Print

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

[sean_hyer]

In preparation for another embark as humans, I decided to do some optimisation testing.

Humans have access to almost every weapon (all of them besides large daggers, I believe), and a few of them look potentially powerful. So I plan to run combat tests in the arena.

I haven't got there yet. The first thing I did was fight humans against dwarves. Humans are larger, dwarves have better stats - which is better. The answer appears to be neither. From peasants to grandmasters, most fights ended up with only a few percentage points of differences in winrates. Dwarves have an edge in large combats thanks to trancing, but that's to be expected.

Then I found an oddity:
If a low-skill dwarf (competent or unskilled) in iron armor with an iron sword fights an identically skilled and equipped human, the dwarf wins 60-65% of the time.
My trials were carried out in runs of 144, and these particular results were strange enough I ran them each twice. 576 is a large enough sample size I'm confident these results are statistically significant.

This should be completely replicable, since I used a clean install of Dwarf Fortress. The only change I made was deleting every creature except dwarves and humans* for ease of use.

Does anyone have any idea why this might be? With the same equipment, but grandmaster skill, the result was a draw (humans win 51%).

*bogeymen, nightmares, and weremice are still able to be placed in the arena. Why weremice? Who knows?

[delphonso]

How often did the trance come into play? Perhaps against grandmasters, the flurry isn't as strong, but against unskilled fighters the flurry that a trance gives could be devastating. That's my only guess.

[FantasticDorf]

Dwarves are harder to hit (even by a margin) because they are smaller, being unskilled affecting the competency of how likely it is to connect and dodging, the gap is closed by competent swordmasters negating this trivial height difference, and masters can hit the smallest creatures deftly most times.

Its why forgotten beasts can't often hit kittens who are widly dodging about, whereas humans are a larger target, even if they only graze the targeted contact area.

[sean_hyer]

I can confirm it's not trancing. That only occurs when a dwarf is fighting multiple enemies.

Once it was suggested, I thought maybe the close layout of the combat cells was fooling the AI, but I ran it again to check and no dwarves tranced.


I don't think that small size is an advantage at low skill levels, because with an iron sword and no armor or skill the numbers go back to even, which wouldn't be the case if size mattered.

Perhaps their greater strength means dwarves are naturally better at wearing armor?

[Funkmaster Rick]

I bet you their increased strength makes them more likely to deal damage through armour. They're not better at wearing it, they're better at dealing with it.

[iceball3]

From what i remember: momentum is actually slightly held back by the mass of items in DF; that is, if you make something heavier, the momentum does not scale up linearly, because the creature struggles to get the weapon to speed.
This even happens with musculature to a more limited extent, i think, though being fat or having less muscle than other tissues would have harsher implications. Anyone happen to have links to the related testing?

Perhaps dwarves at low skill have a momentum advantage concerning their base attributes, versus humans as well, with the following:

Code: [Select]

[PHYS_ATT_RANGE:STRENGTH:450:950:1150:1250:1350:1550:2250]              +
[PHYS_ATT_RANGE:AGILITY:150:600:800:900:1000:1100:1500]                 -
[PHYS_ATT_RANGE:TOUGHNESS:450:950:1150:1250:1350:1550:2250]             +
[MENT_ATT_RANGE:ANALYTICAL_ABILITY:450:950:1150:1250:1350:1550:2250]    +
[MENT_ATT_RANGE:FOCUS:700:1200:1400:1500:1600:1800:2500]                ++
[MENT_ATT_RANGE:CREATIVITY:450:950:1150:1250:1350:1550:2250]            +
[MENT_ATT_RANGE:PATIENCE:450:950:1150:1250:1350:1550:2250]              +
[MENT_ATT_RANGE:MEMORY:450:950:1150:1250:1350:1550:2250]                +
[MENT_ATT_RANGE:SPATIAL_SENSE:700:1200:1400:1500:1600:1800:2500]        ++

[adseaghtrnjty]

I was testing humans vs dwarves and found that an unskilled human could easily kill any number of unskilled dwarves if you're smart about it. The computers are not.

Charge, knock them down, hand on throat, choke out. I've done that multiple times without much injuries as a human.

Now it gets even weirder. For a human vs dwarf, I was trying to do the same thing with these conditions.

iron helm
iron breastplate
iron gauntlet(s)
iron high boot(s)
iron shield
iron shortsword

no skills, no extra stats.

One would think a human could just rush the dwarf, who is smaller, knock them down, attack, but I found that even if my charge worked, I would tumble backwards or tumble together and fall over even though I'm supposedly the bigger entity. I had just even controlled the dwarf and found he was only stunned. For wrestling, you would think no matter a human and dwarf at base strength, it wouldn't be possible for the dwarf to wrestle the human because humans are so much bigger, but I couldn't break the grip of some dwarves. It could be possible that it randomizes stats though.

I had expected something like, you charge the dwarf, the dwarf falls over, is stunned, and you win. Doesn't that make more sense in a successful charge?

[Sver]

Arena randomizes the size of creatures at spawn. Minor differences are hard to notice, but otherwise you can monitor it by viewing their description: if it says they are tall/short, broad/narrow or fat/thin, it will have a noticeable impact on combat.

It is also possible, although uncertain, that the weight of the equipment impacts calculations when it comes to charges, pushes and wrestling moves.

[anewaname]

The weight of the armor is probably a factor; the human's breastplate/mailshirt/greaves weighs about 15% more. If players believe dwarfs with no Armor User skill in full armor are slower, what about "weaker" humans in full armor? Speed or encumbrance might effect both attack and defense.

And, dwarfs do have a higher average toughness rating. The wiki says "Reduces physical damage of all kinds", which I theorize would help the dwarf with hits that were converted to blunt damage, and there might be more hits converted to blunt damage in a fight between unskilled fighters. Without armor, being slashed for a little less damage probably still puts the dwarf on the floor in pain.

[iceball3]

The weight of the armor is probably a factor; the human's breastplate/mailshirt/greaves weighs about 15% more. If players believe dwarfs with no Armor User skill in full armor are slower, what about "weaker" humans in full armor? Speed or encumbrance might effect both attack and defense.

And, dwarfs do have a higher average toughness rating. The wiki says "Reduces physical damage of all kinds", which I theorize would help the dwarf with hits that were converted to blunt damage, and there might be more hits converted to blunt damage in a fight between unskilled fighters. Without armor, being slashed for a little less damage probably still puts the dwarf on the floor in pain.

From what i've gathered plumbing through the occasional infodumps people have left on the forums, i got the impression that toughness was, before the pulping update, really much not useful, as all it would affect was a slim margin of how severe injuries from certain blows were by relatively scaling back the effective force behind them (or something similar to that), and now is slightly more useful as small injuries can accumulate on a body part and eventually mince it to pieces.

[MaxTheFox]

*bogeymen, nightmares, and weremice are still able to be placed in the arena. Why weremice? Who knows?

Arena (and worldgen) werebeasts are random.