Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
MMORPGs and simulated ecology... Futile?
Waffleton
Member Posts: 41
As both an MMORPG enthusiast and an ecologist, few things annoy me more than a poorly simulated ecology. Most MMORPGs today have ecologies that consist of "polar bear respawns here 5 minutes after it is killed, for the sole intention wandering around aimlessly until a PC gets to close". This totally breaks immersion from me. I personally would like to see an MMORPG that simulated a realistic ecological environment; the different species of animals hunt each other, forage for food, mate, etc. Overhunting can lead to plummeting animal populations and even extinctions. Plants and trees use various methods of seed dispersal to grow over time in different climates. Entire forests can become deforested, requiring the player population to constantly migrate for resources as they become consumed.
Richard Garriot, creator of UO, once claimed that a realistic ecology in an MMORPG is futile. I do not share this belief, I think that if it is done correctly, it could add an entirely new layer of immersion. I am tired of walking by to see dozens of different animals stand idle next to each other until I get too close. Why in the world didn't that deer run away from me as I approached? Why is that bear completely ignoring that panther? In current MMORPGs, wild animals are designed specifically for interaction with players. I personally would prefer to see them interact in a realistic ecological setting. Anyone's thoughts?
Comments
Its futile because alot of people don't play for realism, and they want to kill things mindlessly. It might be an innovative breakthrough, but a flop in practice. I wouldn't play a game like that. Well maybe in the beginning, but everything WOULD be overhunted, because many players wouldn't care, or would do it on purpose just to see what happened.
So yeah, I dont think it would work in practice.
Ryzom has a neat ecology system though.
It's futile because it's completely griefable. Someone will kill every animal and chop down every tree.
Then what do you have? Desolation. There is nothing fun in that vision.
well if this would happen then the game would be sold out before it hit the shelves and im not joking around when i say that
lets see here
wouldnt it be rather difficult to do this interaction with other animals
like for experience and whatnot
say two animals are fighting and their health is low and you come own them both what happens with experience there?
its a really good thought when it comes to mind but really how are they going to do it?
so many possibilities
what if the animals produced somehow???
the trees grew back
what if there were seasons?????
there are so many things that could happen and its overwhelmiing
I think it's an intriguing idea.
I wouldn't center my game around it, but it would definitely serve to catch the eye. This is what MMORPG developers need to be focusing on... innovative ideas of this sort to make their games seem more for the 'next generation' of gaming. I think 2005 will be that fine line that splits games like WoW and the more innovative games that will begin to emerge. Right now we're in the age of 'refining' age old features... late 2005-2006 will see a burst of games that advertise never before seen features, imo. An expansive ecology, more complex economies, etc... should all fall under this category.
In the frame of a persistant online world the eco system could perhaps grow and evolve on its own, to a certain degree. Players could manage it(be a forest ranger), hunt, or it could just be eye-candy for the main game.
Tacklebury --}>>>
In a large enough world, it could take a lot of work to destroy an entire population of plant or animal. With different continents on a gigantic globe, such an undertaking could take months or years.
With a realistic ecology, we could see players who specialize in animal husbandry, taming different animals as a source of food and resources, as well as players who specialize in hunting (using skills to track certain animals and kill them). As certain animals become more and more rare, players could see all sorts of creative developements, such as big game hunting and valuable animal pelts, etc. A realistic ecology directly supports a realistic economy. Instead of "farming" sheep by killing them over and over again as they spawn, you could actually "farm" the sheep, by taming them, breeding them, feeding them, and eventually slaughtering them for food, etc.
You know, I was just thinking of this the night before last myself. It would take an system I can't even fathom to do it, but I'd love to see things like this!
I'd like to see animals occassionally cross-breed, and for new types of animals, same with plants. I would love to see what you mentioned, with the populations. That would be amazing.
I would LOVE to see this type of thinking put into game, but I have the feeling that these things might be a little ways off, after a bit of technological advances.
But man, do I love the feeling of a simulated ecology!
I really hope that *insert game name here* will be the first game to ever live up to all of its pre-release promises, maintain a manageable hype level and have a clean release. Just don't expect me to hold my breath.
I would expand this to not only include ecology, but economy as well. I think ecology/economy are often poorly modeled for two reasons: first the people making games know how to write code and make 3d models and textures and so on. When's the last time an MMO maker had an economist on or ecologist on staff?
I remember looking at the UO box back in 1997 where it described a vibrant player-driven economy. Needless to say the economy was neither vibrant nor player driven. Garriot had some innovative gameplay features with UO, but a good economy wasn't one of them. I majored in math in college, but i took numerous economics courses and i am not very impressed with most economic models in MMOs. Being an ecologist i can understand your frustration with the poorly represented ecology in MMOs.
The second major reason ecology is poorly done in MMOs has to do with technology. From both a processing power and connection standpoints. Making animals that aren't mindless mobs requires good AI techniques. AI is notoriously processor intensive and the last things MMO makers want to do is make smart monsters that bring the server farm to a standstill. AI is improving, but there is still alot to be desired. I have seen isolated instances of the wildlife interacting in WoW. While i'm quite unimpressed with WoW overall, i have seen wildlife attack each other on more than one ocassion. As far as the "infinitely respawning monsters" go, i couldn't agree more than camping the monster spawn location is neither realistic nor fun.
The other bottleneck with making interactive environments concerns internet connections. Basically all current MMOs target 56k modem users. By doing this, however, they severly limit the amount of content they can fit down the conneciton pipe. Many people can't even connect at 56k and a more typical connection speed is 28.8kpbs. That's kiloBITS per second. Converting that to kiloBYTEs it's about 3,600 bytes per second. This might sound like alot, but it's not. take for example that for each game object the server must send to the client things like it's X,Y,Z location, it's orientation, object state, direction of travel, hit points, any coverings (like clothes), and so on. The X,Y,Z coordinates would likely be stored at unsigned 32-bit integers. That's four bytes of data for each coordinate. let's say we could store each of the other attributes i mentioned in 1 byte each. That's 4x3 + 5 or 17 bytes of data for EACH game object.
So to send the current object state for a single game object (whether it's a tree or a grizzly bear or another player) takes 17 bytes of data -- please note that i'm over simplifying the math here a bit because the data would actually be greater because of TCP overheads and coders would actually try to compress the data as much as possible. Nevertheless for sake of argument we'll use the 17 bytes of data for each object. Now, how many times a second should the server send the data to the client? Once a second won't be often enough. Just imagine how much players/monsters/objects would "jump" around on the screen if you only got their data once a second. Thus the data must be sent more often. Ideally the client would get each objects state once for every animation frame (say 60 times a second for a 60 frame-per-second frame rate), but this is also unrealistic.
So let's say we will send our object data to the client 10 times a second. This will mean that we won't have up-to-the-second info for every single animation frame, but it should be enough that the user won't detect any lag. Thus we have 17x10 or 170 bytes of data per second being sent from the server to the client. No problem you say since our modem will deliver 3,600 bytes per second. So we have lots of overhead right? Well if the server only needed to send state information for a single object then yes there would be alot of overhead, but what about all the other players nearby not to mention all the monsters roaming about. Doing the math we can see that our connection will hold approximately 3,600/170~21 object states per second. More than that and it overwhelms the connection.
This is where making interactive environments becomes a problem. It's bad enough if you are in an area with many other players and some monsters. In WoW with 5-player instances plus maybe a dozen monsters nearby that's about all the connection can cope with.
This doesn't even begin to address things like trees that can be cut down or burned, or that might regrow. In order to enable such gameplay, each tree has to be programmed as an object (this is in direct contrast to how it's currently done where trees are just pre-compiled parts of the landscape). Want to run through the forest with your friends chasing the black dragon? Now the server must send the object state for each tree (and for everything else you would like to model in the environment) and you could see that a handful of players + 1 black dragon + hundreds of trees isn't going to work over an analog phone modem.
This isn't to say it can't be done. What i am saying is that it can't be done while trying to retain players on 56k modems. If an MMO maker is willing to make a broadband only MMO, and pay for the additional bandwidth such a game would require it could work. In addition to the additional servers needed to perform all the logic, then yes such an interactive environment can be made.
MMO makers want to capture the entire market, but in doing so have to design for lowest common denominator. This, IMHO, holds back game design. People will endlessly complain that an MMO should not be make for broadband only, then again there are those who would also argue that games should be make to run on pentium 133s or even 486 computers. Sure it can be done, but you give up alot to make it run on everything. At some point you have to break with the past in order to reach the next level of development.
www.TheChippedDagger.com My 90-day 2D Java MMORPG project
They that can give up essential liberty for temporary safetey deserve neither. -- Ben Franklin
If opportunity doesn't knock, build a door. -- Milton Berle
In a nut shell, what I finally realized is that, the economy system in every single mmorpg is never perfect.
Realism and fun are not mutually exlclusive. It's just like the people who say gameplay is more important than graphics. In reality both are very important. Realism can add to gameplay and fun, it can also detract from it. The acid test should be wether adding a particular gameplay feature adds to it, not where it makes it more realistic.
I can understand that for you it's all about mindlessly killing mobs. That's great and there are plenty of games out there that cater just to you. Please try to understand, however, that there are lots of players who want a healthy dose of stragety in their games. To many gamers mindlessly hiting the fire button on their BFG 5000 becomes boring very quickly. Many of these players welcome a much deeper gameplay experience. It's not that one is right and the other is wrong, but they are different types of games and different games appeal to each.
www.TheChippedDagger.com My 90-day 2D Java MMORPG project
They that can give up essential liberty for temporary safetey deserve neither. -- Ben Franklin
If opportunity doesn't knock, build a door. -- Milton Berle
www.TheChippedDagger.com My 90-day 2D Java MMORPG project
They that can give up essential liberty for temporary safetey deserve neither. -- Ben Franklin
If opportunity doesn't knock, build a door. -- Milton Berle
All the more reason MMOs shouldn't have experience points or level systems. Skill by use is the way to go. MMO makes seem to nearly always try to import a tabletop D&D game into an MMO then can't figure out why it doesn't work quite right.
www.TheChippedDagger.com My 90-day 2D Java MMORPG project
They that can give up essential liberty for temporary safetey deserve neither. -- Ben Franklin
If opportunity doesn't knock, build a door. -- Milton Berle
www.TheChippedDagger.com My 90-day 2D Java MMORPG project
They that can give up essential liberty for temporary safetey deserve neither. -- Ben Franklin
If opportunity doesn't knock, build a door. -- Milton Berle
"This doesn't even begin to address things like trees that can be cut down or burned, or that might regrow. In order to enable such gameplay, each tree has to be programmed as an object (this is in direct contrast to how it's currently done where trees are just pre-compiled parts of the landscape). Want to run through the forest with your friends chasing the black dragon? Now the server must send the object state for each tree (and for everything else you would like to model in the environment) and you could see that a handful of players + 1 black dragon + hundreds of trees isn't going to work over an analog phone modem.
This isn't to say it can't be done. What i am saying is that it can't be done while trying to retain players on 56k modems. If an MMO maker is willing to make a broadband only MMO, and pay for the additional bandwidth such a game would require it could work. In addition to the additional servers needed to perform all the logic, then yes such an interactive environment can be made."
I don't know much about programming, but why would you update the trees status that often? They're pretty static, they grow very slowly. They would only require network updates when a player enter that world(current status), and when a player somehow 'interacts' with a tree, like cutting it down. Maybe an hourly update on the growing process(to swap it for another scale/model), not much more than that.
But yeah, the glory days of the 56k modems seems to be over anyway.
truer words were never spoken
I really hope that *insert game name here* will be the first game to ever live up to all of its pre-release promises, maintain a manageable hype level and have a clean release. Just don't expect me to hold my breath.
Pinkdaisy,
I like your vision for MMO's. I wish more game developers had thoughts like that entering their mind...
Dont worry. We do.
The problem isnt thinking up the content or even producing it...its the time that it takes to produce it.
DragonEmpires was going to have a working ecology..but we all know what happened to that team.
There key for making such ecologies is three-fold:
a) make a hierarchy of a foodchain
b) give each group of animals a group of goals and needs
c) give each group of animals a 24-hour schedule with game-realistic behavior
You cant make creatures that dont respawn out of thin air...but you can allow some rules for replacing them which simulate a dynamic effect on the world. That is to say if you kill lots of creatures they will move to new territories that will suit their goals and set up homebase again.
Another thing that needs to be considered is "world processing power" - that is to say that some spawn groups may be reduced to offset server lag.
Oh well. I digress. Definitely a feature I see coming further in the next few years.
Althought I mostly support Garriott point of view here...Instanced realism can be fun, I did this so that happen.
Enforced realism based on others actions is a sure way to piss me and grief the game and everyone in it if I keep playing, until a GM will ban me.
Auto-Assault will remember the damage you do to the environment, yet only the damage YOU do...that can be fun, it may be an interesting way, yet some new stuff need to developp or decay regardless of my actions, my actions need to only be a small part of it(unless I work hard on something to change the ecology for exemple, like killing all X mobs, then neighbors mobs or those I implement may prosper). But that is hard, not having the same mobs for all the players is certainly a puzzle for the devs.
Realism, like everything else, is FUN, only when it is an option, not an obligation.
- "If I understand you well, you are telling me until next time. " - Ren
You are absolutely right. In my attempt to try to keep my comments short i skipped many of the details: one you obviously pointed out. You could send state information continuously for all objects near the player, but that would waste alot of bandwidth and processing power. A much better approach is to only send state information for objects whos states have changed since the last time the player (client) was updated.
I was showing worst case scenario (but failed to mention it). Infact much of the time the bandwidth will be considerably less. In some cases there will be virtually no object state information being sent to the client -- consider a player standing still in an area with no other players nearby, no monsters in the vicinity, and no trees catching fire.
Unfortunately we must design for worst-case scenario because those are the situations that will cause undesirable gameplay experiences for the user. Consider a player running through a forest. Say the last time the player passed through the trees were green with leaves. Suppose since then an evil wizard cast a fireball and set the trees aflame. If the state information saturates the connection, the player could be well into the forest before the client gets the updated info that the forest is onfire. To the player running through the forest the trees would go from green with leaves rustling in the wind to a burning inferno instantly! Obviously game designers would want to avoid this situation. Although this example would be more of a graphical glitch than a gameplay one, the same situation could easily translate to other players or monsters.
Thus if the game will function under the heaviest loads, then it should function at all points in between. That's why i only considered worst-case scenario. Sorry for the confusion.
Regards
www.TheChippedDagger.com My 90-day 2D Java MMORPG project
They that can give up essential liberty for temporary safetey deserve neither. -- Ben Franklin
If opportunity doesn't knock, build a door. -- Milton Berle
Yeah, the player interaction with the trees (like cutting them down or setting them on fire) should be instantly sent to the server, and then back to all the clients in that world. Maybe a burning tree could be a local client object effect or animation, triggered by the server once every minute until the tree is totally destroyed? Maybe a small fraction of the networking code every second could be reserved for these effect/animation controls? The hourly update is only useful for growth and other rare events.
Like you mentioned, it's gonna be heavy on the server processing too, with maybe thousands of animals 'living' on the server. Each with their own advanced AI. Well it would be fun to one day see something like that, i'm sure it could be useful in several types of games.
I'm still worried about griefing, I don't believe the world could possibly be large enough to protect players from wanton wood-choppers. Or, here's a thought. Lets say I'm a sheep farmer. Ok, if this is going to be realistic I'm going to need a Sheep-Farm (heh). Well, figure a pen for my sheep and a hut. Then I let them wander around during the day.
Lets suppose that this sheep-farming sim is a lot of fun, I'm not sure how long it'd stay that way but lets just suppose that it is. Ok so now I'm sort of bound to this flock of sheep I put so much time in on and the area in which they like to live and their sheep-pen.
What if some asshole decided that sheep farmers were "n00bs" and cut down all the trees nearby? I mean, I've got to take care of my sheep but lets say I need wood to repair my house or make furniture for it or to maintain my sheep-pen? What if I just broke my Shepherd's crook and need a new one? I mean, what am I going to do now? Wait for someone that is not bound to a flock of sheep to travel off into the wild blue yonder to find a tree?
How much will it cost sheep-farmers to buy wood? I mean, this guy isn't going to travel however long it takes across this giant world to find where the griefers ran out of steam and there are still trees for free is he?
With what I said earlier, I don't think I'd have fun playing Sim Sheep Farm for very long. I mean, yeah, its not a bad idea but it sounds a lot more like work than gaming. Like grinding that never goes anyway except that my number of sheep grows ever larger? Would I be able to sell my flock when fighting off wolves got boring?
For economies, I honestly don't think that they would be worth modeling. They are enormously complex systems. In a matrix they would have hundreds if not thousands of entries. In a fantasy setting, you'd have rules that are bizarre by modern standards if it was set in the past (actually if it was set in the past, virtually all players would be serfs with no money anyway, that would simplify things). How would you model the future? Extrapolation would be fairly outrageous since the economy would change in huge ways over any period of time more than a hundred years. Even if you did extrapolate, you'd end up with something even more unsupportable complexity.
There are a few examples of what could be done to aquire wood in your scenario. Assuming all wood relatively near your pen has been forested, one simple solution is to find a player who specializes in the gathering of wood, a lumberjack, if you will, and purchase your desired resources from him. Perhaps a good place to seek such a dealer would be at a nearby market establishment, as it is inevitable that players will erect their businesses in certain populated areas to hawk their goods. No good marketplace nearby? Then you may have to do some traveling. It may be a good idea to hire the assistance of a mercenary player to protect you on your journey if the route is poorly gaurded.
But where will you get your money for the lumber? Well, seeing as you chose the path of a humble shepard, one likely answer would be in the selling of your sheep's wool to tailors and merchants, or selling their meat (one can only live so long off of foraged berries!). Of course, if you do not find the tending of a flock very fun, there could be hundreds of other options for obtaining wealth. Sell yourself as a hired arm! Become a traveling merchant! A blacksmith? A mason? A tailor? hell, even an inn-keeper. I would find any of these options more fun and immersive than choosing from one of several fantasy classes (druid, barbarian, etc.) at the beginning of the game and painfully grinding them through meaningless levels by nothing but combat.