Print

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

[Urist Tilaturist]

Is the AI failing to get the syndrome completely random, or are there any shared conditions between the times when it happened not shared by the times when it did not?

[Grimlocke]

The only pattern I have been able to make out is that it tends to fail only at the start of a fight.

Also Sergarr, I presume with that 200 you refer to the bone density thing? I made my test subject a grand master puncher (which sounds very silly), then made an unconscious test guy and punched his hand with the heavy attack until it exploded. The idea being that if it doesn't happen in the worst possible circumstance, it won't happen any other time.

So far the only thing that got it to work for me is modifying the creature attacks themselves, setting the velocity to 500 and contact surface to 1000 (you can set that past 100% it seems).

Though I have been messing with my metal raws, now that I think about it that would explain the different result.

I have been trying to sneak it in through the gaits creature variation, but can't get it to place the velocity tags right.

[Urist Da Vinci]

...
Huh, 2000:30000 worked. The wiki claimed 5000 to be about as high as this goes, guess that's not how it is then.

Anyhow, after generous use of a toad leather whip on a very unfortunate test subject, I obsidianized my lasher and watched the test subject crawl away and pass out right as the syndrome ran out.

Neat!

There is still a bit of an issue with the AI occasionally failing to actually get the syndrome after performing the interaction. Anyone have an idea what happens there? When I take control of the creature the interaction shows as having been used, but no sign of the syndrome.
...

Re numbers, 32767 = (2^15)-1 so higher numbers than that overflow

Re Syndromes/AI:
- Test syndromes should always incorporate the zombie flash, so that you have visual proof that it has hit
- Time is different in Arena oversight/Dwarf mode than in Arena direct control/Adventure mode. This can screw with syndrome start and end times.

...
The best way to do this would still be just altering the creature attacks one way or another, but that involves changing every single creature entry in the game, as well as any modded ones that might be added. Not looking forward to that.
...

Why not mod just the creature attacks on the intelligent civ races, especially the playable ones? Does anyone care if *random animal* kicks too hard? Actually, alligator vs cow natural attack modding is a topic of its own.

[Grimlocke]

...
Huh, 2000:30000 worked. The wiki claimed 5000 to be about as high as this goes, guess that's not how it is then.

Anyhow, after generous use of a toad leather whip on a very unfortunate test subject, I obsidianized my lasher and watched the test subject crawl away and pass out right as the syndrome ran out.

Neat!

There is still a bit of an issue with the AI occasionally failing to actually get the syndrome after performing the interaction. Anyone have an idea what happens there? When I take control of the creature the interaction shows as having been used, but no sign of the syndrome.
...

Re numbers, 32767 = (2^15)-1 so higher numbers than that overflow

Re Syndromes/AI:
- Test syndromes should always incorporate the zombie flash, so that you have visual proof that it has hit
- Time is different in Arena oversight/Dwarf mode than in Arena direct control/Adventure mode. This can screw with syndrome start and end times.

...
The best way to do this would still be just altering the creature attacks one way or another, but that involves changing every single creature entry in the game, as well as any modded ones that might be added. Not looking forward to that.
...

Why not mod just the creature attacks on the intelligent civ races, especially the playable ones? Does anyone care if *random animal* kicks too hard? Actually, alligator vs cow natural attack modding is a topic of its own.

Thanks for the number info, guess I went a little overboard with that then (its percentual increase + arbitrary increase).

For the syndrome I used [CE_DISPLAY_TILE:TILE:'U':1:0:1:START:0:END:1000], mostly because it was the first thing I ran into. It consistently fails to show on creatures who have actually used the interaction. What I have noticed is that the failure seems dependent on the target, if I sic a whole group of creatures onto a single creature then either the whole group lights up, or none light up at all.

It happens both in arena mode and arena 'adventurer' mode, switching back and forth does seem to mess up the phase count yes.

I am beginning to suspect this might be something buggy with the arena AI since I have had several cases where creatures simply refused to attack or even defend against things trying to eat them.

I guess the entity creature attacks would be the most important ones, but there is also things like war dogs, trolls and the occasional incident where my fully armored adventurer gets trashed by a donkey or something.

What I have done now is hijack the crawling gaits creature variation entry, which conveniently is used in just about every creature that can move at all. Tested by switching every tissue to spleen tissue, which was funny

It works for increasing the surface of various melee attacks and keep the worst sillyness at bay and keeps standing armored humans safe from a grand master puncher/kicker, except for the occasional fluke which only bruises the head but still knocks them out. It's a bit radical as it is as it just takes every attack surface token that is 100 and makes it 1000, but for the stock creatures it should not result in anything too weird.

[Sergarr]

Have you tested how fast the unarmored grand master punchers kill each other with this change?

[Urist Tilaturist]

Animals should also have sensible attacks, so alligators bite and drag the target underwater to drown it, for example, but the first priority should be to get sapient races and maybe beasts fighting right.

Legendary punchers should have a fair chance to break small bones (ribs, hand, jaw) on a hit, but pulping the hand entirely should not happen.

Fist fights should only usually turn lethal if extremely violent wrestling moves like chokes or slamming the foe head first into the ground are used. A weak dwarf might die from a hard punch to the chest or head, but it should not be common even for legendary punchers to kill their foes with a punch.

[Sergarr]

Have you figured out how to make knockouts happen more often in combat?

They happen really, really rarely when you hit the head, and I'd like for them to happen more often.

[Urist Tilaturist]

Does anybody know if there is a raws value which decides the chance of a knockout?

Knockouts should be possible for any hard hit to the head, but likely if the hit was under the jaw and jerked the head backwards.

[Urist Da Vinci]

Have you figured out how to make knockouts happen more often in combat?

They happen really, really rarely when you hit the head, and I'd like for them to happen more often.

As far as I can tell, it is a random chance when you hit the [HEAD]. Damage to internal components not required. In the disassembly it is nearby the part for lungs-breathing and guts-nausea. It used to be a random chance when a THOUGHT part was hit back in 0.34.11 or so, but that made it near impossible to knock out without also killing the person.

[Grimlocke]

Knockouts seem to happen randomly on a high-contact area, high-momentum blunt attack to the head, bit like organ damage from blunt attacks. As far as I know there is no way to regulate it with raw changes.

I have been getting them semi-regularly with punches and some large contact area edged attacks from polearms.

They also happen with a puncher hitting a helmet. That really needs to stop.

The unarmed and unarmoured combat I have now mostly causes bruises and organ damage, but also 'pulps' hands according to the combat log, but weirdly enough they are still usable and heal fairly quickly. Not sure what's up with that. I also saw a few combat reports that supposedly punched a hand off, but when I looked it was still there...

Anyhow most fist fights end reasonably quickly with one of the fighters getting knocked out or getting a leg broken and getting their head pulped. Wrestling is definitely more effective for destroying limbs, though the AI does not use it anywhere nearly as often.

[Sergarr]

I'm afraid fixing combat further is going to be impossible since the game doesn't really model the attacking material physical properties enough for that.

Right now every single blunt attack is performed as if it was by an absolutely rigid object. That means that all "soft" materials will not behave as expected when used for attacks.

[Urist Tilaturist]

Head pulping should just not happen with fists. Severe damage and bleeding, yes, but not the whole thing being crushed. Only a beast should be able to achieve that bare handed.

[Grimlocke]

Thing is, if the punch attack is too weak to kill anything the AI will get stuck on unconscious enemies and just keeps punching them until the attacker passes out from exertion. And it's not like it splashes heads in a single punch, usually they take 5 to 10 attacks to do any major damage.

[Urist Tilaturist]

Is it possible for the brain to be damaged without crushing the skull? Or for the heart to be damaged without its actually being pulped?

Even on no quarter, combatants should leave a downed and battered foe for dead rather than punching it over and over again.

[Urist Tilaturist]

So will there be an update to this thread, condensing all the current raws changes into one downloadable package?