Believability

[Ed: Posted on behalf of Anders Tychsen.  See also http://www.rpgdot.com/phpBB2/viewtopic.php?t=176736]

During a recent analysis of the relatively new Diablo-clone Sacred, it struck me that I felt absolutely no suspension of disbelief while playing. It did not bother me too much that I had to slaughter hundreds of creatures that wandered around aimlessly, seemingly without any purpose except to kill my “overfeatured”, bikini-clad avatar. What ruined the experience was the fact that hordes of goblins would jump me whenever I got within sight, but stay away from e.g. an unprotected farm. They would even follow an avatar into the farm, without ever considering the farmer a viable target. Later on in the game, a 10 ton dragon attacked using wings that was seemingly attached with duct tape, ignoring the demands on the body structure it takes for an organism of that size to have wings, or even use them in a 1G environment.

Such discrepancies in games are not uncommon, and are somewhat alleviated by the willing suspension of disbelief of the players when they enter a game world. The suspension of disbelief is however ruined when something occurs, within the physical parameters of the virtual world, that is obviously not realistic. While a “Plane of Fire” as physical environment is perfectly acceptable to most, having a creature spawning out of thin air every 30 minutes in an MMO, for the sole reason of providing XP for the grinder, is not (e.g. Ultima Online). It is not even realistic within the fiction of the game world, and ruins the suspension of disbelief. If such problems could be avoided in a realistic, believable (within the game fiction) and consistent manner, games would be more immersive and the believability greater - my all-too-obvious and age-old point here bei! ! ! ng that some features we can accept within the parameters of the illusion, others break the illusion.

Added realism does not mean eliminating any fantastic elements, but adding depth and dimension. For example, a fantasy world might feature magic, and such a pseudo-argument can be used to explain why a 10 ton dragon can fly. However, it does not justify why the dragon has to spawn every 30 minutes at a given place. Thinking about the ecological setting of the world, we might be able to come up with a better system for the appearance of the dragon, which is plausible. Also, if biophysically a bit better modeled, the dragon might even feel more realistic.

Not all creatures of all games should be like real world organisms. Not everything in a virtual world should be explained using the laws of science. However, various features of the real world can inspire, and be directly utilized, to enhance virtual worlds by adding realism. This is an old, worn-out discussion, but I will try to give it a new twist by proposing that by applying natural sciences theory and techniques, a range of problems in games could be solved. Not only can the natural sciences be used to solve specific issues, but they also offer a way of completely changing the way that the virtual worlds themselves, and its related mobiles, act.

I see at least two venues for this type of thinking: 1) Physical world modeling (including ecology) and: 2) Biology, biophysics and behavior (ecological relationships) of creatures. I will here focus on the latter subject:

It is quite obvious that MMOs do not emphasize ecological realism - just think about predator-prey relationships. It is common to walk around in a game world meeting nothing but ferocious, meat-eating monsters (intelligent or otherwise). It is also common for mobs in a game to feature only one of two types of behavioral patterns: Either they are killing machines, intent on the destruction of the players, or they are harmless critters, whom the players have virtually no way of interacting with.

Unless explained within the fiction of the world, this is not very realistic and hurts the suspension of disbelief. Even a cow will be angered or frightened if stabbed or shot by a player, and its frantic mooing may bring down the wrath of the nearby 1,500 pound bull. In short, the ecological setting is unnatural, and the behavior of the organisms is one-dimensional.

With a bit of programming effort, ecological principles could be used to generate a more viable ecology for a given virtual world. In the short term, ecological and biological principles could be applied to specific problems, such as modifying the kill-monster-for-xp-system with the introduction of realistic predator-prey abundances. This also needs to new types of challenges, e.g. environmental management.

To give a primitive example of ecological thinking, take spawning and camping: A dragon with substantial treasure spawns every 30 minutes, leading players to camp around the spawn point. The players know that the spawning is a control measure, with no in-game explanation for its appearance given. This breaks the illusion. 

The problem is that with the XP and leveling system being the way it is, it is necessary for the players to have a certain amount of mobs to handle in each time frame. How could we make sure that there was a dragon there every 30 minutes of game time, or a similar challenge, while maintaining a plausible explanation? The answer could be migration. The dragons have a nursery in a mountain nearby, an inaccessible mountain where adult dragons flock, spawn and rear their young. Every once in a while, the young leave the nests, and seek out their own paths in the world. This means that every X time – which could be randomized a bit – dragons would migrate to the surrounding area, giving the players the challenge they need. This means that there may not be a dragon spawning every 30 minutes, but 10 migrate into the territory every 3 hours. The XP-effect is the sam! ! ! e, and players knowing the migration routes can still camp out. But now there is no spawning, there is a reason why the dragons are there, which forms the basis for more varied quests. There is no need for instanced worlds, either. By adding some fuzzy logic, the areas where the young dragons settle could be varied (as in nature), as could the intervals – e.g. as a figure of the time of year – and thus camping becomes not viable. Players would have to walk around a bit to find a dragon, work out any migration routes, or maybe rumours would spread when it was known where a flock of the young dragons had settled. This takes a bit more programming than the spawning solution, but might it be worth the effort?

More advanced uses for the natural sciences could be ecological modeling, population statistics and prediction of player behavior, simulation of weather effects on biological organisms, seasonal variations/migrations, biophysically guided modeling of creatures and even evolution of organisms:  Imagine an area where creatures evolved rapidly according to adaptive theory: Use fireballs and they evolve to compensate, until random mutation removes the adaptation strain, alters it, or the players eliminate every mob with the mutation strain. The end result is constantly challenging opponents.

To sum up, the question is whether biological/geological/etc. thinking has a place in games and if so, whether the subject is of interest in lieu of the cost-benefit thinking in the industry. I believe that the gains of working on integrating elements of the natural sciences in games could be huge, but I lack the experience to determine whether this is correct – which forms the basis for posting these musings. What do you think?

Anders Tychsen

Comments

[Cenn]

1.

I think its awesome - and something that is needed i feel to turn a button mashing game more into a RPG - a game where we are supposed to be pretending to be citizens of an imaginary world.

[magicback]

2.

Great idea, hard to implement, and hard to maintain ecological balance from the get-go.

Better to stick with theme parks for money's sake :)

The industry could fund a small project to test ideas and then adapt the polished design for bigger MMOs. But with the talk about the rush for the golden 1M milestone, everyone is too much of a hurry to work out any kinks.

My humble rant

[ricky]

3.

Very great ideas!
Simulated Ecosystems, though, shines better in concept. If you define strict ecological rules, then the endless waves of player hunters would very easily disrupt the chain. It wouldn´t work, were it to be strictly realistic. I imagine some players would even want to "greif" the ecosystem, by, say, driving an important predator to extinction....However, if you are accurately able to make it look and feel like there is an ecosystem, without actually simulating it, then you have immersion and control at the same time (for instance, basic migratory patterns like you said, and basic hunter-prey situations amongst creatures without player interference, etc).

The idea for genetics and evolution for creatures is great...having creatures develop defenses for certain spell types would make the game much more fun, especially on higher levels, where the player would then have to think intelligently how to defeat a creature that he previously easily killed with a simple fireball...

[Joseph B]

4.

Just an idle thought here.

One alternate way to solve the "assembly-line dragon abattoir" problem would be to have one group actually kill a single dragon, and subsequent groups "re-live" the experience through the proxy of in-game storytellers. Bards of some sort would create an instanced space, in which people could re-do the dungeon/raid/etc.

I suppose one could invent all kinds of convoluted mechanics to maintain the level grind ("I have told you what I know of the dragon's death; if you would hear the true story, seek out Thorvald, who stood with Hafnir's band that fated day").

[Raph]

5.

Odd that UO is mentioned as an example of the "spawn every 30 minutes" sort of game, when it was originally conceived as (and even ran during beta as) an ecological sim exactly like what is described here. It's been described (generally negatively) in lots of places--I suggest some Googling.

The failings proved to be:

- ecological crash when players disrupted the chain by overhunting
- economic crash when players hoarded everything in the world
- player dislike of unpredictability--they WANT the dragon to be in the same place every time, because it provides shared experiences with others, keeps people from feeling "robbed" when someone else gets the dragon and they have no chance to, and provides more predictable play session length and entertainment value

In other words, simulationism is great as long as it's fun, and fun is often more easily achieved by stagecraft rather than simulation.

That said, I tend to think simulation is the future anyway. Getting there will require mastering "stagecraft via sim" and some of it will require adjustment of player expectations ("look, the dragon is dead, kaput, finito. Go find another").

-Raph

[peterb]

6.

I addressed this topic very recently, here:

http://www.tgr.com/weblog/archives/000228.html

The summary is that "realism" is a canard. No one wants realism in games. People _say_ they want realism in games, but that desire only extends up to a certain, extremely limited point. They want _iconic verisimilitude_, not literal realism.

Literal realism in games is so boring that it has been shown in a number of serious academic studies to induce anuerysms in spider monkeys (see footnote 1). Computer and video games are used as ludological proxies precisely because of their transrealistic characteristics. No one plays _Madden_ because they want to throw a football. Rather, they play _Madden_ because they want to throw a football better than Dan Marino.

Hope that helps,

peterb

Footnote 1: I just made that up.

[Flatfingers]

7.

I was going to make the point that, while I like the "complex systems" approach and would enjoy playing such a game, most people don't want "surprising," they want predictable success.

But as Raph and peterb have made that point, I won't belabor it. Instead, let me take a slightly more optimistic tack.

Achieving believability doesn't require massive complexity. A better word to describe what's really necessary is "connectedness."

A relatively small world with few objects and a minimal number of available actions will inspire belief as long as those objects appear to be connected to each other in plausible ways. To put it another way, if you apply a force to something over here, something over there should change as well, and in a reasonable way.

If the wind blows, the candles should flicker and go out; if new members join a group, they should be able to affect the group's organization. The Havoc system achieves something of the former capability -- where is the equally powerful system for allowing highly-connected social dynamics?

Thinking of systems (including human systems) in ecological terms is common. It's a useful model for understanding consumer/producer (or predator/prey) relationships. But living systems are much more connected than this! As James Grier Miller showed in his Living Systems work, living systems show a connectedness at both smaller and larger levels than the simply ecological.

NASA's Life Sciences group has a table of Miller's Living Systems model. The interesting thing about this model from the multiplayer game designer's point of view is the way interactions are broken down for system sizes of "group" and larger.

These large-scale systems are a useful way to think of massively multiplayer environments, because they spell out the functions required by living systems at these levels. A game world whose objects and player actions are so designed as to allow actors (players/NPCs/creatures/other) to fill the roles of Ingestor, Producer, Associator and so on will have a good chance of achieving functional verisimilitude. And this will happen not because the game world's designers deliberately set out to create a full-blown ecological system from scratch, but because ecological systems arise naturally in a (game) world that is a living system.

For some of us, such a world would be fun to play in purely for its simulation aspects. For most others, you'd also need a game... which is where Raph's and peterb's comments apply. Designing a fun game is hard. Designing a stable but adaptable living system is hard. Doing both together is very hard.

But since when does "very hard" mean "let's not even try?"

--Flatfingers

[Paul "Phinehas" Schwanz]

8.

Flatfingers> But since when does "very hard" mean "let's not even try?"

Best Guess: When trying is dependent upon the funding of vision-lacking and/or risk-averse publishers.

I'm optimistic that it isn't, but I may be overly optimistic.

--Phin

[Damion Schubert]

9.

It's long been commented by those of us in the anti-realism brigade that realism != fun. That being said, it should be pointed out that realism != immersion either.

I define experience loosely as 'you're so deep in the experience that you forget you're behind a keyboard'. As such, anything that jars or frustrates you is going to break immersion. (Humorously, the number one enemy of immersion is not endless monster spawning or breaking crates to get potions - it's interfaces so bad they make you want to throw your computer into the river).

All that being said, realism can be a huge enemy of immersion. In the early days of UO, trying to find a monster to kill so you could advance was a huge immersion breaker - although a perfectly realistic result of overhunting. Trying to find a vendor who would buy your shoes when the market was flooded was a huge immersion breaker - but a perfect realistic result of a flooded market.

[magicback]

10.

So maybe believability is not the key. Perhaps rationality is.

When you throw a ball up, it is rational that the ball will come down due to gravity. This is the given expectation.

We just need to design for rational expectations. Quests are reduced to fedex. Complex & believable fixed monster ecology is reduced to spawn and camp points. Normadic monster ecology is reduced to herds and so on.

It's reductionist, but that's the current rational expectation. While feather and a rock fall theoretically should fall to Earth at the same rate, rock just fall faster! When I kill a dragon, I as a RPer don't expect the dragon to come back. But, as an item farmer I do expect to camp the Dragon.

[greglas]

11.

Damion> It's long been commented by those of us in the anti-realism brigade that realism != fun. That being said, it should be pointed out that realism != immersion either. I define experience loosely as 'you're so deep in the experience that you forget you're behind a keyboard'.

And arguably fun and immersion are one and the same -- see the flow concept of fun/pleasure/happiness.

[Raph]

12.

The guaranteed presence of a dragon, or the guaranteed sale of your good, is also an enemy of fun. Predictable success is boring over time.

There are a couple of reasons why those of us in the simulationist camp pursue it. The pursuit of "realism" (defined as mimicry of reality) generally isn't in the top five reasons. Instead, I'd offer up the following as goals:

- greater plausibility (e.g. stagecraft)
Ex.: when the dragon breathes fire, things actually catch fire

- easier content creation
Ex.: A dragon spawner forms where the right conditions arise, and you simply create the conditions with the worldbuilding tool, rather than creating the area by hand

- larger solution spaces for challenges
Ex.: You can trick the dragon by feeding it, or sneak past it, or slay it in the time-honored manner. More playstyles become supported organically rather than by special-casing them.

- self-generating or dynamic challenges
Ex.: simulations can give rise to emergent situations, rather than every challenge being hand-crafted; the difficulty might be that the dragon is blocking the cave entrance to a popular mining spot, or it might be that it is eating all the newbie deer in the area

- consumable content with consequence
Ex.: After the dragon is slain, we can give proper credit to the heroes--they have made a real mark and a real difference. The next dragon will be a different dragon, in a different place, with somewhat different characteristics.

All of these have challenges of their own, of course. Improved stagecraft is undeniably leaning on greater reliance on simulation, going forward--I think that Half-Life 2 is a stellar example of that. But the others are more controversial. Simulation used as a content creation tool has not yet achieved the level of interest that handcrafting can. Larger solution spaces also increase the odds of exploits and balancing very difficult (though games like GTA, Thief and Deus Ex have shown that they can be very fun). Self-generating or dynamic challenges may disrupt the fun that other players are having (the newbie who loses all his deer to hunt to the newly spawned dragon will be pissed, especially if they cannot do anything about the problem). Content with consequences has the issue that people DO crave predictability (cf my book, due out any day now).

All that said, I tend to think those are surmountable issues.

[Greyhawk]

13.

Personally, I hate this idea. It does have some merit, and it would add believability, but it would not make the game more enjoyable.

With that said, I think a limited implementation of this idea could be used to improve the ingame experience. The dragon mirgration idea, for example, would be far more effective if implemented for mass mobs such as goblins and orcs. If the game has a sentient, semi-civilized mob, then there ought to be cities, towns, and villages where those mobs congregate. All existence of those mobs could then be the result of pathing from the settlements into wild areas, providing both individual targets in the hills and fields, and massed targets in the villages.

Lineage II does this partly, EQ2 also partly addresses this potential. I think much more could be done, but I would hate to see a virtual world so realistically modelled that players could drive a mob species to extinction by overhunting. What would be the fun in that?

[Anders Tychsen]

14.

To go back a bit, and continue along with Ricky and Peterb, I would venture the thought that it is not necessarily realism that is the ideal with regards to the natural ecologies of MMOGs. Rather, it is believability. In order for ecologies to be believable, however, they must operate within the world fiction.

The goal seem to be to create ecological systems that allow the game developers to maintain control with the game world in order to avoid ecological collapse (due to player activities such as selectional hunting and hoarding - Greyhawk), but at the same time approximates what a realistic ecology would be for the given world fiction in order to make the virtual world believable and immersive. Another important point in this endeavour is that realism and believability must never supersede the games themselves, i.e. the fun factor (peterb). The afore mentioned dragons must be present, but their distribution and biophysical presence should be guided by ecological principles. In other words: A form of sectional ecology, based on parts natural science and parts game design.

Would it be possible to generate systems that feel like _realistic_ natural ecologies (or connected systems – Flatfingers) – that is, a believable system, without ruining the predictability and while retaining a resistance to exploitation? (and thus which are sufficiently risk-free – in relation to the financial aspect).

[Jake]

15.

It would seem that the dynamics being considered must be weighted more towards believability than realism. In that light, operating on the pseudo-realism of the fictional setting itself would then seem to me to be the ideal.

In recalling a number of my own UO exploits, there were a tremendous number of things that didn't make any sense at all. Decaying houses, monster spawns, lag...heh, we knew that was coming, one hopes. ;)

Of all those things, and those not mentioned, it was never really those that jarred my gaming experience. Rather, it was almost invariably the other players.

In almost every instance I recall, from my hardcore roleplayer perspective, other players breaking character, or merely playing in an out-of-character fashion, was the biggest drag of my entire 5-year UO stint.

Operating further from my perspective, it would seem that cultivating the player-base to function in certain fashions might just as key to propagating believability as tweaking the mechanics.

And it certainly goes without saying that the greatest of systems will be no more and no less than what the players, themselves, make it to be.

Just my two cents.

-Jake

[M. Scott Boone]

16.

Has anyone here played Saga of Ryzom? It appears to be an attempt to answer some of these criticisms regarding believability, at least as it relates to ecosystem and creature behavior. The quote below is taken from their website:

"A Lush, Living Ecosystem

Known as the Living Planet, Atys doesn't behave quite like any world you've experienced in an MMOG. To start with, Atys is growing, slowly sending tendrils out which can unveil new lands and cover up existing ones. Due to this constant growth, the climats and weather on the surface can vary greatly, even in a short time span.

On the surface, the creatures behave even more intelligently. Each creature has its own set of behaviors, and natural enemies. Wild gingos travel in small packs, willing to attack anything they think they can kill for food. The more docile bodok will travel with its herd, but fiercely come to each others defense should they be attacked. Some creatures will migrate naturally; others will prefer to remain in the same general area.

Seasons change naturally, altering not only the landscape of Atys but also the resources available to harvesters and the behavior of the creatures that live there. The weather can also change these things, from raging sandstorms in the desert to thunderous storms in the jungle."

The occurrence of in world events also appear to be affected by player behavior, in a realistic manner. If certain monster levels are not properly managed, in other words if their numbers are not kept down, the monsters will engage in large scale attacks on various gaming areas (cities, forts, etc.).

I haven't played, so I was wondering if anyone had. And if so, whether they thought these attempts helped with believability and/or enjoyment, and whether playability was affected.

Scott

[andrew stern]

17.

Once you blend simulation and stagecraft, which is very typical in games -- rarely is there a "pure" simulation that succeeds as entertainment -- we probably shouldn't call it simulation anymore. During a discussion last year I initially argued that process-intensive games that mix simulation and stagecraft could be thought of as "simulating the logic of a non-purely-realistic worldview", but towards the end of the discussion I realized that's unnecessarily stretching the definition of the word "simulation". (However we didn't come up with a better term more specific than "process-intensive computation system", or perhaps "dynamic representation".)

The dramatic laws of a virtual world, e.g., the set of rules and/or goals, plans, what have you that control the behavior of the elements of the world -- e.g., when dragons should appear and how they should behave -- can and should be as complex and robust and "bottom-up" as pure simulation-type rules, but that doesn't make them a simulation, if by "simulation" you mean modeling some real-world dynamic system.

Drama management (a more nuanced term I think than "stagecraft") can be implemented as a layer above the core virtual world rules and behaviors, that is always observing what is going on in the world, and based on its own (hand-crafted) dramatic logic, carefully alters the state of the dynamic system. This could be a way to allow finely balanced low-level rules with emergent behaviors operate much of the time, allowing for emergent behavior, with occasional from-on-high alterations of the system -- ideally believable ones, not clumsy deus ex machina actions.

In my own experience building these types of things, I lean towards drama-managed yet heavily process-intensive computational systems. I find systems like Petz (which had metabolism, aging, learning, memory, genetics, as well as high-level dramatic goals) more engaging than the less drama-managed Creatures, in which Norns were primarily controlled by neutral networks. Creatures' greater degree of open-endedness and bottom-upness offered more opportunity for emergent behavior than Petz, but the pacing and lack of dramatic moments left me bored. In this 1999 paper I warn against pure simulation:

Because purely bottom-up approaches do not give authors direct control over the behavior of their characters, I challenge the idea that biologically-inspired A-Life alone will be the answer here. Real life, while endlessly rich and complex, is often marked by long stretches of dullness. Users that want to be entertained are not going to be willing to wait very long for something funny or exciting to happen. For example, a recent article about AI in Game Developer magazine mentions that the A-Life techniques used to control non-player characters in the Ultima Online virtual world had to "be compromised in the interests of game play".

As with almost everything, the solution lies somewhere in the middle.

[Raph]

18.

I don't define simulation quite that narrowly--to me a sim can be modeling something fairly abstract. I think there's tons of games that are pure sim and succeed brilliantly, ranging from highly abstract games such as Tetris to more advanced ones like SimCity. I see it as being more about coherency in the ruleset. Traditional spawning mechanisms aren't concerned with coherency--they perform their actions by fiat ata a rate imposed by fiat, and they have no interdependence with any other elements in the system.

I agree that drama management is where we need to go, but I have to emphasize that barring very noticeable exceptions, almost nobody uses simulation to any real degree. There's nary a drop of it in World of Warcraft or EQ2, for example. It doesn't hurt the games any, except for the cost of making the game content and the longevity of the activities you can engage in.

[Mark Asher]

19.

If the MMO spawns a dragon I want a shot at killing it. If someone else kills it first, that sucks. I'm not comforted by the knowledge that a new and different dragon will spawn somewhere else soon -- chances are someone else will kill it before I do.

Simulation is fine as long as it doesn't frustrate me. I don't play games to be frustrated.

[Walter]

20.

At the risk of overjargonizing, it might be useful to distinguish between simulation and verisimulation. (I also think there's some potential for using the word 'isomorphic' with regards to simulation, but whatever.)

If you haven't read it yet, Chaim Gingold's thesis offers some insightful analysis and quotes on Will Wright's work, extremely useful for world simulation. Personally, I think it may be more interesting to direct analysis about the coherency of simulation to games like Morrowind and the upcoming Oblivion, where there's this rather obvious striving towards deep, broad, ground-level simulation (a real shame that Origin is no longer around to carry on the same project). I'm beginning to strongly feel that any game positing a deep, broad world needs to be designed by first working out all the significant causal relationships in the world and making sure there are as few arbitrary origination and termination points as possible at various levels of abstraction. If all of that isn't feasible, either change the world so it is, or (less enthusiastically) whittle away the simulation until it fits.

[Raph]

21.

If the MMO spawns a dragon I want a shot at killing it. If someone else kills it first, that sucks.

Mark, you're going to spend your entire life frustrated and depressed with this worldview. There's ALWAYS a dragon out there that someone else got to before you.

I think there's more to dig into there. You're making some assumptions about ease of access to the dragon, I suspect--thinking that perhaps it's rarer BECAUSE it's simulated?

[Richard Bartle]

22.

Mark Asher>If the MMO spawns a dragon I want a shot at killing it. If someone else kills it first, that sucks.

So let's say there are 10,000 people signed up for your server, and the dragon spawns every 30 minutes. Assuming they all have the same attitude that you do, that means one day in about 208 you get to kill it. The other 207 days someone else gets there first.

If everyone shares your opinion about constant dragon spawning points being important, you'll only get to kill the dragon once every 208 days on average. If you can play a game for 207 sucky days in a row, just to have an unsucky 208th day, well, frankly you'll put up with anything and another couple of hundred sucky days while you find yourself a replacement dragon isn't going to make any difference.

If only a few people share your opinion, OK, so you'll kill the dragon more often. However, a designer might figure that the number of people in favour of fixed spawning points is small enough for the will of the majority (who at the very least don't care about it) to prevail. Such a designer may decide that making the virtual world a richer, more satisfying environment would make the virtual world a more attractive place.

If you already think the virtual world sucks most of the time, making a change to it isn't going to alter your opinion a great deal. It may, however, alter the opinions of other people quite a lot, done right.

>I don't play games to be frustrated.

Other people don't play games to be bored.

If you want your fun predictable and unstoppable, try Progress Quest.

Richard

[Damion Schubert]

23.

If only a few people share your opinion, OK, so you'll kill the dragon more often. However, a designer might figure that the number of people in favour of fixed spawning points is small enough for the will of the majority (who at the very least don't care about it) to prevail.

It's not so much that players want fixed spawn points as much as players want to a directed experience to the dragon. They don't want to have to wander an endless countryside with no direction looking for a mob that could be anywhere. That's the MMO equivalent of the adventure game pixel-hunt. It is both frustrating AND boring.

ProgressQuest cracks are all good and well, but in the grand World/Game schism, the World advocates will always lag behind until they start actively addressing the problem that, in most simulations, its more fun to run the ant farm than to be in it.

[Raph]

24.

Again, though, the goal has to be not to run an ant farm, but to provide a more interesting GAME.

I agree that the directed experience to the dragon is pretty critical. But there's absolutely nothing preventing us from drawing a bright yellow brick road to the dragon, if we want to. That's not a failure insimulation, it's a failure in accessible design...

[Mike Darga]

25.

M. Scott Boone:Has anyone here played Saga of Ryzom? It appears to be an attempt to answer some of these criticisms regarding believability, at least as it relates to ecosystem and creature behavior.

I played Ryzom’s beta for a month or so over the summer, and loved the world. I actually intend to purchase it over WoW and EQ2. I intended to mention it when I first read this thread, but as always I didn’t get around to posting until after the thread died. =)

The text you posted from their website is all accurate to how the game actually works. You can learn a lot about the world and the mobs by watching them interact and fight with each other (which are aggro, which will gang up on you, how powerful things are), but the main benefit as I see it is immersion. Nevrax has done an excellent job of creating a world that feels alive, and of illustrating that simulation isn’t necessary to satisfyingly mimic a living world. As far as I know, animals don’t attack each other because they are “hungry,” they just occasionally fight with each other. The effect, however, is just as pleasurable and convincing to observe.

The changing of the seasons does affect resource gathering quite a bit (on top of being beautiful), but as far as I know, the animal behavior doesn’t actually affect gameplay much. When last I played, the animals didn’t do any sort of traveling to drink water from lakes and streams, and didn’t affect each other’s population numbers in any meaningful way. However, it seems like the technology to do those things is the same as what they’ve already implemented.

I would find it very satisfying to have to wait for a member of a predatory herd to stray away for a drink of water before attacking it, to have to kill the predators attacking my pet/livestock, or to have to compete with predators that are killing off the same thing I want to kill. I really don’t see how any of these things necessarily require simulation to be effective- they could be driven by the standard percentages and probabilities for results that would be very convincing.

-----------------------------
Damion Schubert: It's not so much that players want fixed spawn points as much as players want to a directed experience to the dragon. They don't want to have to wander an endless countryside with no direction looking for a mob that could be anywhere. That's the MMO equivalent of the adventure game pixel-hunt. It is both frustrating AND boring.

Raph: I agree that the directed experience to the dragon is pretty critical. But there's absolutely nothing preventing us from drawing a bright yellow brick road to the dragon, if we want to. That's not a failure insimulation, it's a failure in accessible design...

So much has been stolen from Tolkein over the years, but nobody has successfully adopted his ability to reflecting a location’s inhabitants in the location itself. We all know that good characters are products of their environment, but Tolkein knew that good environments are also products of their denizens. Because of that, he is always able to foreshadow what lurks in a place by describing the approach to the place; how the terrain changes as it gets closer, what animals are or aren’t there, mists, lack of vegetation, etc. Reading this thread, the first thing that comes to mind is the withered heath surrounding dragon territory (I wish I had the books on hand to quote from!), but some other great examples are descriptions of approaching the heart of the old forest, the barrow downs, Shelob’s lair, and Mordor itself. No one in those books ever knows exactly where enemies are at, but they always know where to look. It’s a very simple “you’re getting warmer/colder” dynamic that’s very effective in letting people know how close danger is. (Foreboding also increases feelings of heroism, as characters are required to knowingly press forward into obviously increasing danger. I would say that this is one of the main themes of the LOTR books and movies.) Perhaps now that worlds">http://terranova.blogs.com/terra_nova/2004/12/world_of_warcra.html">worlds are beginning to achieve true atmosphere, changes along these lines are not far off.

We’ve long had technology for directional/localized audio for ominous sounds in the distance which increase as players approach. The same technology we use to change seasons could easily be used to burn out the area around a dragon more the longer it has been there/the more dangerous it is. Z-depths are now typically large enough to allow a player to catch glimpses of fleeting shapes in the distance. In a system like Ryzom that includes realistic wildlife interactions, it would be very effective to make an area around a powerful spawn conspicuously empty of mobiles, or better yet to have normally-aggro predators run past the player “in fear” i.e. without attacking. So if you’re in what should be the dragon’s territory, but you can’t hear or see anything strange, the snow on the ground isn’t melted, and you see bunnies frolicking around, you know to come back later and try again. As Raph says, it’s not a question of technology at this point, just design.

---------------------------------
Another thing I wanted to mention is how little it can take in some cases to achieve believability within certain worlds. Toontown has the most ridiculous way of spawning enemies (robots fly down from the sky, using helicopter blades on their backs, wander around mindlessly until they bump into something, and if not they sprout the helicopter blades again and fly away once their patrol is finished), but it is perfectly believable within the world they’ve set forth. It really causes no problems at all with lack of immersion as far as I’m concerned.

Having dragons spawn in the same place all the time wouldn’t be a problem for me at all, if it was handled appropriately. Having it just pop into existence in a specific cave doesn’t do it for me. Having it hatch from an egg in a cave that was chosen by its mother because of lava streams or steam vents providing lots of heat makes much more sense.

For full grown dragons, it would be great to have them leave their caves each morning to hunt and terrorize the countryside or a random city. Players might have to fight it off from selecting their livestock to take home, or try to lure it toward an enemy village rather than theirs. After some period of time, the dragon would make its way back to its cave with corpses to feed its young. Lowlevel players would wait for the dragon to leave so they could go fight its young, and highlevel players would wait for it to get back so they could fight it. The townspeople and crafters get a little excitement now and then, everyone gets a chance to fight the dragon and feel like a hero (either to chase it off or to kill it at the end of the day) and it’s all fairly appropriate to the reality of the world. When dealing with a scenario like that, I don’t think that I’d mind much if a new dragon took up residence in the same cave a day or two after the last one was killed (and maybe if the cave is surrounded by lava and sulphor, it’s just prime dragon real estate?).

I hope someone out there is still reading this thread!

[Anders Tychsen]

26.

Very interesting post, Mike. I generally agree with most of your points, however, I think it would be really interesting to do something more detailed than the ecology of Ryzom. From what you describe, it seems that they have gone some of the way - and way longer than most MMOs, but not explored in depth what a well-designed believable/realistic ecology could do for an MMORPG.

I agree that the design of an online ecology is not something that should be left to pure science (and that we do not need a lot of new technology _at first_), i.e. the true simulation of an ecology, but rather a mixture of stagecraft and science, with the science (the biological and geological sciences) forming the basis for outlining an ecological system, and stagecraft used to implement it and modify it to suit the game, the fun factor etc. With my background in the natural sciences, I am possibly a bit over-biased towards having more realism in games, but what I think is the big challenge is to use the natural sciences in these fantastical, unrealistic world settings to form the basis for ecologies that are true to the world fictions - AND the game, AND the fun, ......

Having prime dragon real estate is not good enough for me, unless dragons are really numerous! - especially if dragons are modelled as intelligent creatures. If 250 dragons has been slain in a cave, why go there? Unless you did not have prior knowledge, e.g. the migration example that was talked about in the thread above.

I think that games would be improved if there was a consistency in the way that the virtual worlds were constructed, and a deep, geological as well as historical, background.

Excellent point on Tolkien - I agree wholeheartedly. I do not know about the realism of MiddleEarth in scientific terms, but it is believable because it is consistent and mimicks what it would be like if the world actually existed - erhm ... with some leaps of faith

Something one of the guys said over in the WoW is WOW thread struck a note: The most believable worlds are the ones we do not notice. I wonder how this implicate the believability/realism/simulation discussion?

Just some random thoughts ...

[matt]

27.

DUDE ITS JUST A GAME HAS NOTHING TO DO WITH REAL LIFE WHAT-SO-EVER AT ALL PLAY THE GAME AND QUIT THINKING ITS REAL LIFE STEP AWAY FROM UR COMP GET SOME FRESH AIR AND BE HAPPY U HAVE SOMETHING TO DO INSTEAD OF BITCHING ABOUT IT...
Sincerely yours,
Matt..