AI when it comes to animals I'd like to see a breathing world without the need of the player to discover it. Animals hunting, plants growing while animals eat them, etc.

Simulations are not games, so integrating a simulation in a game is always a nice idea, but it brings a very heavy burden with it. You can compare it to hollywood movies, you don't need to simulate an whole ancient city with a million population to make 3 scene shoots in it.

The problem is, that the player would not recognise it as simulated world, there's even the danger of missinterpreting emerging behaviour as bug. L4D is a great example of a virtual populated world which is only simulated close to the players.

Your correct in your response, saying that a true simulation would almost be irrelevant because of the burden if it was an RPG. L4D is an example, but if say within 30 feet AI interactions would coincide with objects in the world or other AI animals in the wilderness would be ideal in my opinion.

Skyrim would be a good example of what im talking about. Just more AI interaction with the environment along with other AI.

Simulations are not games, so integrating a simulation in a game is always a nice idea, but it brings a very heavy burden with it. You can compare it to hollywood movies, you don't need to simulate an whole ancient city with a million population to make 3 scene shoots in it.

The problem is, that the player would not recognise it as simulated world, there's even the danger of missinterpreting emerging behaviour as bug. L4D is a great example of a virtual populated world which is only simulated close to the players.

Car racing and flight sim fans would care to disagree. I agree that granular simulations are usually surplus to requirements. However I do think that simulations with a degree of LOD could be worthwhile. For example, simulate animals nearby in a granular way, and for other areas simply simulate broad predator/prey dynamics and migration patterns. I agree that emergent behaviour can be tricky. Perhaps truly emergent behaviour is not desirable, but rather representative set pieces, e.g. if a herbivore runs from a carnivore the carnivore should always be coming from a direction that the player can see. But overall I would welcome some sort of nature simulation as a source of new types of "random" encounter and to avoid that creepy "all these creatures exist only so you can kill them" feeling.

I think the real question is not how to make wilderness interesting (because it is trivial, just make a lot mini secret locations with secrets and quests and hidden treasures) but how to make it interesting in a cheap way (because you want to focus your development effort on the locations, not the wilderness).

The biggest thrill of wilderness travelling is discovery of secret locations, so I guess there is no way artound it, you need to make at least a few unique handmade mini locations without telling the player the coordinates of these). Everything else should be generated randomly.

There could be:
- 10 old bunkers (shape "generated" cheap way by putting together 3x3 sections made of 6 predefined sections), each holds a standard random treasure and 1 old electronic chip (3 of these bunkers will have broken chips), when ypou collect all working chips you can build something out of it. The exact coordinates of these bunkers changes every game and are totally random
- many hidden caches with random resources, come in 3 versions: small, medium, large.
- there are 20 parts of machine X, generated randomly among the bunkers and caches above as a rare treasure (and very rarely could be found just plain on the ground), when you collect 5 you can build something awesome
- rare plants that can be processed into something (so plain "found something" thing)

I also recommend checking "Barbarian Prince" and "Tales of the Arabian Nights" boardgames, these had a lot of randomly generated things with auto generated secrets and storyline.

[quote name='Ashaman73' timestamp='1341983638' post='4957893']
Simulations are not games, so integrating a simulation in a game is always a nice idea, but it brings a very heavy burden with it. You can compare it to hollywood movies, you don't need to simulate an whole ancient city with a million population to make 3 scene shoots in it.

The problem is, that the player would not recognise it as simulated world, there's even the danger of missinterpreting emerging behaviour as bug. L4D is a great example of a virtual populated world which is only simulated close to the players.

Car racing and flight sim fans would care to disagree. I agree that granular simulations are usually surplus to requirements. However I do think that simulations with a degree of LOD could be worthwhile. For example, simulate animals nearby in a granular way, and for other areas simply simulate broad predator/prey dynamics and migration patterns. I agree that emergent behaviour can be tricky. Perhaps truly emergent behaviour is not desirable, but rather representative set pieces, e.g. if a herbivore runs from a carnivore the carnivore should always be coming from a direction that the player can see. But overall I would welcome some sort of nature simulation as a source of new types of "random" encounter and to avoid that creepy "all these creatures exist only so you can kill them" feeling.
[/quote]

I completely agree with you jefferytitan, The fact that sabre tooth cats aren't running around chasing foxes running across the screen instead they go after me is a little, how you say, boring.

I think the real question is not how to make wilderness interesting (because it is trivial, just make a lot mini secret locations with secrets and quests and hidden treasures) but how to make it interesting in a cheap way (because you want to focus your development effort on the locations, not the wilderness).

The biggest thrill of wilderness travelling is discovery of secret locations, so I guess there is no way artound it, you need to make at least a few unique handmade mini locations without telling the player the coordinates of these). Everything else should be generated randomly.

There could be:
- 10 old bunkers (shape "generated" cheap way by putting together 3x3 sections made of 6 predefined sections), each holds a standard random treasure and 1 old electronic chip (3 of these bunkers will have broken chips), when ypou collect all working chips you can build something out of it. The exact coordinates of these bunkers changes every game and are totally random
- many hidden caches with random resources, come in 3 versions: small, medium, large.
- there are 20 parts of machine X, generated randomly among the bunkers and caches above as a rare treasure (and very rarely could be found just plain on the ground), when you collect 5 you can build something awesome
- rare plants that can be processed into something (so plain "found something" thing)

I also recommend checking "Barbarian Prince" and "Tales of the Arabian Nights" boardgames, these had a lot of randomly generated things with auto generated secrets and storyline.

I don't think you got the point of this, me personally I get bored of the same old stupid ruins I come across of, or the fort. I want to focus generally speaking on the wilderness. I would like to see more diversity in the wilderness

and a lot more interaction with everything in the wilderness. The forts, secret locations, hidden treasures are all fine, but I want to see a real living breathing wilderness around the player

Consider the pacing of the gameplay as well. Pretend you are the player moving through that wilderness the first time, is it important that the player relax after a hard fought battle, build tension towards an intense encounter, experience something new (because the games been pretty quiet), learn something useful for the area they are headed to, etc? These factors will help decide what the reason for the gap between locations is an how big it should be. Now pretend you are the player who's been grinding for 4 hours in that god forsaken wilderness. The sections between player way-points can be very useful. Be the player, test it in your mind before you start filling it with assets. I don't know if I agree with removing it because travel and discovery are important aspects in an RPG they help the player prioritize objectives, number crunch, let the game and the OMG moments sink-in.

Explore the weather and the seasons. All wilderness experiences incredible things worth showing off from climate changing. So many RPG's have a static world. Plants move and change,animate them, they deserve it! And animate them for a reason. Have encounters with bad guys that blend in with the forest in fall, or grow hardier during winter, etc.

Hope this helps.

Consider the pacing of the gameplay as well. Pretend you are the player moving through that wilderness the first time, is it important that the player relax after a hard fought battle, build tension towards an intense encounter, experience something new (because the games been pretty quiet), learn something useful for the area they are headed to, etc? These factors will help decide what the reason for the gap between locations is an how big it should be. Now pretend you are the player who's been grinding for 4 hours in that god forsaken wilderness. The sections between player way-points can be very useful. Be the player, test it in your mind before you start filling it with assets. I don't know if I agree with removing it because travel and discovery are important aspects in an RPG they help the player prioritize objectives, number crunch, let the game and the OMG moments sink-in.

Explore the weather and the seasons. All wilderness experiences incredible things worth showing off from climate changing. So many RPG's have a static world. Plants move and change,animate them, they deserve it! And animate them for a reason. Have encounters with bad guys that blend in with the forest in fall, or grow hardier during winter, etc.

Hope this helps.

I always thought it would be cool project to create an game environment with interactive wilderness like your saying, then building a game into it. I mean, first building a world that is very immersive and interactive. Then making a game world within it, even if only a few players at a time can enter it. It would be a great learning experience don't you think?

Wilderness is (apart from crafting) another thing that Ryzom got just right (well, in my opinion, given the game's success, the majority of people didn't seem to think that way).

Most things in "wilderness" (especially in the Roots) was strong enough so you could survive being attacked by one, if you were high level and skilled maybe two. Most animals would group together somewhat loosely, sometimes one would maybe wander off a bit. Also, the groups would move around according to daytime and season. This usually left a "clear path" for the skilled player to move through, a big challenge in some regions.

Predators would chase herbivores, but this would inevitably bring them into your aggro range sometimes. Also, some herbivores seemed to run towards you (for protection? who can tell?). Nothing as entertaining as seeing a lumper being attacked by 3 kinchers, and then it comes running to you -- with the three following.

The downside is of course that it isn't precisely an easy entry, nor does it cater the typical crowd of players who are used to only click on the "win button". That's probably the reason why every company that went near Ryzom went bankrupt