World vs. Game, Emergent Gameplay, and the Fun Loop

[This is part of a discussion of my indie MMO, codenamed "Gorgon" for no particular reason. This week is about some of the earliest questions you have to ask before you begin your MMO.]

Realistic or Gamey?

One of the first MMO design decisions is figuring out where your game falls in the “game” versus “world” spectrum. The idea here is that every MMO is a simulation of real life to some extent. However, it’s supposed to be a lot more fun than real life. So you have to make changes to the world that aren’t realistic.

If combat were realistic, a single sword-hit would maim or kill any human being. If travel were realistic, you would need to spend 80% of your time traveling between interesting spots. If commerce were realistic, … well you get the idea. Too much realism sucks. Every game has to find a point along this continuum.

For any given game, you make this choice based on who you imagine playing the game. And most MMO designers right now will tell you that the “casual audience” (by which they mean “nearly everybody in the whole world mwahahaha we’re gonna be so rich”) prefers an extremely directed game that drops realism by the wayside in order to focus on fun.

I generally agree with this sentiment: highly-directed games are more accessible, meaning they have more players. But my game isn’t designed for “everybody in the world” or even “everybody on Facebook” or even “everybody who loves Farmville” or even the rather pathetic demographic of “everybody who likes World of Warcraft”.

No, I’m making an indie MMO, and for once I don’t have to peg the world/game meter on “100% game”. I want my game to appeal to a niche of players who are willing to play those “100% game” games, maybe, but would prefer something a little further down the spectrum.

This doesn’t mean financial suicide, though, because there’s actually a ton of these people. And I don’t even need to find a ton of them in order for my indie game to be financially successful.

(Unfortunately, too many people think “more realism” means “more like EverQuest”, which is hilarious on so many levels that it makes me want to cry. So to be clear, I do not mean “realism” in the sense of forced grouping, not having a compass on screen, or camping monsters for hours or days on end. Stay with me here!)

The Tyranny of the Fun Loop

Normally, if I was contracted to do an MMO design, I would almost entirely ignore the “world” and focus on “making the most fun game possible.” A common successful approach to making a fun game is to divide and conquer: first you make the game fun in tiny 30-to-60 second chunks. When you’re confident that the lowest-level thing you do in your game is fun to do over and over and over, then you step back and make a fifteen-minute “fun loop” (or some similar time window). Thus in WoW, killing a monster might take 30 seconds, but completing a quest takes 15 minutes. These are loops: you are rewarded for completing them and are then pushed toward doing the loop again.

This is a very effective way to make a highly directed game. I’ve used it before with success, and I will no doubt use it again in the future. I’m not knocking this method. But it’s not a good approach if you want the game to have more “world” in it.

Consider the difference between WoW and Fallout 3: in the former, you always have numerous quests and have exceptionally clear directions on how to achieve said goals. In Fallout, you get quests, but they are often of the “big picture” variety, and you’re left on your own to figure out how to do them, or even if you want to do them at all.

I remember spending 30 minutes once I left the tutorial area of Fallout 3, doing basically nothing. I was picking over the post-apocalyptic remains, occasionally hitting a giant bug with a stick, and slowly getting my bearings. It was kind of boring at the minute-to-minute level. But it was lots of fun in the bigger picture.

Flash forward to a few hundred hours spent in Fallout 3, and I can confidently say that Fallout 3 doesn’t give a damn whether I have fun every 30 seconds or not. Sometimes I have fun for a straight hour, and sometimes I wander around being vaguely bored for an hour. It’s an inconsistent game.

(However, Fallout 3 does have a few long-term loops. Almost all games do! For instance, the inventory system causes you to return to town every hour or so of play, which causes you to shop, which causes you to spend money, which causes you to need more money, which causes you to go out and explore some more… a classic behavior loop.)

I’m not saying Fallout 3 was made carelessly, because it definitely had a ton of care put into the world. It just uses a different reward schedule than other games. It’s not chunked up or signposted nearly as orthogonally as Diablo or WoW.

This is one of the reasons that Fallout 3 feels more like a “real world” than WoW does. Without those directed loops to constantly guide you, you have to think about the world more and figure out what to do, and the more you think about the world, the more immersive it can be. However, lots of people find this lack of direction to be frustrating or boring. Fallout is certainly not for everybody.

But let’s do a quick reality check here: Fallout 3 sold 4.7 million copies. Even if that’s not “as big as WoW”, it’s incredibly big. Clearly there is room in the marketplace for games that don’t tell you what you should be doing every single moment of the game.

One of the neat things about a game that doesn’t try too hard to tell you what to do is that players find their own things to do. In fact this is often the most fun part of the game, and if it’s something the designer didn’t specifically think about when making the game, we call it “emergent gameplay” because we like fancy terminology.

Emergent Gameplay Comes From The Interaction of Game Systems

Many designers who hear “emergent gameplay” think it means “adaptive enemy AI”. That’s because FPS games use AI as the primary way to get unpredictable scenarios. But that’s just a common example. Actually, you get emergent gameplay whenever you have a complex set of interacting systems. Any systems.

I’ll give you an example from Asheron’s Call 1. There was a dungeon in the middle of nowhere that was full of mosswarts (your typical frog-like baddies). I don’t recall it being a very important dungeon, except for one thing: a few of the monsters dropped special loot. They would randomly drop either an “acid axe” or an “ice tachi”. These weapons were magical, but not particularly good. (Not even good to look at! The ice tachi’s “ice particles” were simple white triangles, so when you ran around with it, it looked like you were dropping tons of White Cheddar Doritos on the ground everywhere you went.)

But you could get lots of these weapons, and they were worth a lot of money! If you were a certain kind of gimpy character, at a certain level range, then this was the best money you could get.

So there I was, having filled my inventory with Dorito Swords and Green Jelly Axes, and wondering what to do next. I didn’t have the ability to teleport. I could barely move because I was so encumbered (each weapon weighed a ton, and the system penalized you for carrying too much weight). And I was scared to travel in this condition, because if I had to fight anything tougher than a mosswart on the way, I would probably be killed from the encumbrance penalties to my combat skills.

Fortunately, there was a small town relatively nearby, and the NPCs would gladly buy these weapons. However, money in Asheron’s Call had weight. That meant I couldn’t just convert all this stuff into money and be done: the weapons were literally worth more than their weight in gold, so selling them would make me even more weighted down, to the point that I wouldn’t be able to move at all.

The answer: I had to convert the money into bank notes, at a 10% fee. (Yes, those bastard designers implemented bank notes in their MMO.) But here’s the dilemma: the nearby town only sold tiny 100-coin banknotes.

Each of these weapons was worth thousands of coins. That would require a lot of 100-coin bank notes! And each bank note took up an inventory slot, so I couldn’t just buy a million notes. I had inventory slots for a few dozen, but not the hundreds I’d need.

So there I was with a complex dilemma: did I try to limp to the nearby town and exchange some of this stuff for crappy bank notes, or just go straight to a better city, which was more dangerous? Or should I drop a bunch of these weapons behind a tree somewhere and hope that I can get back to them before they disappeared? There were lots of other perfectly legitimate answers, too.

This was emergent gameplay. And it was fun! (That’s the thing about emergent gameplay: it doesn’t always sound like fun when you describe it.) This gameplay came about because the game implemented many extraneous systems: encumbrance penalties, item-count restrictions, bank notes, the ability to drop items on the ground without them instantly being destroyed, and so on. These various systems are why I had a dilemma.

“Fun Loops” Don’t Need Extra Game Systems, but Emergent Gameplay Does

If the designers were focusing on making the game fun in 15-minute loops, it would be a stretch for a designer to think “You know what this fifteen minute experience needs? It needs for money to have weight, so that if you pick up too much money you move slower.” Somebody would slap that designer.

When you’re making gameplay loops, it can be hard to take the time to make those extra systems. By themselves, they seem somewhat extraneous, until you’ve created enough of them to start to see interactions.

And even if you do make time for these systems, you’re kind of going against your goal! The reason you’re making loops is to ensure a consistently fun experience, and emergent gameplay isn’t reliably fun. Sometimes the gameplay that emerges is dumb and boring. If you’re making a game for a target audience that will wander away if they’re bored for 30 seconds, this is unacceptable.

Relying on emergent behavior means being willing to have bored players sometimes. But for the right demographic, the payoffs are quite pleasurable.

I’m not trying to recreate any particular experience in my MMO. I don’t think my MMO has any “weight” system at all, let alone “encumbrance penalties”. I think I can create systems that are a little more fun than that, maybe. But the point is that the game will have dozens of these simple systems in it, chosen for having a lot of interesting interaction points, which causes emergent gameplay to happen.

Emergent Gameplay and Realism

I’ve conflated some different things here. I talked about “realism” earlier, but you don’t need realistic systems in order to get emergent gameplay. Any systems that interact with each other can cause emergent behavior.

But for whatever reason, a game that models a bunch of real-world concepts will automatically seem more “realistic” to players, even if the concepts are modeled extremely unrealistically. (In Asheron’s Call 1, even with the “realistic” encumbrance system, I could carry more than fifty axes. How realistic is that?!) In other words, it’s purely an illusion, but it works in our favor!

So when creating  systems, it makes sense to use (and dramatically distort) real-world concepts like eating, sleeping, or banking rather than inventing systems that have no real-world analogue. You don’t have to go overboard and base every game system around the real world, though. A little bit goes a long way.

Emergent Gameplay = Fun Anecdotes

My favorite experiences in any PC roleplaying game come from emergent gameplay. Usually I don’t even remember the plots of those games, but I remember tons of anecdotes about being able to do some crazy thing by taking advantage of this other thing, and then this unpredictable thing happened, and it was so great. That’s the best-case scenario of emergent gameplay: a memorable anecdote where things came together and the player kicked ass by exploiting interacting systems.

Those anecdotes are very memorable because they happen on a random reward schedule rather than the minute-to-minute schedule of a loop-heavy game. Plus, they’re often entertaining enough to retell to others, which makes them all the more valuable to us. In a single-player game, those memories make it easier to sell sequels and add-ons. (The reason I was excited about Fallout 3 was because of my happy anecdotes from Fallout 1 and 2.) For MMOs, it means those memories can help improve your game’s rebound cycle.

Like I said before though, this design is not a mainstream choice. The game I’m making is much less directed than WoW, and WoW is like Realism City compared to most Facebook pseudo-MMOs. If you’re aiming for millions of players, the smart money’s on a directed game with tight controllable experience loops.

And just to be clear: I like those games. I enjoy playing WoW and even Facebook games. But that isn’t what I’m doing this time, because:

  1. I don’t think I can beat a major company at that design: they’ve been honing it to an art form.
  2. Facebook-style games are what I typically work on while contracting, so I’m kind of tired of making them.
  3. For an indie MMO, I need to stay self-motivated by making a game I’m excited about playing myself. And I do miss the systems-heavy MMOs of yore.

Hopefully a few thousand people per month will like the game enough to pay for it, and it will be a successful indie MMO. Perhaps a hundred thousand people will pay for my game each month and I will be rich beyond my wildest dreams. (I can’t even conceive of a scenario where a million or more people play my game.)

To be fair, there’s a very good chance the game will be a complete flop. But as an indie, I can afford to take that risk. And I can’t afford to play it safe.

This entry was posted in Design, Project Gorgon. Bookmark the permalink.

12 Responses to World vs. Game, Emergent Gameplay, and the Fun Loop

  1. Nils says:

    Wonderful post. To everybody who knows my blog it certainly doesn’t come as a surprise that I welcome a more ‘worldy’ game. I’d advise caution with the semantics, though. No fantasy game is ever realistic, no fantasy game needs to be realistic to feel ‘worldy’.

    Fantasy games become more ‘worldy’ when they are a more credible and consistent simulation of a fantasy place. While simulations of real life are realistic, simulations of Middle Earth are never realistic and shouldn’t be.

    Moreover, by using the ‘realistic’ term, you make yourself vulnerable to the ‘fireballs are not realistic’-crowd. That crowd is right: fireballs are not realistic. But they can be credible and consistent with your simulated fantasy world.

    One more thing. The simulation is what most games advertise. For example, WoW’s famous advertisement was “It’s not a game, it’s a world”. That is non-sense, especially nowadays, but it is effective. Very worldy games have a big advertisement advantage, because players don’t have fun with abstract gamey rules, but enjoy the idea to play a Night Elf Hunter in a simulation of the World of Warcraft.

    Had Tetris not been in the gameboy’s starter box, nobody might ever have discovered it. The game is too abstract to appear fun to anybody who hasn’t tested it. Advertising it effectively is almost impossible, as it is 99% abstract and only 1% ‘worldy’, or so.

  2. Fitdz says:

    I think your recent decision to drop the punditry has already re-vitalized your blog. Fantastic post.

  3. wufiavelli says:

    Remind me of richard garrette recently talking about how they create a housing market aspect of the game my accident.

  4. Eric says:

    Nils – I actually talked about that a little bit in an earlier draft, but cut it to try to keep the monstrosity under 2000 words. I agree with you about games not needing real realism in any way, but I also think it’s valuable to be able to hype “realism” (as opposed to just “immersiveness” — works better depending on who you’re advertising to), and the people who point out the unrealistic magic and so on are usually a minority. At least that’s the hope.

    Fitdz – heh, thanks!

  5. Jon says:


    As a player who broke my teeth on Asheron’s Call, as it were, I’ve enjoyed your blog since I first learned of it. I’m just a player, not any sort of developer … but if you manage to get through this project, I’d be quite interested in playing it :) (Or, you know, testing it … if you need that sort of thing).

  6. William says:

    Can’t wait to play it!

    …sounds great. The ideas that you have expressed embody how I’ve always felt about MMO’s. I can’t remember individual modern-day quests (tasks), but I think back fondly of specific events that have happened during my playtime that stand out from the norm.

  7. Stephen Veiss says:

    I’m another player who cut their teeth on Asheron’s Call, and to echo Nils’ comment, I would use the phrase ‘internally consistent’ not ‘realistic’ when debating the merits of AC vs other games. This applied especially to the next generation MMOs of the time, which made a fairly strong shift in the direction of ‘game’.

    I didn’t care that waving my fingers to shoot fireballs while carrying 50 axes and wearing a cloth robe buffed to infinity+ protective stats wasn’t ‘realistic’; I cared that all of those things fit within a mostly internally consistent set of rules and that I could use those rules to figure out new things.

    Modern games lose that consistency in lots of ways. To use the inventory example, player inventory in Asheron’s Call combined variable item weights, encumbrance, heavy coins/light trade notes, droppable items and a high item slot limit to give an internally consistent system. Players could use the rules of the system to derive new bits of gameplay — eg, an expensive but lightweight spellcasting currency, platinum scarabs, became a defacto currency for quite a while.

    Modern MMOs don’t have that sort of consistency. Some items you can fit one to a slot, some you can fit 500, others are secreted away in a little currency wallet and don’t take up room at all. There’s not much room here for players to figure out a creative solution to one of the symptoms of inflation.

    Asheron’s Call also launched with a monster population which was geographically distributed based on species/appearance. You’d find the white Mattekars over at the peaks of the mountains, and the Sand and Water Golems on the beaches. Each monster type had a fairly distinct loot profile, and almost all of the items looted were usable, somehow. There wasn’t the inconsistent split between grey vendor trash items and everything else — if you picked up a plate you didn’t want to sell for money or hang on your wall, you could toss it at the next monster you met to get their attention.

    Later updates moved the mobs to locations which split the game world into level/challenge based areas, and normalised the loot tables somewhat. Both of these changes made a lot of sense from a gameplay point of view, but I always felt that breaking the internal consistency of area->monster and monster->loot detracted from the fun of the game by taking away some of the flavour of the world.

    (Of course, I’m sure those two changes also reduced the data maintenance burden on the small live team considerably — something I appreciate much more now than I did a decade ago!)

    There are plenty of other emergent-ish systems which I miss from Asheron’s Call which I doubt I’ll ever see implemented in a modern AAA MMO, for both practicality reasons as well as ‘game’ vs ‘world’ concerns.

    I miss owning my lone cottage on the hill and the amazing fun I had when I discovered it was open and raced to be the first to purchase it. I miss sacrificing myself to bunnies to buff them, and I even miss fizzle dancing. I think all of those were emergent results of other decisions (no instancing tech for housing, monsters following a player leveling curve and able to spend XP, spell discovery by combining components trial-and-error), and the latter two show internal consistency in the systems involved.

    I hope someone will figure out how to bring those positive experiences back without the terrible landscape blight of the non-instanced groups of houses, long-living bosses buffing themselves to invulnerability and the pain that was spell research before Split Pea. To bring this back to the original post, internally consistent doesn’t mean consistently fun..

  8. Maria says:

    I’m going to chime in with those saying that ‘realistic’ is not the word you want to use. The concept as you explained it is very niche, but it’s also quite useful. Coin a new word or use one that isn’t usually used when talking about games.

    How about ‘convergent gameplay’? It sounds like emergent gameplay, avoids the idea that it’s purely concerned with AI or fighting, and emphases the interaction between several systems.

  9. Jon says:

    Stephen made a fantastic post. If you can give us something that looks adequate and features the same sort of convergent gameplay that we all loved about Asheron’s Call, you’ll have a customer here. I would still play Asheron’s Call now except there is some hella-old network code that means I disconnect and have to repair my installation every time I need to go through a portal … just can’t play that way.

  10. Ian Welsh says:

    I think immersion is the word I like to use. One immersion break I still remember was in WOW. In Un’Goro crater there used to be a lost expedition. Un’Goro was a pain to get to because there were no flight points, and in the old days before flying mounts, you had to ride there (generally on a 60% speed mount.)

    But it felt right. It was “lost world” area.

    Then they added a flight point, and they put it in the camp of the lost expedition, without changing any of the quest dialogue. So you had people whining about how to get back to civilization, and you’re thinking “uh, there’s a flight master 20 feet from you.”

    Hated it. Don’t break your own world.

  11. Ian Whitchurch says:


    Did you ever do stuff with RQ ?

  12. Pingback: NPCs as Systems Nexus | Elder Game