Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  
Pages: 1 2 3 [4] 5

Author Topic: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*  (Read 18731 times)

PatrikLundell

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.03 combat [Long]
« Reply #45 on: January 24, 2017, 12:56:47 pm »

True, but untrained/poorly trained civilians/part time militia are helped more by something that doesn't slow them down than better protection. However, if the shield is shattered in two strikes it won't provide much help. Professionals going up against tough targets probably won't achieve much with a shield bash regardless of the weight (within reason, and I don't think DF can make slade shields even with a mood).
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: From Iron to Steel: A survey of v0.43.03 combat [Long]
« Reply #46 on: January 24, 2017, 07:13:11 pm »

Even if it could, shield weight isn't affected by armor user skill, so you'd be slowed down.

Also, shield isn't shattered in two strikes. I think you can get about 8 full pages of bashing things with wooden shield before it breaks. Since one is pretty light, I suppose you could set up uniforms as weapon, shield, shield and automate always keeping X wooden shields in stock.

fragfish

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #47 on: January 25, 2017, 06:24:38 pm »

So, finally, after spending the entire evening recompiling numbers and rerunning a set of falsely calibrated combats, the update is there. Wear mechanics and armour removal make very interesting additions to the gameplay.

As for the shield question: It depends on what you're fighting. Against armoured foes, your dwarves might well splinter their wooden shields quickly.

A weird part of the update, this time steel hammers landed a bit below iron ones (last time they were a bit above) in efficiency (they still won more combats). It's probably down to deviation, but as before steel hammers show no signifcant advantage over iron ones. Time to test the silver ones : )
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #48 on: January 25, 2017, 10:50:30 pm »

On the updated 43.05 results:
Interesting. Looking at the undead tables, you mentioned hammers and axes as winners, but spears ultimately outperformed hammers, without suffering from unkillable reanimated part issue or a body part for every zombie rising up. That this happens with both steel and iron ones suggests that spears improve with skill there, unlike iron axes which I expect to become effectively sharper with skill (I think).

Though they managed to actually kill several times slower than axes, so....They did actually put down some corpses for good, right?

Meanwhile, I guess inorganic really hands the advantage to steel axes.

Also, goblin side, mace performed worse in hits/kill, dropping by 0,8 to 18,8, while whip improved from 22 hits/kill to 16,6 hits kill. The first is well within the range of standard deviation, granted, but the second one is not. However; dwarven survival against goblin macemen dropped between versions and is all-around worse versus macemen than lashers in 43.05, with iron spears in particular doing lot worse due version upgrade.

It is odd that this doesn't  reflect in hits/kill, even if it does in killed dwarves (where maces perform best).

Also, interesting find about inorganic enemy armor removal. Guess that's what Toady meant by "non-sentients shouldn't be able to remove armor, but we will see".

As for wear, it seems like the standard adamantine mail shirt would almost never get any wear, going by that table, while rigid pieces got wear against everything, suggesting that armorsmithing mood should perhaps create steel breastplate rather than adamantine mail shirt, depending on the amount of available adamantine.

Whip causing wear despite ignoring armor is interesting as well, as is the right gauntlet focus (for artifacts, at least).

PS: I hoped you saved the 43.03 post as well. I recalled some the differences from memory, but it's important for both comparison and LNP purposes.

fragfish

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #49 on: January 26, 2017, 03:59:22 am »

Looking at the undead tables, you mentioned hammers and axes as winners, but spears ultimately outperformed hammers, without suffering from unkillable reanimated part issue or a body part for every zombie rising up. That this happens with both steel and iron ones suggests that spears improve with skill there, unlike iron axes which I expect to become effectively sharper with skill (I think).

Though they managed to actually kill several times slower than axes, so....They did actually put down some corpses for good, right?

Actually, Spears killed more Zombies, but looking at reanimated Zombie/Part kills, those tended to simply get up again. Maybe they killed one or two for good, but Hammers have the advantage that stuff they killed stayed dead the first time.

As for the unkillable parts, that occured for spears in v0.43.03 as well. I think it happens when the brain is destroyed, but the head is not, so it can still be reanimated. Since Axes/Spears tend to never pulp the head completely, they can't kill it any more afterwards. In v0.43.05, the undead have become a lot more deadly, meaning they don't have to rely on unkillable heads to overwhelm the spears (and, often, not even for steel axes).

I would like to try a setup with reanimaton that can't be stopped by chopping up a necromancer, but far less initial zombies. That should show the strength of hammers in simply putting stuff down once and being done. However, I don't know how to make the arena reanimating.

Also, goblin side, mace performed worse in hits/kill, dropping by 0,8 to 18,8, while whip improved from 22 hits/kill to 16,6 hits kill. The first is well within the range of standard deviation, granted, but the second one is not. However; dwarven survival against goblin macemen dropped between versions and is all-around worse versus macemen than lashers in 43.05, with iron spears in particular doing lot worse due version upgrade.

Whips tended to need a lot of hits previously, as they would usually only chip bones unless they hit the neck. They had the highest "Dwarf suffocated" rate of all weapons. I'm guessing torsion damage helped here, transfering more blows to the neck. I could look at the data again to verify this.

I think it's not so much dwarves doing worse after the upgrade as goblins doing better, in particular short swords with armor removal and maces/whips with torsion damage. Piercing weapons (spears and pikes) don't benefit so much from those mechanics, so they fell behind.

It is odd that this doesn't  reflect in hits/kill, even if it does in killed dwarves (where maces perform best).
From the data about hammers, where steel also had more hits/kill while doing better in overall kills/wins, I'm thinking that the "large" bashing weapons (non-whips) knock opponents helpless relatively quickly. However, as they do some damage through armour, their wearers don't remove helmets to kill the helpless opponent in one strike, instead killing in half a dozen by simply bashing helmet and head to a pulp.

Again, I could look at how "early" in the combat fighters went down (falls over/looses consciousness) to check if this is truly the cause.

PS: I hoped you saved the 43.03 post as well. I recalled some the differences from memory, but it's important for both comparison and LNP purposes.
I did. I also still have all the files. I did not want to start another thread, though. I could just add the results here as a post, for comparison and LNP?
Logged

PatrikLundell

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #50 on: January 26, 2017, 04:24:20 am »

I'd recommend updating your second post to contain the 0.43.03 results so both sets are close to the beginning of the thread, and thus easily accessible.
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #51 on: January 26, 2017, 05:14:53 am »

S) So for steel spears, their actual progress is more like 28 killed zombies - 15 reanimated corpses?

Probs not, given the 1821 corpses for steel axes, but kinda hard to gauge, there - especially with steel axes only making 506 attacks, a fraction of killed enemies!

Overall, I guess I don't really know how to parse the implications of that table properly.

For unstoppable reanimation, why not just put the necromancer in an unclimbable location with complete line of sight to the battleground?

W) That hypothesis could work. Another one that would work is whips' longer prepare and recover leaving them more vulnerable and counterattacks to that situation being now more deadly - though I doubt this is the case, beyond the general buff to all attacks.

PS: Yeah, what PatrickLundall says.

fragfish

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #52 on: January 26, 2017, 08:42:35 am »

S) So for steel spears, their actual progress is more like 28 killed zombies - 15 reanimated corpses?
Yupp. As for most weapons.

Probs not, given the 1821 corpses for steel axes, but kinda hard to gauge, there - especially with steel axes only making 506 attacks, a fraction of killed enemies!
Oops! Guess I shouldn't post updates at 1 in the morning. I must have forgotten to replace that number. I'll go and put in the correct ones right away!

Overall, I guess I don't really know how to parse the implications of that table properly.
I'll fix the numbers, then it should be clearer. Previously, I had listed total values; now the tables mostly list averages.

For unstoppable reanimation, why not just put the necromancer in an unclimbable location with complete line of sight to the battleground?
I tried. Either they climb it and kill him, or he looses sight of the combat. Maybe I'll try some more.
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #53 on: January 26, 2017, 10:29:35 am »

Hm, yeah, that makes more sense now.

Lets see....32 zombies standing at the end of hammer fighting, taking about 32 attacks per killed zombie....

Hm?

Iron axe, total attacks/strikes/kill = 155. Killed things total 168. 372/168=2,21....Which is close, but not quite 2.4?

Steel axes are ok.

Iron spears, 361/(23+5+21)=7,36....Which is not 8.9 listed?

Ditto mismatch for steel spears and hammers.

Anyway, substraction still doesn't work out. How many zombies ended up standing at the end of typical steel axe fight, at least as far as the 13 losses went? 48-144? Nah, that makes no sense.

As for arena, not sure what the problem is. This box works (sealed from bottom and top with liquid's plugin obsidian walls)


The cats or troglodytes don't climb through fortifications, and both on one side get attacked. (Though other side archer doesn't attack back even then, for some reason.)

What setups have you tried?

PS: How do you get a necromancer in, by the way?

fragfish

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #54 on: January 26, 2017, 10:48:20 am »

Hm, yeah, that makes more sense now.

Lets see....32 zombies standing at the end of hammer fighting, taking about 32 attacks per killed zombie....

Hm?

Iron axe, total attacks/strikes/kill = 155. Killed things total 168. 372/168=2,21....Which is close, but not quite 2.4?

Steel axes are ok.

Iron spears, 361/(23+5+21)=7,36....Which is not 8.9 listed?

Ditto mismatch for steel spears and hammers.

Anyway, substraction still doesn't work out. How many zombies ended up standing at the end of typical steel axe fight, at least as far as the 13 losses went? 48-144? Nah, that makes no sense.

Careful using the numbers to recompute averages. Attacks, Kills and Efficiency are all rounded averages computed over a number of combats. That is to say, for each combat #Attacks, #Kills and Efficiency were computed separately, then averaged. As you can see from the large deviation for efficiency, they did not tend to be the same between combats. For Steel Axes, the efficiency numbers are so small that differences probably disappear when rounding.

Also, "Zombies killed" refers to kills on the initial 50 zombies, while those that got back up and were killed again are listed under "reanimated corpses killed". I split those because I wanted a rough number on how often the undead had to be killed. Would it be more clear if I labeled the first kill column "Initial Zombies killed"?

The number of corpses left standing is not in the statistics. I don't think it's possible to compute that from the numbers I've given.

As for arena, not sure what the problem is. This box works (sealed from bottom and top with liquid's plugin obsidian walls)
[SNIP]

What setups have you tried?

PS: How do you get a necromancer in, by the way?
I tried that setup once, sealing off a combat area and the part with fortifications one level above it. The necro got killed pretty much instantly. Same when standing on a tower. I made the box on only one space big, though, will have to try making it bigger.

Necromancers can be set by pressing "u" for Effect when adding a unit.

Anyhow, thanks for all the questions! It's nice to see someone go to the effort of checking the data. I'd have hated for bad results to remain in there (such as the 1821 kills).
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #55 on: January 26, 2017, 11:51:49 am »

1) Ah, well. No, I understood it fine, no need to rename....

The number of corpses left standing is kinda interesting to gauge what would be the average number of undead the group could handle. But even if I knew that number, for all I know it'd could be harder for iron axes to kill a group steel axes lost against once due undead heads, or the group of undead being larger after steel axes attack.

Hard to gauge. Don't even know how good spears were at putting down zombies for good - better than axes, maybe, with only two third a kill for each fully dead initial zombie - but for all I know, their slow kill speed meant just not getting more and still raising comparable amount to axes.

2) It'd be easier if you could use buildings, as they're unclimbable in general and would provide windows in particular; no idea about arena mode setups.

I assume you've tried a moat as well.

And, hm...



No jumping through the fortifications with these humans, here [43.03]

PS: Haven't really checked all numbers, tbh - just these looked weird.

Galena

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #56 on: February 06, 2017, 07:43:42 am »

Has anyone tested how wrestlers fair in the newest version? I'd also be interested and seeing if a squad with 50% wrestlers, 50% weapon users is more effective than one with no wrestlers.
Logged
He listed off what sounded like a lot of painful-sounding procedures that needed to be done, then proceeded to immediately throw a party.

SquareUnit

  • Escaped Lunatic
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #57 on: February 06, 2017, 04:27:44 pm »

Has anyone tested how wrestlers fair in the newest version? I'd also be interested and seeing if a squad with 50% wrestlers, 50% weapon users is more effective than one with no wrestlers.

Could be worth in considering an immobilised enemy is an easy target for the armed dwarf standing next to the wrestler
Logged

mikekchar

  • Bay Watcher
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #58 on: February 06, 2017, 07:23:44 pm »

I've never done any kind of objective study, but I often send out squads of wrestlers along with my armed squads.  What I've noticed is that actual *wrestling* is practically useless.  They hold one limb and then release it over and over again.  I guess it stops the adversary from dodging, which is about it.  However, *striking* seems super useful.  Because they all have legendary observer, they can find weaknesses and attack them.  I'm guessing it's the same as the blue exclamation mark when fighting in adventure mode.  Whatever it is, I find my wrestlers often striking the third toe (or something useless like that) and exploding it into gore. The adversary falls over and is quickly dispatched by the axe dwarf.

I'm not sure how practical it is, though, because it is *very* random.  Also, without shields, they are sitting ducks for archers.  If there were some way to run the fights a bit more tactically, I think they could be quite powerful, but as it is it's probably not worth it.
Logged

Fleeting Frames

  • Bay Watcher
  • Spooky cart at distance
    • View Profile
Re: From Iron to Steel: A survey of v0.43.05 combat [Long] *VERSION UPDATE*
« Reply #59 on: February 14, 2017, 06:15:51 pm »

Recalled that whips sometimes kill colossi, while in my previous tests inferior material morningstars fail to significantly wound them unlike steel or adamantine ones (still cutting, after all).

So ran a quickie in 43.03

Spoiler: dwarf (click to show/hide)

Spoiler: before (click to show/hide)

Spoiler: after (click to show/hide)

Literally none of the 133 bronze colossi died. But...


Two lost an arm ("Dwarf 47 lashes Bronze Colossus 59 in the right hand with her silver whip and the severed part sails off in an arc!" "Dwarf 50 lashes Bronze Colossus 8 in the right hand with her silver whip and the severed part sails off in an arc!"), more had some or several white parts. Better result than gold morningstars, that's for sure, though worse than actually sharp ones. This could give a reason to use whip squad instead of warhammer or macing one, if you have no good sharp weapon materials and have to choose.

Bit harder to track without log analyzing how much death was due exhaustion and how much due injuries, but this at least satisfied my idle thought a bit.

(Of course, single omni-grandmaster is a different thing from 10 omni-proficients. )
Pages: 1 2 3 [4] 5