MMOs, Operant Conditioning, and Progress

[NewClassic_v1legacy]

I had a discussion earlier today that got me thinking on the nature of online gaming, and how it could be best achieved given our current technology. It started with a discussion about EVE Online.

For me, EVE has always been both a subject of much fascination, and a sign that a solid concept doesn't make up for lackluster execution. The game itself is supremely built and designed in the best interest of the player who wants to do something unique to themselves. It's a game that as easily rewards trading as mining, security for player guilds to piracy of the aforementioned traders, miners, and security. One can make their bread and butter as a transport ship as easily as they could any other build, offering to be an interstellar taxi should they so choose.

Then when you get out of the individualized play styles, which incorporate entire player-wide colonies of team members, coordinators, and player operated and established stations, then the malleability of the world as a whole is almost perfect. Add in the player-managed economy, and it becomes the sort of free market on which economists could write novels. It manages a free-thinking living breathing world. Something that more MMOs should strive for, but it also does that in a way that manages to be alienating to the individual players, and often give the villains of free-space too much space to be villainous. Cost of freedom is risk.

That, and a system which is designed to discourage casual play. Someone who wants a game to kick on for an hour every day and play a bit has no place in EVE. It's a monster that requires time, devotion, and for all of its efforts, provides very little for the player outside of more work for enough resources to continue working. Which is where Operant Conditioning comes into play.

Spoiler: Brief Description of Operant Conditioning/Skinner's Box

For those unaware, Operant Conditioning was an experiment wherein a man named B. F. Skinner trained various of animals to do a certain task after being given a stimulus. In this case, a light would flash, a sound would play, and they would have to pull a lever. If they did so, food would come out. If they did the wrong thing, they would be shocked. If they did nothing, nothing would happen.

This experiment proved that given enough positive reinforcement, the creatures learned to repeat the motion because of the light and sound, not for fear of reprisal or want of the food that dropped, but simply because they had been conditioned to do so. This stimulus yields this result. Almost without fail.

He proved that it worked on humans, too.

The discussion devolved as Operant Conditioning became the work/reward mechanism for most of the titles we could think of. World of Warcraft, certainly. Ragnarok Online, Phantasy Star Online (and later Phantasy Star Universe), Everquest, Ultima Online... All of these relied on our stimulus-response to a game. "Ding!" is a noise that most WoW players should recognize, and almost immediately go to either congratulate a passerby/guildmate/party member/themselves, and often don't think about why they do this. Operant Conditioning.

We train ourselves into habits with a lot of online games, and when we stop to look at them from a distance, we have difficulty seeing the appeal outside of the interaction. Which is why when I think about this, I wonder where the "next step" of MMOs can be, simply because of the very structured nature of almost everything on the market right now.

So my big concern was how MMOs, especially the RPG types, can get away from this particular habit.

I find that RPGs especially have a bit of an issue with MMOs. The narratives can't possibly culminate to any epic, world-shattering/saving shift into the world and how the game works because the world would have to resume some kind of conflict for all of the players who weren't the absolute savior of the world. This tends to yield a lot of issue because most non-online RPGs focus almost exclusively on the "Save the World" arc, which has been consistent with the beginnings of RPGs, either in the western or eastern games.

Without that kind of narrative backbone, then the only solution is to either have a conflict that can't be solved terminally (IE, bosses that respawn instead of dying forever), or conflicts that work on a scale that's player specific. For example, uprooting the mayor of the player's birth village. This is large enough to be able to construct a campaign (as doing something similar would take years of effort in a real world scenario), and small enough not to shoot all of the other players in the foot while still being able to have a persistent impact on the world.

For the sake of argument, let's not talk about the world segregation techniques used in World of Warcraft and Guild Wars. It's a good idea, but since it's been done, we'll argue that it's too player-base dividing to be functional in a grand scale.

Then comes the combat and gameplay, which relies very heavily on the operant conditioning, and rarely is fun for its own sake. (How many times have you heard complaints about grinding or equipment micromanagement from someone who plays a game full-time, professes to hate it, yet continues to play? Operant Conditioning at work.) This is the problem that seems to crop up the most, and I see very few solutions to it. The problem is that most single-player games that use this mechanic can do so because while each individual level and item acquired build up to a certain level of skinner's box mechanics, it avoids reliance on the techniques simply by having a final goal and a narrative end. MMOs don't have that luxury. At least, not really.

Part of the reason I mentioned "given our current technology" in the opening line is because I like toying with the idea of an MMO DM AI. A computer that is capable of creating and maintaining five or six major, world-wide conflicts and ideas, making a sort of serialized conflicts that each have their own major consequences and effects within the game world, but can only be achieved by players once ever, and then that particular conflict cannot come up again once resolved.

Proceedurally generated narratives have the benefits of world persistence without the flaws of game-breaking for all of the other players, given that every player has the same chance to find and fix major conflicts, and once a conflict is resolved, a new conflict is around to take its place. Or has alternatives to be chosen.

The problem with this is that the technology is well out of the realm of possibility. That would require a server that could take stock of world situations, adjust the world accordingly for everyone, which involves manipulating to replacing NPCs, towns, textures, and entire worlds of other things. It's a level of procedural generation that has no peers. Not even for games as variable as Minecraft, Fuel, or Synth.

Despite that idea, I'm actually not too sure how MMOs can really get away from their current format without also changing the way the gameplay mechanics have been done for years.

So, conversation prompt time. What could be done to MMOs so that they rely less on very simple stimulus/response mechanics, and make themselves more fun, progressive, or different?

 

[F4LL3N]

This post is really long and my mind's not all there today, so forgive me if I misinterpret what you're saying, or I'm hard to understand.

I'm somewhat designing an MMORPG for fun. The technology needed for it is probably 20+ years away, and ignoring the general theme/universe it's held in, some general ideas I have for it perhaps relate to what you're asking.

As a core, the 'main' game engine would be best related to the one used for TES: Oblivion, only much more perfected in terms of visuals, physics, animation, etc. To build onto this, players can collect 'Physic cards', which in turn would completely change the feel of the game. For ease of understanding, let's say you could pick up a 'Call of Duty' card, or a 'God of War' card. This would completely change the feel of the game, for your character. It wouldn't change the way water flows, for example. I think this relates to Combat and general gameplay/feel. It would also relate to classes and gamestyle a lot. If you're a gunner class, you'd ideally use a 'CoD' card or other gun related card. Chain weapons you'd use the 'GoW' card, etc.

In terms of story, my idea is that there isn't a 'main plot' or game ending story. There would be your typical side quests, and then more major plotlines. The latter would somewhat be like normal main plots, but this is simply for your 'storyline fix'. However, the real action would be in limited-time, 'single use', possibly world changing mini-updates. These would be coded and released regularly by the develper. For example, one update could see an alien invasion or a 'super boss' attack. This would all be in real-time, and depending on what the developer wants, could see anything happen... However, once the threat is over, it's over for good. So if you go on a 2 week holiday, you could miss this quest forever.

Every action in the game would, or could cause a 'chain and effect'. It's been attempted before, but not very well. The AI is extremely intelligent, as in it doesn't even need to be scripted. For example, let's say an NPC witnessed his parents killed as a child. As an adult, this person may commit suicide if he doesn't receive enough support from his community. Maybe he becomes a drug addict. Maybe the murderer had a scar on his face, so this person now spends his life hunting for their killer. Ideally, the AI would have completely un-scripted individual thought. A peasant could one day become the king of an empire depending on how his lived his life. And NPCs can do everything a player can do, i.e. choose a player class and complete quests, etc. All completely from their own thoughts. An NPC could become an inventor or a scientist. If their smart enough, they can code and draw their own objects, items, etc. The scientist could genetically engineer two animals to create an entirely new species. All completely out of nothing... The 'beast design/artist' could log in one day and see one of these creatures and think "what the hell, I never drew that into the game'.

The self-thinking, intelligent AI and chain and effect concepts would play into every aspect of the game. They could start one of the super quests, as mentioned above. Two major kingdoms could start a war with eachother for various reasons. And this would effect all aspects of the world around them. Although, ideally non of this would negatively effect gameplay/be really stupid. Even taking into account entire kingdoms or 1 billion+ strong religions can be destroy. To explain that further, for example, NPCs can start their own religions, all from the thought of one person. However, they're not going to start a religion based on worshipping flies. They could, depending on how much that person wants to worship flies, but ideally they won't.

Players can also do essentially anything (except add code). Do you like fighting games? Then play the game like a fighting game, and it'll feel like an AAA fighting game. Do you like racing/flying games? Play it like it, and ignore the rest. In that sense, the game caters to pretty much anything you want, while still essentially being a fantasy MMORPG at it's core.

Another large aspect: a player can create their own empire, or town, or what have you. There would be various tools to do this. For example, they could switch to a top down RTS view to build, using predesigned buildings, or something along the lines of Minecraft for custom buildings. From their, they would have to expand it's population through recruiting, birth, slavery, etc. They then have to manage it in the same way you would manage an empire in real life. Everything from military, politics, education, economics, religions, etc. You can't just click on someone and they instantly become a soldier or a priest, but you can influence your empire. Well, you can force them through slavery/conscriptions, etc. From their it'll become a PvP/PvNPC situation where you can conquer other empires/land or create alliances. You can force other empires to pay you regular amounts of X or you'll conquer them, etc.

TL;DR: It's essentially like a real-life fantasy/sci-fi world in the sense that anything is possible within the universe's general guidelines. It's like a living game-organism... Did I say 20+ years? Let's give it 40 years.

But I think the same concept could be applied using today technology. Actually let's say within the next 5 years. No where near to the extent I am talking about. But some of the concepts. A lot of it's already been done over and over, just not very good. Btw, I'm still not thinking straight. So I'm not sure if much of what I said makes sense.

 

[FieryTrainwreck]

I think we're going to see some legitimate divergence with Guild Wars 2. They're really emphasizing the dynamic nature of the quests as well as the more tactile and reactive combat mechanics. Assuming they pull these things off, I'd say they're poised to alter at least some parts of the standard MMO equation. At the very least, it could be an evolutionary game.

With MMOs, the problem is longevity. A successful game keeps people playing for years, and it becomes increasingly difficult for them to cut ties with so much accrued virtual equity. The "me too" nature of this (and most) industries means a wake of copycats trailing behind every big money-maker. Combine the longevity with the copycats and you've got pretty lengthy "cycles" between opportunities for design innovation. That doesn't mean they aren't coming, but you can expect them to take much longer than the more standard genres.

 

[TimeLord]

My answer? Make MMO's less about doing jobs for idiots standing on the same space for the entire game and make actual missions that might involve doing something other that running to a point on the map and killing everything there and on the way.
Basically, remove grinding and make the story take president more. But as a man once said; "Removing grind from MMOs is all very well until you think about it. Because grind is what keeps people playing MMOs for so long in the first place!"

 

[Seishisha]

I personaly would like to see some kind of immergant gameplay where in your overall objective remains the same but has multiple solutions, for example the quest is to kill ten bandits, now this is a very average scenario for mmo's and games in general i suppose, my point is it would be nice if you could plant a bomb/detonate somthing nearby to kill them, or poison the food suply, or maybe even just steal all the horse's so the bandits bugger off, now looking at somthing like world of warcraft all those examples are probably in a quest but they are specific and must be completed or are only used once for a specific quest. I'd like to see a system that has many solutions to the same problem with a more flexible completion criteria, do you storm in spells blazing and swords swinging in the "old fashioned" manner or do you use the area to your advantage? I guess to have this much freedom in an online game would cause immense problems unless it was heavily instanced though which isnt a bad thing in its self but i feel it somewhat defeats the point of playing in an online game if you segregate yourself from everyone else.

Im not a games designer but i do have some idea of the depth and complexity making even somthing simple sounding like killing ten bandits actualy is, throwing in a mix of online competition from other people also wanting the bandits and adding numerous solutions would be an absolute programming nightmare. As a concept though i think it has merrit and would be worth exploring if and when it becomes practicle.

On a final note i'd love to see some innovation in the combat aspects of mmo's almost all of them are hotkey based, which again in its self isnt a bad thing its just a little dull. If at all possible i'd love to see a more active combat system that relies less on hotkeys and more on a tactile response from say your mouse or the angle and direction your character is facing, im aware age of conan tried somthing simular to this with having an almost combo style directional attack system but from what i've heard it didnt work out too well. Idealy it would work somthing like this: dragging the mouse to the left whilst holding left click (<-------+click) would result in your character swinging his axe to the left, or causing his fire spell to arc, again though this is realy a technicle limitation issue when you factor in latency and so forth so right now it would be very impracticle to try and create.

I guess my point is that the world is full of people with bright ideas and on how to make things better but more often than not they have to compromise and do what works best not what is best and that is a real shame. Talking about adding new features or mechanincs is a wonderfull thing, anyone with a real passion for gaming probably has thought to them selfs if only this game had that feature. Innovation is often the lifeblood of games design or atleast it should be in my opinion but actualy implementing these amazing ideas is realy hard or just down right impossible with the engine your using or the time limit your given. I could go on but i think i'd just be wasting everyones time with an elongated post.

 

[BloatedGuppy]

Snip

Good post. I've been really interested in operant conditioning and skinner boxes and how they apply to MMO's for a long time. I actually think we're starting to move further and further away from this as the numbers in these games are dragged more and more above the hood. I think Everquest was the perfect example of operant conditioning in action because the positive feedback was so random. These days, we know exactly how many lever pushes are going to give us a biscuit. The element of "randomness" has been largely removed, and I think the games have become less psychologically addictive as a result.

What do we do to make these games more about "fun" and less about "busywork" and pushing levers? Well, fun is different for everybody, but I think the genre needs to move away from virtual tedium for virtual rewards and make the game play its own reward. But how to keep a game from becoming stale after hundreds if not thousands of hours? You need to lean on the multiplayer aspect, and you need to lean on it hard. A mob can only do so much, as a challenge. It can only ever do what it's programmed to do, and even the most complicated enemy will soon have solutions on how to overcome it posted everywhere. A human opponent is unpredictable. A human opponent is endlessly dynamic.

What you need to do, then, is marry your PvE and PvP, and find a way to create an engrossing PvP experience that doesn't get hobbled by player population issues or player behavioral issues. As to how to do that...I don't really know yet.

 

[F4LL3N]

Snip

Definitely. I believe it's really the technology that's preventing MMO's moving forward. There's only so much you can do with things such as 3rd person perspective, lock-on, point and click, hotkeys... basically the World of Warcraft approach. This is what I was getting at in reference to the Oblivion engine, and these so called "physic cards", in terms of gameplay as the reward, rather than grinding and flashy lights.

Currently, there's only one style of gameplay in any given title. There's PvE, PvP, and questing; however, the gameplay doesn't change with any of these. There's generally multiple classes and races in any given MMO; however, this doesn't change gameplay either. Not to mention this single core option of gameplay can be rather bland to begin with. The difference between using a Warrior class and a Mage class is non existant. Strategy differs to a certain extent, but not core gameplay.

I'm going to use real game titles as an example, which is basically what I had in mind, except it would be intergrated more into the core gameplay and 'core' world, in addition to 'instanced' mini-games.

I'll start by using Assassin's Creed as an example. In one instance, you fall asleep and transport to the normal Assassin's Creed world. Here, you're traversing buildings in a parkour style, you're pick-pocketing, etc. The game feels exactly like Assassin's Creed. From there, let's say you wake up and than enter a different world. It's a racing game. This game actually feels like Need for Speed or Gran Turismo, for example. You awake and enter a different world. It's a FPS, and it actually feels like Call of Duty, or Battlefield, or TF2, etc.

I guess what I am describing is aesthetics. Currently, there's little difference between the way a Warrior feels and the way a Mage feels. Not only do they feel the same, but they don't really feel like what they're meant to be. Warriors don't actually feel like warriors, neither mages or any other class. Riding a horse doesn't actually 'feel' like riding a horse. This is true even with the best of games, in my opinion.

As I mentioned, this could also be applied more in-game rather than in a dream world as described above - by merging PvE and PvP. I think MMO's could take inspiration from FPS in this regard. Several FPS have intergrated vehicles into core gameplay in a very aesthetically pleasing way. For example, in Halo 3, I believe there's several instances where you can fight a Scarab Tank with a human tank. Walking and driving a tank are two very different game mechnics that generally work pretty well. Unlike in MMO's where riding a horse can feel somewhat unnatural.

One scene I have in my head: You're driving along a desert road on a motocycle, and out of nowhere a giant worm smashes through the ground. You're in first-person mode, you pull out your sword or rifle and start trying to kill this thing. It's all unscripted, and as he goes in and out of the ground he's completely destroying the enviroment. Visibility drops as sand is filling the atmosphere, giant rocks are crashing to the ground, in which you have to dodge. You know you're fucked... Well, that's basically all I have. Just full on chaotic un-scripted scenes, each one being unique and straight out fast-paced fun.

I kinda rambled a bit and I'm not sure if I actually got my point across. Something to do with aesthetics.

 

[NewClassic_v1legacy]

I'm not a games designer but i do have some idea of the depth and complexity making even something simple sounding, like killing ten bandits, is actually throwing in a mix of online competition - from other people also wanting the bandits to adding numerous solutions - would be an absolute programming nightmare. As a concept though, I think it has merit, and would be worth exploring if and when it becomes practical.

Genuinely, for this, I think that much of that ideal came with games like All Points Bulletin or WoW, with player-populated opposing factions. Granted, in practice, this makes for a pretty sloppy moment because a majority of the gameplay has to give more or less fair digs to the player, which creates fairly simple PvP-style setups. It makes a wide world play more like a small Capture the Flag map on an FPS, or a PvP arena that could be found in any game.

Although the idea of multiple solutions does sound a lot more manageable with current technology. It just means that each event has five or six different "victory flags," which while coded ahead of time, could come from a list of 10 or 15, and give the player more variety in their problem resolution. That alone should be a more solid dynamic, gameplay-wise.

I kinda rambled a bit and I'm not sure if I actually got my point across. Something to do with aesthetics.

The biggest issue with code like this is code unyieldingly varied it has to be. Even with something so simple as switching between first and third person (from motorcycle to sword-and-shield, and back) takes a lot of code, and the driving physics and detail will often take a backseat to first person mechanics. Think about Grand Theft Auto IV as an example of showing what happens when so much code has to be spent on each individual facet of a game.

Then tack onto that the destructible environments, so more code to be had there. Then when you add the fact that the server has to keep track of the damage (not just the individual player's CPU), and update it for all of the other players, whose actions continue to tax the CPU for map maintenance and variable control. Then introduce other player-types to the scenario you mentioned. If an RTS player trips over this worm and sends an army as part of combat, the first player at the scene has to contend with kill stealing from a whole different angle of battle.

Then comes dividing experience, and who achieves what rewards for combat, not to mention "loot" of a sort. And if you instance this sort of thing, how does it handle a persistent world?

There are a whole mess of variables involved here. So while the variability of the play styles does appeal, you do have to contend also with the idea that there's going to be a lot of debugging to do from a code standpoint, and a lifetime of planning in order to avoid server failure and player interaction misfires like those described above. It's a very, very tall order for any developer.

 

[NewClassic_v1legacy]

I didn't read your OP, no time, just read the bottom discussion thingy.

That's a terrible to habit to keep, especially on a mod's discussion thread. If you didn't have time to read, then why did you bother posting in the thread to begin with? Terrible, terrible habit.

In either case, had you read the opening post, you'd've seen that I discounted the idea of instanced worlds, for the sake of argument and division of the playerbase. While I don't think it's a bad idea, I prefer to keep the thread more on the topic of progression. Or, failing that, then approaching the different worlds for storyline reasons in a way that would help unify the worlds rather than divide them based on story progression.

If anything, I like the idea of having found a way to divide everything, but I still don't like the idea that the world in very specifically this way or that way. I'd prefer there to be, at the very least, player factions which can hold "control" over buildings, and adapt them in various pre-built configurations, instead of having "The world is this way until cutscene A, then this way until cutscene B." It keeps the world static, which is part of the problem to begin with, and also divides the playerbase.

I also feel like Guild Wars - though I can't speak for the second one as I haven't been keeping up with it - is a bad example because it's very linear, in as far as MMORPGs go. The story is this way, leads to the end, and then stops. It's such a small narrative for a big world, and keeps the players separate from one another, not only by level, server, or location, but also storyline.

 

[BloatedGuppy]

I also feel like Guild Wars - though I can't speak for the second one as I haven't been keeping up with it - is a bad example because it's very linear, in as far as MMORPGs go. The story is this way, leads to the end, and then stops. It's such a small narrative for a big world, and keeps the players separate from one another, not only by level, server, or location, but also storyline.

He's referring specifically to Guild Wars 2, which is significantly different from Guild Wars 1 in approach. Guild Wars 1 wasn't even really a MMO so much as a co-op fantasy game with a rubbish story. Guild Wars 2 is focusing on large scale public events (think WAR's public quests writ on a larger scale) that have multiple phases depending on whether the players are successful or unsuccessful.

However, as much as Guild Wars 2 is wanting to be all things for all people, I doubt it will attend to many of your desires as far as pushing the genre forward, beyond possibly being a little less of a chore to play. It's still very much ensconced in the MMO traditions that we've grown up with since Everquest.

I agree with you 100% that the brightest future for the genre would be in procedural generation and emergent game play, but the game worlds I imagine in my head (deformable ala Minecraft) and the game play I imagine in my head (true play style differentiation...a rogue might play the game like "Thief"), is still a long way from being realized I think. This is part of the problem, too, with WoW casting such a long shadow over the genre right now, is we're seeing very little in the way of experimentation or ambition in MMO design. Guild Wars 2 looks groundbreaking compared to the parade of me-too clones we've been subjected to, but in reality it's fairly staid and traditional. I miss the earlier days of the genre, when UO was wildly different from EQ, and the sky seemed to be the limit. Now it's a question of losing our gourd because we won't have to spend 45 minutes looking for a goddam healer.

 

[NewClassic_v1legacy]

[W]hat exactly isn't progressive about instanced worlds? That rectifies the problem the world lacking monsters for all players to kill, leveling issues, loot, and it now allows for original worlds based on your choices.

The world not being static is an issue for those who may not play 24/7. If clan A controls territory A but they can't meet at the right times then clan B takes over without any input from A. Oh gee the great gameplay there is astounding!

I don't really see the issue with separation by storyline, most MMOs have a "tutorial" world.

Depends entirely on the system and setup. For the faction concept, I was genuinely thinking something more along the lines of either Urban Dead (with the human faction and the zombie faction) whose map is static, but whether or not a building is barricaded or overrun with zombies is variable day-by-day, and often changes because of how many zombie or human players are on and have action points. It's not like a building is likely to be ganked because the human players are asleep, but rather that it just happen to be more zombies on at the time.

Same could be said for No-Sec (0.0) space in EVE. Or most of the planets in Planetside. There's more than enough playerbase, usually, to ensure that AFK ganking isn't common. Possible, but generally implausible. Clan-warfare, I'd think, would happen more in field outposts. Tents and such, but that's really splitting hairs.

As for the instancing, I agree with BloatedGuppy in saying that limiting the world that way separates the playerbase. In my experience with Guild Wars, I've actually had more MMO-like experiences from Halo or Diablo, rather than the actual MMO that is Guild Wars. The instances worlds and hugely varied servers just kept the players too far withdrawn. Gameplay-wise, the game was perfectly fine, just didn't really feel like an MMO to me.

I feel like that problem will persist assuming the worlds get instanced heavily. A bit of population makes a game a terribly different ballgame. Just my thoughts on the matter.