KD7BCH

WOLVES

Recommended Posts

In the current build even if you use Autoclicker by Shocker to click as fast as your cpu will register to flood the interrupts the game maximizes your clicks at a level which in certain situations is impossible for you to outclick the bar by which the wolf will still mortally wound you. WHAT FUN IS THAT? Why would anybody want to play a game where they invest in dozens of hours on a run, only to have it end where a wolf gets a jump on them when they are tired and instead of letting the player's skill and fending off the game "eliminates" your chance of coming out of it on it's own.

I agree completely with that particular criticism. I just started a Stalker run the other day, and I was doing pretty well, until I tried to shoot a wolf running at me and missed. It was Game Over real quick at about 71% and freezing. Had a knife. Made no difference.

One mistake, and that was it. The current Mash-option seems to work just fine for some players, whereas for others, it's just random and will take a huge chunk out of your health, no matter how fast you push.

However, a spear will not solve that particular issue any more than the torch will. Personally I found the two-button mash system better in that it actually required some skill. Although frustrating at first, you'd get better at it over time.

And again, I dare you to imagine how wolf AI behaviour could be improved, and to plot out how much man-hours development time it will take to program that kind of behaviour into the game.

TWM if you agree the game should not be luck based on that the lets start there. What of the 2 button system? That was better than the masher we have now.

The whole coding of the wolf AI is a separate issue, if you need a system that doesn't take much coding time build a similar type of system as the Fallout targeting system that stops time, could do that right in the UI that already exists, make it a system of die rolls with modifiers. I don't know how to code at the level they are in TLD but I wrote up a bit on how to handle the wolf actions.

Wolf has 100 pts before it dies. If wolf hp drops below 100 it dies and can be harvested.

Wolf attacks with X points each time it tries to bite you.

Wolf starts with base Z chance to bite you. Die roll of 20 modified by how fatigued you are and how encumbered you are by gear. +1 under 30lbs ,+2 30-60lbs , or +3 over 60lbs to all rolls for how encumbered you are, and +0,+1,+2,+4, for how tired you are based on the fatigue bars, +4 is when you can no longer run.

For 1-14 roll Bite misses.

For 14-15 roll Wolf bites, tears clothing.

For 16-18 roll Wolf bites, causes Condition Damage A Die roll of 3-8 condition loss.

For 19-20 roll Wolf bites, causes Condition Damage B Die roll of 5-15 condition loss, also causes "blood loss" needing bandages, potentially infect-able by current mechanic.

Player strikes with Y points each time wolf tries to bite. Die roll or 20 modified by what weapon you have. +0 Fists, +2 for prybar +3 spear, +5 hatchet, +8 knife.

Die roll of 20.

For 1-10 roll player misses, causes no damage to wolf.

For 11-12 roll player causes 20 pts of damage to wolf. (severe hit)

For 13-15 roll player causes 40 pts of damage to wolf. (brutal hit)

For 16-20 roll player causes 50 pts of damage to wolf. Wolf immediately disengages running away.

For each hit, roll a D20 to determine if wolf runs off returns back to agitator state or continues to try to bite. Each hit is for Y points is cumlative, so for 2 hits above roll 2 die to determine wolf actions. for 3 hits, use 3 rolls, Use highest roll to determine wolf action.

For 1-15 roll wolf continues to try to bite.

For 16-17 roll wolf returns to agitator state.

For 18-20 roll wolf runs off.

Right now once you puncture a wolf and it runs off it will eventually end up dead somewhere. The nature of blood is such that it is designed to close up a wound and prevent this. As is wolf saliva in fact. So the current system should have another such accessory system which uses die rolls based on overall hps brutality and number of hits. Could also be used to indicate how much blood a wolf is losing by dropping more blood spots on the ground as it retreats. Nothing too gory just not the same for every situation. If you cause 4 hits severe hits, the wolf should be bleeding from 4 places, and should drip blood from each of those. A system of losing hp each minute with a chance for the wound to close would keep the game play of a wolf that is severely injured largely the same and being able to find it dead soon or if the wolf took a heavy hit lost alot of hp in one go and ran off, it might very well survive having appeared to decide not to mess with you after having the bejesus knocked out of it. Yet that wolf will live to come around and bite you another day. This level of AI is not impossible or impractical to code, a staple of 8 bit gaming was hp based combat. Do you need to see the hp at work no? Do you need to see more blood for more wounded animals yes. Do you need to be able to survive these attacks from a position of relative strength yes. Does the current system do these things rather than give the whole "luck" the real shake? No the current system lets you click all day long and doesn't let you live if it pre-determines you won't even from high levels of condition.

If you want to keep the current system fine. But adjust it so players at high condition dont die because of one interaction and don't concentrate six wolf interactions in an area where there is no logical reason for them to be in.

Share this post


Link to post
Share on other sites
I am still confused as to why a bow and arrow, both of which are complicated pieces of equipment and require a deep level of knowledge to make, are "ok" for the game, but a spear isn't.

Bow and arrows: deep knowledge of physics, how that particular piece of wood is going to react when bent, how archers paradox will affect the arrows, etc

Spear: sharp length of wood.

I am still aggravated that a bow and arrow was included, just throwing that out there.

As for the whole argument of "torches are just like spears!"...... I know of no fire in the world where you can start it in 30 seconds, on snow mind you, and after throwing in a couple of frozen twigs you can pull out a flaming brand that burns for hours :roll:

Spears are infinitely more "realistic" than torches, if for that reason alone

Not just from a realism prospective but I think they would bring better gameplay. The reasons for objection to their introduction to current gameplay don't really trump the fact that current wolf gameplay has not been that good for a long time.

"I hope they never change TLDs Wolves because the are just so fun to interact with." Said nobody anywhere.

Share this post


Link to post
Share on other sites

And again, I dare you to imagine how wolf AI behaviour could be improved [...]

Imagine: Instead of a wolf derp-charging you as soon as you enter a specific radius, the wolves will just run away from you when you make contact ( so long as you are healthy). You shrug, and head back to your chosen base. Out of the corner of your eye, you see movement suddenly. Looking back, you see a wolf back by the last bend in the trail. Then again, you see movement. And again.

You are being followed, and surrounded. They aren't doing anything, just following you...... watching.

Perturbed, you hurry to your base. Maybe they will be gone in the morning. And, they are. But you find plenty of tracks around your base. Circling, waiting, even getting close to the door and windows. "Whatever", you think, you have a gun/bow/spear. You can handle yourself.

All throughout the day, you keep noticing a wolf in the background, watching. You start to get freaked out when whenever you turn around, you see a wolf (different ones each time) get closer and closer. They aren't attacking (yet), but they are definitely interested in you. You try to hustle back to base, only to get cut off by a wold, right in front of you! It advances towards you, snarling, and you hear the same noise behind you. You are surrounded and cut off from shelter. The wolf in front of you advances, and you move to defend yourself, only to get jumped by the wolf behind you. Game over.

Wolves can already:

-"smell you"

-track you

-chase you

-"threaten" (by this, I mean "advance to you snarling)

ALl they would need is some adjustments to the "engagement" AI to be actual living, breathing, intelligent predators.

Seriously, for a game that touts the whole " we don't have zombies!" shtick, the wolves in this game are totally little more than furry zombies.

I think maybe you might be underestimating how difficult it would be to successfully implement what you describe . But it'd certainly be a big improvement if they got somewhere close!

But really with a resource pool as they have, and a daily interaction in the game as they are, should they not have a team working on this? I mean really shouldn't one guy at least be devoted to animal AI? Look at how cheesy rabbits are in the game world, they run at you sometimes but when they do they run fast, so you don't catch them. They run away from you, they squeak when they get eaten by wolves. They are substantially more

believable than the wolves which you find just walking around. If the AI was better, then the need for so many of them would be much less.

Share this post


Link to post
Share on other sites
All throughout the day, you keep noticing a wolf in the background, watching. You start to get freaked out when whenever you turn around, you see a wolf (different ones each time) get closer and closer. They aren't attacking (yet), but they are definitely interested in you. You try to hustle back to base, only to get cut off by a wold, right in front of you! It advances towards you, snarling, and you hear the same noise behind you. You are surrounded and cut off from shelter. The wolf in front of you advances, and you move to defend yourself, only to get jumped by the wolf behind you. Game over.

Thank you Boston. See, now we're finally getting into a conversation that is actually stimulating.

And I agree, this kind of wolf behaviour would be more immersive. At the same time, you'd have to find a way to be able to postpone this kind of outcome. Otherwise it just means you're simply spending even more time building your character only to have him/her killed off inevitably anyway (which is part of the complaint about Stalker).

It might get repetitive and therefore tedious after a few times as well. At a certain point I picture myself calculating how many days it will take for the wolves to make their final move. So you'd need a balance between predictability and unpredictability in wolf behaviour, to keep the game fresh but yet offer the possibility for the player of finding a counter strategy that will work.

Most importantly, I think you're underestimating how much coding would go into developing such an AI. It wouldn't just require individual wolf AI at different stages of contact, it would also require coding for collective pack AI. And then you'd have off encounters with a lone wolf every now and then, which might either keep it's distance, run off, stand it's ground or defend it's kill. All of which require further coding.

At this point, wolf AI is pretty straightforward and simple. They patrol a certain area, with some minor variations and periodes when they're absent; they attack the player pretty much on sight and will halt or run off given certain player actions. The number of variables is relatively limited. As someone noted: they're basically functioning as zombies.

What you're suggesting though requires way, way more sophisticated coding, we're might be talking months of coding, bugfixing, and if the result isn't satisfactory either, that's just a huge waste of resources flushed down the toilet. It would be a big gamble for Hinterland. I'm not sure if I would bet on that in this stage of development if I was them.

  • Upvote 1

Share this post


Link to post
Share on other sites
All throughout the day, you keep noticing a wolf in the background, watching. You start to get freaked out when whenever you turn around, you see a wolf (different ones each time) get closer and closer. They aren't attacking (yet), but they are definitely interested in you. You try to hustle back to base, only to get cut off by a wold, right in front of you! It advances towards you, snarling, and you hear the same noise behind you. You are surrounded and cut off from shelter. The wolf in front of you advances, and you move to defend yourself, only to get jumped by the wolf behind you. Game over.

Thank you Boston. See, now we're finally getting into a conversation that is actually stimulating.

And I agree, this kind of wolf behaviour would be more immersive. At the same time, you'd have to find a way to be able to postpone this kind of outcome. Otherwise it just means you're simply spending even more time building your character only to have him/her killed off inevitably anyway (which is part of the complaint about Stalker).

It might get repetitive and therefore tedious after a few times as well. At a certain point I picture myself calculating how many days it will take for the wolves to make their final move. So you'd need a balance between predictability and unpredictability in wolf behaviour, to keep the game fresh but yet offer the possibility for the player of finding a counter strategy that will work.

Most importantly, I think you're underestimating how much coding would go into developing such an AI. It wouldn't just require individual wolf AI at different stages of contact, it would also require coding for collective pack AI. And then you'd have off encounters with a lone wolf every now and then, which might either keep it's distance, run off, stand it's ground or defend it's kill. All of which require further coding.

At this point, wolf AI is pretty straightforward and simple. They patrol a certain area, with some minor variations and periodes when they're absent; they attack the player pretty much on sight and will halt or run off given certain player actions. The number of variables is relatively limited. As someone noted: they're basically functioning as zombies.

What you're suggesting though requires way, way more sophisticated coding, we're might be talking months of coding, bugfixing, and if the result isn't satisfactory either, that's just a huge waste of resources flushed down the toilet. It would be a big gamble for Hinterland. I'm not sure if I would bet on that in this stage of development if I was them.

In terms of animal AI, it might be interesting to check out the development of Rain World (devlog to be found on tigsource). Animal behaviour there is apparently determined by a number of variables that are kept for every animal, like hunger, fear, curiosity etc. Animals make decisions based on these variables and the situation they are confronted with, eg. slight hunger - see prey - approach - see larger predator nearby - fear rises - hide - wait - fear level goes down - surface again and carefully look around for predator etc. The mechanism can be fairly simple for every single animal but allows for very interesting and complex interactions not only with the player avatar, but also between different animals. Sometimes things happen that even surprise the devs. This kind of thing can make a game really fascinating in my opinion.

I am not a programmer, but I would think that a more complex wolf behaviour could be done technically without demanding too much resources. The big question is what it would do to the game. Currently the wolf mechanics are quite simple and therefore easily understandable and predictable. Most players understand where the wolves are (after running into them here and there of course), how they work and how their mechanics can be "exploited". For some people, myself included, understanding and outwitting these mechanics may bring about a feeling of satisfaction and be part of the fun this game delivers. But as we can see, even the current simple wolf mechanics are too much for some players. If these mechanics became more intricate, the game might be confronted with even more people complaining about wolves and how impossible it is to avoid/survive them, and that they desperately need a spear or a bazooka or whatever to be able to survive on stalker.

It's also interesting to note how you can advocate for the introduction of a spear and a wolf AI overhaul for all the right reasons (Boston) or - on the other hand - for all the wrong reasons (OP).

What my ideal wolf world would look like is this: We would have a few solitary wolves roaming the map freely and unpredictably. Their behaviour would be determined by some variables like fear, pain, hunger, anger etc., and they would generally detect and avoid the player from long distance unless... well, unless the wolf is hungry and the player is hurt for example. I am unsure whether these solitary wolves should be able to catch and bring down a deer... maybe it would be better if they couldn't. I would like them to roam the map in search of deer carcasses and finish them off. We could have the occasional deer carcass spawn somewhere on the map, and the player would have to go searching for them in the morning. And sometimes there would be confrontations with a lone wolf over a deer carcass. It would be incredibly cool if those loners would be recognizaable individuals with a certain character and memory, and if they acted according to that.

And then there would be packs that would enter and leave the map at certain intervals. For the sake of fairness, the player could get a warning for the time that a pack is roaming the map, maybe a red wolf icon on the screen or something. And then he/she would have to decide whether to stay at home and wait it out or whether to take and risk and go hunting/gathering/looting while the pack abounds.

Share this post


Link to post
Share on other sites
All throughout the day, you keep noticing a wolf in the background, watching. You start to get freaked out when whenever you turn around, you see a wolf (different ones each time) get closer and closer. They aren't attacking (yet), but they are definitely interested in you. You try to hustle back to base, only to get cut off by a wold, right in front of you! It advances towards you, snarling, and you hear the same noise behind you. You are surrounded and cut off from shelter. The wolf in front of you advances, and you move to defend yourself, only to get jumped by the wolf behind you. Game over.

Thank you Boston. See, now we're finally getting into a conversation that is actually stimulating.

And I agree, this kind of wolf behaviour would be more immersive. At the same time, you'd have to find a way to be able to postpone this kind of outcome. Otherwise it just means you're simply spending even more time building your character only to have him/her killed off inevitably anyway (which is part of the complaint about Stalker).

It might get repetitive and therefore tedious after a few times as well. At a certain point I picture myself calculating how many days it will take for the wolves to make their final move. So you'd need a balance between predictability and unpredictability in wolf behaviour, to keep the game fresh but yet offer the possibility for the player of finding a counter strategy that will work.

Most importantly, I think you're underestimating how much coding would go into developing such an AI. It wouldn't just require individual wolf AI at different stages of contact, it would also require coding for collective pack AI. And then you'd have off encounters with a lone wolf every now and then, which might either keep it's distance, run off, stand it's ground or defend it's kill. All of which require further coding.

At this point, wolf AI is pretty straightforward and simple. They patrol a certain area, with some minor variations and periodes when they're absent; they attack the player pretty much on sight and will halt or run off given certain player actions. The number of variables is relatively limited. As someone noted: they're basically functioning as zombies.

What you're suggesting though requires way, way more sophisticated coding, we're might be talking months of coding, bugfixing, and if the result isn't satisfactory either, that's just a huge waste of resources flushed down the toilet. It would be a big gamble for Hinterland. I'm not sure if I would bet on that in this stage of development if I was them.

So dismissive. Fine. Fuck the AI. Give me a SPEAR! I'd also like to add torches are no guarantee of survival.

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.