Reduce group spawn of wild animals. #25792
Conversation
|
I will have to see how well this PR will work in practice though I do like the idea, as well as the one you have for future additions, as per additional notes. I think doing it like quest markers upon an attempt to track an animal/pack rather than making it like hordes would be better, but that's a discussion for when you'll get to it. |
|
I agree there is a lot that could be improved, but reducing spawn rate is a good measure in the meantime. |
|
While it might be that wild spawn is too high, I don't like dropping it to 1/10th of current spawn rate. What's wrong with reducing it to 1/2 first and see how that works? Why such drastic change? Or should I take the name of your branch "no-meat" as a hint what is true intent here? |
|
A good point raised, barring the personal jab at the end - 1/10 may be a bit excessive, slightly less extreme change gradually adjusted till it'll be 'just right' may be safer. |
|
"no-meat" means that a character without survival skills and propper infrastructure will barely be able to get fresh meat, until he will occasionally kill a moose with his deathmobile or a pack of wolves will find his base. I want hunting to be more similar to Unreal World, where it takes few days to track a deer, but his meat will be enought for a season. Trapping, for example, is one of the most important skills for surviving, but is useless as long as we have annoying animals everywhere. |
Great, how about nerfing current spawns after these ideas are implemented. Not the other way around. Not to mention that it is currently not possible to sneak upon non aggressive animals. That should be implemented first as well. |
|
I like the general idea, as I do not like the overall amount of wildlife either, but I have to agree it's done the wrong way around... There should first be a way to reliably track and hunt the game, before you reduce it this much... |
|
I dunno, 1/10th seems about right as right now you can't cross between two cities without getting attacked by coyotes repeatedly and having to kill 15-20 of them. No need for hunting, the meat will come to you just two steps away from your base. |
Here it may be more of an AI behavior than animal abundance. Plus, I've noticed far fewer animals in forests than in swamps so cutting it all to 1/10th all across the board no matter the initial group size may be indeed a bit too crude measure. |
|
Only seems to be a swamp spawn problem IMO too. Not seen in game numbers, but certain swamp tiles do seem to get a monster splosion (lore/gameplay challenge?), whereas forests just get normal spawn numbers. Swamps also get normal spawns, but there is certainly something causing the occasional epic battle of wildlife! |
|
This needs a goal for what density of animals there should be, moving from one person's "seems ok to me" to another person's "seems ok to me" doesn't help us. |
|
I don't have a source specifically for New England but density could be estimated from an average size of territory a pack (or a solitary animal) defends. Encyclopedia Britannica:
Westernwildlife.org:
Wolfworlds:
How much square kilometers/miles is one map tile? |
|
Given that we have breeding and similar in now, this makes perfect sense to rebalance things around it. Another idea just to throw out there: |
One house-sized area, which is one symbol/tile on the overmap view is about 24mx24m. |
|
And a single overmap region is 180x180 of these, meaning it is a square of 4320 meters, or about 18,6 square kilometers. Which seems to suggest that a 180x180 region should contain about one wolf pack, one coyote pack, and 3-4 bears. |
|
How does this fit with city density in the area though? Cities are much smaller than RL, and their distances are much closer. Condensed, if you will. I am not sure if it does make sense to pull one aspect (animal density) and change that to RL value while ignoring the rest. |
|
Prey animals, in general, are more densely-packed than predators. This abstract offers some insight into how big a moose's home range is: |
|
I've always imagined those "cities" were not really cities, but minor settlements and cottage groups. The whole overmap looks like a suburban (big city size) or rural (small city size) settled area to me. I have never been to New England, of course, but where I live it's pretty common to see small groups of cottages every 5-10 km, denser along the highways, railroads and rivers. If these are condensed, then not by much. Though it does surprise me that one tile is only 1 meter wide. I mean, I can cover more than 1 meter in 6 seconds. |
|
I propose that we work off the travel times, e.g. a pedestrian covers 4320 tiles in ~26000 seconds, so if we assume 1 meter per second then we should calculate wildlife densities as they would be for 26x26 km square overmaps. That would suggest 20-30 predator packs (wolf, coyote) and 40-60 of each large solitary animal (moose, bear) per full overmap square. However, this does not yet help work out what numbers should be in the json. E.g. I need to know what exactly the following line means:
UPD: okay I think I understand how it works, so gimme a day or two and I'll do the math and post my findings. (In the name of !!SCIENCE!!) |
How would you explain the size of the farm then? Average USA farm is 442 acres, or 1.8 km^2. So it should be 56x56 map tiles... but it is not. |
|
See above, I suggest that we work from travel times instead -- the farm, incidentally, would be about the right size then. |
So, condensation factor of 6 based of travel times then? Sounds reasonable to me, this is something that Kevin should approve of first though. |
|
Discussions about how large is one tile never ends good. It has size from 0.5 to 25 meters, depends on measurment system. For example a pedestrian walks 1 tile per 6 second and a car at 16 km/h rides 1 tile every 6 seconds. 26x26 km square overmaps sounds good, but recalculating animal spawn to this size will (probably) reduce overall population even more then by 10. |
|
Which is fine as long as the numbers have a basis. |
I agree, we simply dont have "cities".
Scale was never really considered, I want to shrink turn length to 1 second, which rationalizes move speed and walking distances nicely. I'm not against scaling this based on move speed. Amount of ground covered in a day is a major limiter on hunting effectiveness. If we later adjust effective move speed, we can adjust densities again then.
Large features are not scaled properly, in large part because they're harder to make, and the game engine doesnt support them well. Overall I'm fine in principle with taking real life population densities as a goal, if that makes hunting tedious or prohibitively difficult, then we might not jump straight to that density, but instead go part of the way there and look into adding features that bring player time investment and resource balance to where it should be. |
|
I just realized that we can't take RL density value of specific animal, like wolf, and use it in the game, but rather density of animal class, because we don't have every animal represented in game. This would lead to lower than RL wildlife animal density otherwise. |
|
If we assume one overmap (180x180 chunks) to be 26x26 square kilometers, the area of one chunk will be 20k square meters. Average ingame chance to get a moose from This gives moose reduction ratio 6, as it was suggested before. Any thoughts? |
|
For comparison: a realistic scale (overmaps at 4320m x 4320m, or about 18,6 square kilometers) would require reduction by a factor of 216, not 6. But 6 would be a good start. Seriously, wildlife is so plentiful currently that it shatters suspension of disbelief. UPD: |
|
@Marrim |
|
Average animal spawn rate is decreased by a factor of 6, as discussed above. In this case most animals will have 20 km2 home area. More precise tuning can be done later by edit monster groups. |
|
It can be detrimental to use home range as a measurement for animal density as this does not take into account pack/herd size and overlap between home areas. It would be better to use density numbers as a measurement as they already include these considerations. Sweden for example has one of the strongest moose populations in the world with about 1.2 moose per square kilometer of forest (source). Canada has a lot lower numbers at 0.267 moose per square kilometer of forest even though they both have the same home range. The reason given in the linked article is that the prevelence of predators is different between the different countries. Given that one chunk is roughly 20k m2 or 0.02 km2; a swedish moose density would be Density numbers from the above referenced article (translated from moose per milion hectares to mose per km2):
|
|
Good discussion everyone, I'm convinced this is a good first step. |
|
@zheele |
Summary
SUMMARY: Balance "Reduce group spawn of wild animals."Purpose of change
Forests and swamps are overpopulated with wild animals. Such high concentration of predators and preys is possible only in controlled conditions like ZOO. Most wild animals must avoid humans and run away before contact.
As the result hunting is trivial and finding fresh steak is not a problem.
Describe the solution
Decrease spawn of common animals in forests and swamps.
It will decrease chances to meet a group of earth animals and their zombified versions. Ants, triffids and jabberwocks will not be changed.
Additional context
As a long-term idea I want to see animal groups to move similar to hordes, but they will avoid noises instead of being atracted to. Characters with high survival skill will be able to see them on overmap view.