Bay 12 Games Forum

I want to isolate the attributes that are used in dodging.

HYPOTHESIS
The attributes that are trained by practicing Dodging (Agility, Toughness, Endurance, Willpower, Spatial Sense, Kinesthetic Sense) are the attributes that contribute to success in dodging.

METHOD
The Object Testing Arena does not appear to allow you to modify attributes, so I elected to do the following:
1. Generate a pocket world with exactly one type of bogeyman so as to reduce randomness in my foe.
2. Create a Human Hero Adventurer with Average attributes except for one Superior Attribute and dump the Skill points into Dodging solely.
3. I slept all day (to avoid Drowsy) and then waited all night for bogeyman attacks. If nobody attacked before my Hunger/Thirst became two levels deep (about 16-ish hours?) then I simply restarted.
4. I moved to let approximately one bogeyman attack me per round. I did not attack, to conserve energy.
5. I ended the test when a speed-affecting status ailment occurred (nausea, stunned, winded, etc.).*
* I had to modify this approach when I realized that superior Toughness or Willpower will mean you can be so hurt as to be no longer dodging effectively but still at normal speed. I wound up trying to get about 50-100 attempted attacks in per test. Interestingly, with a superior Toughness or Willpower you can endure about 4 times as many attempted attacks from a bogeyman than otherwise.
6. I then parse the gamelog using a previously posted !!science!! perl script, modified for my purposes.

RESULTS
The Strength results were my control. With my methodology, there is some noise in the data. I do not have hundreds of blows, I do sometimes get bounced from one bogeyman to another, and there's normal human variation in my attributes.

The hit rate of the initial bogeyman test was (1.0 = every blow lands):

Strength	.77
Agility	.58
Toughness	.84
Endurance	.75
Willpower	.50
Spatial Sense	.81
Kinesthetic Sense	.76

I then attempted to do some testing based on Human Hero Adventurers with below average attributes in order to get multiple Superior attributes. I quickly discovered, though, that a low Intuition and Focus will cause severe impairment in your ability to dodge (you live such a short time that I have no decent data). I believe this is a function of rolls modified by the Observer skill made to detect incoming blows.

Further experiments with Human Demigod Adventurers showed that adding significant levels of Intuition & Focus attributes did not materially affect the dodging outcomes (.60 hit rate). However, increasing Observer skill a great deal did have some modest effect on dodging outcome (.38 hit rate with Superior Willpower, Focus, and Intuition).

Skill level in dodging remains the single largest effect on dodging chance, of course.

CONCLUSION
Willpower and Agility are the key attributes for making successful dodge rolls. Other attributes don't materially improve the odds (at least in isolation, the only thing I am attempting to study), but subpar Intuition and Focus attributes will definitely harm your dodge rolls. Other subpar attributes may harm your dodge rolls, as well.

NEXT STEPS
It's probably smarter to mod the raw files to mutate the Average Human Male and pit him against a sufficiently skilled dwarf to test attributes. The results would be far more repeatable and it would be less effort to generate more depth of data.

I am now curious how much Observer/Focus & Intuition & Spatial Sense is required to dodge multiple incoming projectiles. I am very likely to test this.

I am also curious how body size, length, and broadness affect the dodging chance. My limited tests seem to show that Adventurers have a 1/36 chance of being extra fast (super thin and short) which may have some impact on miss chance. I'm not sure if I care or not. It is sufficient for my purposes to know that the elven agility boost is important when attempting to dodge.

I think there is also some wound severity stuff going on here. My highly skilled Observer test did not incur any medium/heavy wounds. I may attempt to decipher this.

I feel like there might be some difference between moving and waiting, but as one is generally never waiting for an incoming blow in .34 I don't think I care.

There is probably a minimum set of attributes which must be at least average to successfully dodge a few dozen blows in a single fight. I don't think I care.

Quote from: seht on September 07, 2019, 09:04:28 am

Quote from: Meph on August 24, 2019, 10:18:13 pm

Quote from: mikeydmc on August 04, 2019, 03:31:08 pm

When running Meph+Vordak graphics, some of the walls are displaying as other things such as a cactus on sand

That's mostly happening in adv-mode, it's constructed walls adjacent to so-far unrevealed constructed walls. I did fix that in my own release, the Meph tileset version in the LNP is severly outdated.

So, to be clear, your pack is now fully functional in adventure mode, even when using your tileset and twbt? Or am I misunderstanding?

twbt doesn't work well with adv mode. no matter which pack.
it's a problem of twbt, not the pack.

When running Meph+Vordak graphics, some of the walls are displaying as other things such as a cactus on sand

I think actual migrant waves are still limited to dwarves for some reason.

At least vampires and weres alive from world gen continue to exist (until slain) and they can infect new ones in fortress mode at least (I haven't looked at dates for Legends entries to see if they also continue to spread the curses outside of the fortress, although attempts to infect prisoners and release them to try to bring down invaders as they turn when back home haven't been successful). NEW vampire and were curses in the world isn't much of an issue, as the rate at which that happens is so low that it probably wouldn't happen during the existence of a normal fortress anyway (or the world would be overrun by them).

The forum has had reports of curse generation in fortress mode, so it can be assumed to work, with the tricky part being to get access to a suitable god (i.e. one that dispenses curses).