head of the mace travel faster.

As others have stated, it would also be reasonable to make them attack faster.  I have personally never messed with the ATTACK_PREPARE_AND_RECOVER numbers, but I assume smaller is better.

If you are interested, I can post some more of my scripts I use for large-scale testing, or help you design something more to your needs.  You don't by any chance run Linux, as that would make things much easier?  I could still help you write a script that would set up the entire arena with one click, but using Perl in Linux I could do an automated scan over a range of velocity multipliers. 

If you want to try this, probably the best way to start is for you to tell me exactly what test you want to run, and then I can post a save to DFFD that lets you do this.  For example, I assume you want to run tests of hammerdwarves vs. macedwarves?  What armor do you want them wearing?  What skills?  How many do you want to test at a time (I have found 1152 to be the practical limit for me, but maybe with a better computer more is possible)?  I also assume you would prefer to use my cells for one-on-one battles, as this allows you to get many independent data points at once?

[Art]

I would definitely be interested in your input, though I am using windows 7. Let me tell you about my own testing so far. I have been doing some imperfect 10v10 melees.

With vanilla stats, warhammers just clearly win with every combination of skill and armor I have tried. However, doubling the velocity multiplier makes silver maces win vs silver hammers, with about a 50% survival rate in unarmored fights, at competent and skilled proficiency. Using iron breastplates and helms, warhammers start winning again at about a 30-50% survival rate, again at skilled and competent. However maces pull ahead at Grandmaster with about a 10%-30% survival rate.

It appears that doubling the velocity will make maces superior vs squishy targets, but does not actually result in equal armor penetration, as has been suggested. This is all from a very small sample, so take it with a grain of salt.


I would be interested in testing the attack recovery times, but my methods are slow.

Edit - As for what presets I would like to see, how about:

1000 1v1s

dwarves vs dwarves
silver hammers vs maces(2x velocity)
armored (iron breastplate/ helm)
naked
Competent
Expert
Grand Master

So that is 6 tests total. Can you do that? Or provide me with the tools? Will probably need more tests after.

[Stalhansch]

I, too, struggled with maces and hammers in my own mod.

In reality, hammers specialised in their versatility, often having a spike or beak on one end along side a hammer bite on the other. Maces, on the other hand, were strictly blunt weapons with no (literal) armour penetration capability. They are virtually the same weapon, except the hammer has a bladed spike attack, which leads to problems with DF making bladed attacks get used more often than blunt, and which inevitably drove me to removing the spikes altogether, further stressing the weapons' similarity.

Historically, maces did have the niche of symbolism, much like the sword, and were often represented or used ceremoniously with connotations of authority. The hammer was strictly a brutal anti-armour weapon, through and through.

For what it's worth, as a weapon enthusiast: If you want to give maces their own identity while still respecting the reality of the weapon, it should be kept in mind that maces bite/strike rather than bash, unlike the war hammer. They do pin-point damage with very small contact areas to try and send their force directly through the armour, while the hammer attempts to turn the armour itself into an extension of the weapon with force distribution over a large area.

[Chevaleresse]

So basically the exact opposite of what happens now?

[Stalhansch]

So basically the exact opposite of what happens now?

I suppose you could say that.

It is difficult to differentiate between a mace and a hammer without simply making one of them inferior to the other with the DF system. There simply aren't enough variables. I had a lot of trouble with it myself. Even after all of my work and testing I ended up with some iteration of maces which are nearly identical to hammers. Square one. I'm still working on it.

[Art]

In small scale testing my current mod seems to make maces better vs armorless foes and hammers better against armored.

Needs more testing though.

[Urist Da Vinci]

In small scale testing my current mod seems to make maces better vs armorless foes and hammers better against armored.

Needs more testing though.

If the tissue/armor layers have more "momentum resistance" (i.e. steel plate), then you want to focus more momentum on a smaller area to ensure you get through (hammer/whip/bolt/stabbing).

If the tissue/armor layers have effectively zero momentum resistance (i.e. flesh or leather), then you want a big contact area in order to maximize the volume of tissue damaged (mace, mauls, two handed sword, greataxe). Well, except if the target is megabeast-sized, where you may prefer to damage a small area to a large depth rather than do shallow damage to a large area.

[Art]

Which is what I am going for, and seems to be what the vanilla stats were going, but missed the mark.

I still need to do more testing before I decide on the final stats

[Pirate Bob]

Here is an attempt at some scripts to help you with testing.  I probably messed something up, as I'm tired and my 3-month-old daughter has been screaming most of the evening, but if we're lucky it might work:
http://dffd.wimbli.com/file.php?id=9727

If you aren't feeling adventurous, I'll have a more careful look at it tomorrow to make sure it's working, but if you don't want to wait give it a go.  I'll try to sneak some time at work tomorrow to try running the perl on a windows machine - I think it should work fine, but I'm not sure.

Anyhow, good luck, and let me know how it goes.

Oh - just remembered - perhaps most important.  To change the mace velocity you need to adjust the 2000 at the end of [ATTACK:BLUNT:20:200:bash:bashes:NO_SUB:2000] under mace in /raw/objects/item_weapon.txt to be something else.  Set it to 4000 for twice the velocity, or 3000 for 1.5x.  Increasing this velocity number will increase damage and armor penetration, but not attack speed.  If you want to change attack speed, change the next line.  I don't know exactly how those numbers work, but I assume smaller numbers means a faster attack?

[Urist Da Vinci]

Pirate Bob cancels Script: Interrupted by Baby x100
...
 If you want to change attack speed, change the next line.  I don't know exactly how those numbers work, but I assume smaller numbers means a faster attack?

[ATTACK_PREPARE_AND_RECOVER:3:3]

The first number is the number of time steps between when you decide to attack, and when the weapon hits. The second number is the number of time steps after the weapon hits in which you are off balance and more vulnerable.

If you use a weapon with [ATTACK_PREPARE_AND_RECOVER:1:1] for example, you can attack and recover before an opponent with a [ATTACK_PREPARE_AND_RECOVER:3:3] weapon has a chance to hit you. You then still have enough time to block/parry their attack and then hit them while they are recovering. Because you attack more frequently, your "DPS" is higher and you are more likely to get a useful cyan opportunity attack.

At the moment most weapons are [ATTACK_PREPARE_AND_RECOVER:3:3], except a few large weapons are [ATTACK_PREPARE_AND_RECOVER:4:4]. This is mostly to give room for quick or wild attacks to decrease the time of the attack. This is probably a good area for modders to introduce some variety.

[scamtank]

I'm thinking the default prep/rec numbers should be higher across the board. Since I assume we're dealing with integers only, there's not a lot of nuance and differences you can stuff between 1 and 3.