Ten Lessons Learned

This talk was delivered at Konsoll 2018, held in Bergen, Norway. 25 years making games has given a lot of hard-earned lessons. Here are ten of them, spanning a range from design tips to teams to business advice, drawn from real-world game development situations.

Slides

Video

PDF

Transcript

Ten Lessons Learned

HOST: Okay, so our next speaker is a veteran game designer who has been professionally credited in almost every area of the games industry, and he is widely recognized as one of the world’s top thinkers in game design. He’s here to share some of the lessons that he’s learned from his endeavors. So please welcome Raph Koster.

[Applause.]

RAPH KOSTER: You hear me? Ah, there we are. Hi. I’ll just wait for the slides to come up.

I have to tell you, looking at the crowd, I feel old. It’s been about 25 years since I started making video games for an audience. How many of you are 25? A couple of you. Yeah. So I feel old.

I hope that you don’t mind if I indulge myself a little bit as we do this talk. I thought I would just pick ten things, ten lessons, ten bits of information that I wish I knew when I was 25, and I’ll tell them to you. You won’t agree with them. Or maybe you will, but it won’t matter, because I’m not sure that lessons like that are the kind of thing we can learn from being told. They’re the kind of thing that you have to actually go through in order to really absorb.

Hopefully we have the slides. Otherwise this will be a much looser talk.

My name is Raph Koster, and I’m probably best known for doing MMOs. I started out in the days of text-based MUDs. Has anybody here ever played a MUD? Oh good, a bunch of you. This is the wrong talk.

Ten lessons learned. It was set up on the other laptop over there. There it is. Hooray.

I started out on MUDs, and I was lucky enough to be the lead designer for a game called Ultima Online. Did any of you play that? It came out in ’97. See, you’re young. They’re young.

Then I went on to do Star Wars Galaxies, and I’ve ended up working at EA and at Sony and at Disney. I founded my own studio. I had it bought. I’ve had C-level titles, and I’ve been a grunt designer. I’ve done soundtracks. I’ve done graphics. I’ve done the game design. I’ve done business. Almost everything now.

And I’m tired.

So let’s just—I just want to indulge myself and share some lessons here.

The first one really has to be this: the only way to do all of those things, and really the only way to keep growing in this career, is to keep learning.

You can’t stop. If you stop, you fall out of date in about three and a half weeks and are no longer relevant.

The industry moves fast, and it can seem like there’s too much to learn. The sad news is that that is true. There is in fact too much to learn. You can’t learn it all.

I was lucky. I started my learning when I was really, really young. I started playing Dungeons & Dragons and being a DM when I was like twelve. That was the early 1980s. I was born in 1971.

Look at those lovely mountains. Remember what those look like. It’ll be relevant later.

I started making board games because I loved playing video games, but I couldn’t make them when I was twelve. So I started making board games, and some of them were actually ports of video games that I loved to board games that I would then take to my school, and we would play with the other kids during breaks.

I didn’t know it at the time, but I was teaching myself game design, because it turns out that porting Pengo and Q*bert from real-time action games to turn-based games is hard. If you asked me to do that today, I would consider that real work. I would charge you for it.

But I was doing it for myself, for my friends, for fun when I was twelve. And I learned an awful lot about game design by doing that.

Eventually, of course, I did get an 8-bit computer with nice floppies that actually floppy. Floppy is so old that they don’t look like a save icon. I started with the 8-inch ones, actually.

You can see here from a very early age, I branded myself as Super Game and started learning marketing in order to tell other kids at my school—you can see that one says, “This one is already a classic.” So I was already learning the tricks of marketing.

Some friends of mine and I, when we were about fourteen, formed a company that we called Protocom. You knew it was official because the splash screen for the game had a copyright symbol next to the name, which was legally wrong, but we thought it meant something.

This was one of the first games I made, and in hindsight it actually set the pattern for my entire career. I wrote it mostly in BASIC with a little bit of assembly language, although not at first. Not until stage five. On the Atari 8-bit computers. We tried porting it to the Commodore. It didn’t go so well.

We were actually trying to sell games in Ziploc baggies to classmates. We sold one copy. It was awesome.

I started out by cloning, of course. I cloned Snake, which was not Snake at the time; it was Tron light cycles. Snake is actually a clone.

Anyway, I started out with light cycles, and then I said, “Oh, that was fun. Now I want to make a shooting game.” I didn’t know how to do collision detection, so I put the two ships at the opposite sides of the screen and I just checked if they were at the same height to see whether or not you hit the other guy.

Little did I know, that is in fact half of how the collision algorithm works.

Eventually I did figure out collision, but I didn’t know how to make bullets travel independently of the ship movement, so all actions stopped while very slowly that laser beam went across the screen. I couldn’t make pretty explosions, so I used X as an explosion. That’s how you died.

Then I had a brilliant idea. I should have patented it. It really was a good idea. This did not exist at the time. There were no games on the market that did this: capture the flag. I made capture the flag.

All of these were two-player games. They were multiplayer. They were couch-play games. These actually all connected. When you won a match in one, it loaded the next one off disk. So this was a five-stage head-to-head combat. Right there, capture the flag: who can get to the flag first?

Finally, I figured out how to actually do real graphics with sprites, and that is where I first started doing procedural generation for levels, when I was fourteen.

By the age of fifteen, I had already started, without knowing it, putting myself through a course of study.

Today that means that my bookshelf looks like this. This is not all of it. This is only two shelves of an eight-foot-tall bookcase of books that are relevant to the kind of games I make. Since I have mostly made online communities, games with connected audiences, that means it is full of books on law and on economics, books on music theory and composition, books on service, books on marketing, books on film.

Because all of these are fields that are relevant to making games. Games like to eat all the other fields and swallow them up and take them inside of games and blend them into something new.

Over the years I’ve continued to try not just learning, but figuring out what else to learn. What else do you need to go after? I currently have this list. I call it a map of game. I’m going to walk you through it just to show you how big this can be.

I did a book a while ago—it’s been quite a while, about twelve years, I don’t know if any of you have read it—called A Theory of Fun. In that book I wanted to figure out what fun is for a very simple reason: a bunch of players had told me that my newest game was cool, but it wasn’t very fun.

So I wondered, well, wait, what does that even mean, really? We hear it, but what does that mean? So I set out to figure out what fun was, and I wrote a whole book about it.

I concluded it’s about learning. It’s about forms of deliberate practice, although players do in fact use it for other purposes. We leverage fun in order to get exercise in the case of, well, pretty much any hobby that you do that involves sport. If you go hiking, if you like cycling, whatever.

Players will fall into games and use them to meditate. You’ve all done that, right? It’s relaxing. We use it to solve boredom. We use it to fall into flow.

But there are whole books about flow. Flow is a large, complicated topic. Exercise is a large, complicated topic. So books found their way onto my shelf about sports training in order to help understand what was going on there.

What I arrived at was the conclusion that players are looking to solve problems. They are engaging in balancing resources and juggling timing and solving problems as they play. The games are always teaching us things. They teach us themselves. That’s what games do, and that is where the fun comes from.

That means as we proceed, we form what are called heuristics, which is a term from either computer science or psychology, depending where you prefer to look, which are interim solutions that are good enough for now.

That’s kind of the definition of playing a game: all right, I can do this for now. Then comes the next level, and you can’t do it anymore, and you need a new solution. So you fall into a standard human-computer interaction loop, where you have to form a mental model for what’s coming and figure out what kind of verbs and affordances are there.

That leads you into the realm of Don Norman and usability in a whole new arena, in order to identify the verbs and the inputs and the actions that lead into the actual game, what I call the black box.

When I started, there were no game design schools. In fact, when I started at Origin in 1995—long-lost Origin, makers of Ultima, swallowed up by EA, digested as EA does; the list of studios that have been eaten by EA is long and sad—when I started at Origin, they had just created the title “game designer.” Before that, they didn’t think designers were a job, because before that most games had been made by programmers. You had some artists, but they were replaceable cogs. Some things never change, EA.

So the scope of what was inside that black box, what it meant to be a game designer, wasn’t really very well understood. We didn’t realize that it was insanely complex, hugely complicated, that there were all of these enormous questions about what is good game design and can we get better at it.

It’s hard to comprehend now, but there were enormous flame wars online over whether or not games were an art form, and most game developers used to say no. They thought that games were a product. They thought the games were, you know, just for fun, just entertainment, and there isn’t anything truly meaningful about them.

That’s really, really wrong. But it took a long time for the industry to be persuaded of that.

Gradually, of course, then you start learning about how to convey the things that you teach people. That leads you into the realm of art. It leads you into the realm of narrative. It leads you into the realm of UX, which wasn’t a term even around eight years ago. UX wasn’t a thing, and today it’s a job name.

Finally, you reach the point where you realize games affect culture. Games teach empathy. Games do in fact reach for art. So I went and got books about every field on this diagram, and yes, I’ve read them. I didn’t just get the books. I read them.

There’s just a lot to learn. It can seem intimidating. It can also seem kind of pointless. “Well, but I do just fine without learning all of that.” Most people do. That’s true. That is true. You don’t have to learn all of this. I’m just saying it helps.

It helps to do it with other people. It matters that there is a scene around you of others who are trying to learn alongside you.

I was lucky in that I got started in text MUDs in the early ’90s, precisely when there was a boom. I was lucky in that text MUDs—this one, by the way, we started in 1993, so 25 years ago—I got lucky that Marc Andreessen put pictures into Mosaic or whatever, and then came Netscape, and the internet became graphical at the exact same moment that I was making games that were ready to become graphical.

Before this, not only was there no Google, but there were no pictures. In fact, there was barely a web.

I remember when the web started.

We ended up forming a tribe. We hung out on Usenet. Any of you even remember Usenet? Yeah. Google—it’s called Google Groups now, which is amazing. They managed to basically buy the entire archive of what used to be the discussion forums of the internet. They are all now locked up inside Google. But there are decades’ worth of messages from Usenet, because that messaging started in the late ’70s.

We ended up creating a mailing list, this community. I didn’t do it; there was a community there of people who believed that the creation of virtual worlds would lead to completely changing the internet, because we were all positive that by the year 2000 all of us would be walking around with brain implants in a perpetual AR/VR world.

We absolutely thought all of that was coming way faster than it actually was. We thought it would be cool if on Amazon you actually logged into a virtual mall that you walked in as an avatar and looked on shelves. We were really stupid.

On this mailing list, we had the people who were creating the MMO. We had Jake Song from Korea, who went on to do Lineage and did Dark Ages. We had the Funcom guys. We had the Turbine guys, who were doing Asheron’s Call. We had the Verant guys, who were doing EverQuest.

There were no guys doing World of Warcraft yet, because Blizzard only jumps in after people figure stuff out. No, it works very well for them. That’s not a knock against them. It’s sound business strategy.

This mailing list is where the ideas that ended up shaping the next several decades of online games came from. Our bosses did not know that we were on a mailing list swapping trade secrets, telling each other how our algorithms worked. And we did it anyway, because that was how we learned. That’s how we got better, by saying, “Oh, well, I have a better idea than that. Have you tried doing this? Here’s the algorithm for that.”

The archives for this are still online. We even started up our own conference. It ran alongside GDC, either the day before or the day after. Some of the folks here—see, some of them were old already. There’s me when there was less gray. Brian Green, he was one of the designers on Meridian 59. In the middle is Richard Bartle. He invented virtual worlds in 1978. Everything that you think of as MMO or multi-user online gameplay, he was pioneering that back then.

Next to him is Randy Farmer. Almost every technique you know about community comes from Randy. Next to him is Pavel Curtis. Pavel ran LambdaMOO, which was probably the first real experiment in granting players governance of an online space. So no Pavel, no Reddit, right?

We held our conferences, and we took notes in the forms of cartoons, and we tried to share what we were learning. If we hadn’t been in a tribe, if we hadn’t been in a scene, we wouldn’t have been able to do it.

I wanted to give this lesson to you particularly because I happen to live right now in a city that doesn’t have a scene, and I really miss having one. So look around you. See the other folks who are near you and start making friends, because it’s one of the most valuable things you can do.

This is what my office looks like. One of the things about being in a scene: it’s a lot like being a musician in a band.

How many of you have heard of the punk club CBGB in New York? CBGB was at the heart of a scene. No Talking Heads, no Blondie without that club. They were hanging out together, listening to each other’s music, and swapping ideas.

The punk scene in London, for that matter, the blues-rock scene in London that gave us pretty much every famous British band: bands come apart, and that’s also a lesson.

When I first stopped working with the people that I had worked with first on LegendMUD and then Ultima Online, because it was the same people, I wondered whether it was all over. The band had broken up.

It turns out there’s a lot of great lead guitarists out there. You can probably hire another drummer. But you do have to be in a place where musicians like to hang out. You have to be hanging out with them so that you know, “Oh, this is somebody I think I might want to jam with someday.”

Of course, the other part—and this is especially true these days as indie has risen—is that a lot of musicians starve, and we need to remember that too.

Now that we know that games are art, we should remember that historically artists do it for love, not money. It’s incredibly important to remember to make games for love and not for money, even if you make the money along the way.

There’s a lot of reasons for that. It’s important to be invested more in the craft than in the outcome. It’s important to love the process more than what ends up happening.

Okay, I haven’t played Fun Run. I’m sorry. And I’m sure it’s awesome. One of the biggest lessons I have in my career is that awesome things can be made and completely ignored. It happens all the time.

I’m very aware of the fact that I was lucky, lucky, lucky, lucky. Fun Run was lucky. Blizzard was lucky. Once they made the pile of money, they didn’t need as much luck. When they started out, they got lucky.

The fact is that is where popularity mostly comes from: from getting lucky, being in the right place at the right time. You can prepare for it. You can hope for it. You can work really hard so that when it happens, you have success. But the fact still remains that luck is an enormous part of it.

So all of the attention towards the craft is so that you stay content and happy and loving what you do regardless of whether or not you get lucky. Hopefully you do it because you love it. If you’re in games to make a quick buck, you are in the wrong business. That’s not the motivation to be in this field.

Craft does help you make games more predictably. It helps you figure out schedules. It helps you make a more fun experience, because there are rules to making fun experiences. We can now analyze games in a way that we simply could not 25 years ago.

This is what’s called a game graph. I call this a game grammar diagram. There are a bunch of these different systems out there. Machinations is another one, where you can do things like analysis of economic flow, figure out game balance in advance, that sort of thing.

Attention to craft: if this isn’t the stuff you love, if this isn’t the kind of problem that you love digging into—well, this is what games are. This kind of flow of resources, the balancing of the numbers and the objects that you have, and the setting forth of challenges packaged up in a way that players can actually learn from. Underneath, games are really pretty mathy.

Learning to look at games in this way, separating yourself from it, will help you start to see market patterns as well.

For example, as an exercise back during the Facebook game boom—I worked in Facebook games for a while—I actually built graphs like this for all of these games, and I noticed the trend line in terms of how the games were growing more and more complex over time.

I don’t know if any of you have ever heard of Dan Cook. He’s a game designer at Spry Fox. They make Alphabear, Road Not Taken, a bunch of other really wonderful games. He had an observation that all game genres go through the same cycle: they start out simple and easy and accessible, and then as players learn how the game style works, developers start adding more and more features into them. That makes them richer and deeper and more complex. Unfortunately, it eventually means that nobody can understand how to start the games anymore, and the genre starts to decline.

This is very, very true, and I watched it happen. I’ve actually watched it happen multiple times. There used to be a wargaming scene. Flight sims used to be a genre, and now if you try playing a flight sim you need like $500 of equipment to get started.

Designing for others in that way means paying attention to the sweet spot of simplicity and accessibility. But it also means being aware that what you want to make isn’t necessarily what the audience needs.

These days, of course, one of the things everybody has figured out is that you need to pile on microtransactions and free-to-play mechanics and DLC and retention mechanisms and all of that sort of thing, and multiplayer. Over my career, I was a big part of pushing some of those trends. I helped make those things happen.

But it’s easy to lose sight of the fact that friendships on Facebook are shallow and mostly meaningless, that true human contact—we only, on average, have five real close intimate friends, the kind of person who helps you bury the body. This holds true across the world.

It’s important for us to work with actual humans and not with our ideas of what a consumer is. We should be making games that make their lives better. Sometimes that might mean designing for smaller. It might mean designing for unpopular things like couch co-op or whatever.

This, by the way, is from research called the Trust Spectrum. That is something that I worked with Google over the last two years on. So if you’re interested in multiplayer dynamics, there’s a whole bunch of stuff. Google “Trust Spectrum.”

It means becoming aware of the fact that the business is cyclical. Every once in a while, a new platform comes along and it offers a new interface: accelerometers, or touchscreens, or the fact that you can actually play a decent game inside a web browser with something like Flash.

What that means is a whole bunch of young hungry people show up and make mostly really crappy games for a little while while they figure out how do you make something fun on a touchscreen. So it’s super cheap, and we get these tiny, scrappy indies, and they make something completely new, something we’ve never seen before, and it’s awesome.

Then that thing becomes super popular, and a couple of those indies make an insane amount of money and turn into Zynga or Supercell or King or whoever. Then they spend all of their previous companions into the ground to prevent them from being able to crack the market anymore, which is kind of where we are now.

So costs start to rise, and tiny guys can’t compete, so a whole bunch of them get bought or they go away. But luckily somebody invents a new game platform, and the cycle just continues.

We’re going through a transition like that right now, and I think it’s important. The lesson I would give you all is, the pendulum will swing again. New platforms will come along.

When the platforms are new, we get games like the ones on the top. Any of you play any of these? Pong? M.U.L.E.—literally the best game ever made. Street Fighter and Doom and Canabalt and Minecraft are games that are driven by systems, not by narrative and story and experience.

Then when budgets are big, it turns out that character and narrative and experience are market defensible, and systems are not. It turns out you can advertise like crazy with cool brands like King’s Quest, or Ultima, with the Mysts and the Zeldas and BioShocks. It turns out you can tell amazing stories the more cinematic you get, like Edith Finch.

People grow up during cycles of the top half or the bottom half, and it turns out they don’t like the other kind of game. That pendulum is also something that swings. It’s just cycles, and the more of them you go through, the more you see it.

I have found that the retreat for me is to retreat back into the fact that game design is game design. It doesn’t matter if you’re designing a new playground game, or a tabletop game, or a video game. It’s still the same underlying mechanisms. It’s still the same stuff underneath.

I try to remember that in order to succeed at this, you have to end up being ambitious and practical.

We made Ultima Online on servers that ran at 500 megahertz. It was a map 4K by 4K, four kilometers with one-meter resolution, and we had 27,000 times less bandwidth than you do at home today. Each packet was 400 bytes, and we had to tune it by hand. Yet each server had over a billion dynamic objects on it running in a fully simulated world, because we were young and stupid and didn’t realize we were building something impossible.

Notice how the mountains on the map are still the same.

Being ambitious yet practical, realizing that it’s about solving these problems in a way that can be efficient and clever and smart, is what ends up giving you the path forward into the future.

Star Wars Galaxies: we had the requirement of launching with ten planets, and each of them was 16 kilometers on a side, this time still with one-meter resolution. Once we did the math, we realized that meant we were trying to ship four gigabytes just of map data. People did not have DVD drives yet, so we invented a system using layers of Perlin noise, used as filters and as generators, in order to create maps like this in 2003.

We shipped ten planets with four gigs worth of data because we had to solve it somehow. Today, of course, you get a No Man’s Sky out of pursuing that line of ideas.

Being practical means realizing that you’ve got to keep it simple. You have to find ways to really focus down on what it is you’re trying to achieve.

Later on, I did a web-based MMO system called Metaplace. We built amazing technology. Metaplace was the backend for The Matrix or Ready Player One. It was that powerful, and that meant we had no idea what it was for. So it failed.

Because we were ambitious but not practical. That’s what happens sometimes. You have to learn to find that balance.

Part of finding that balance is realizing you’re not in control, and you just can’t be. You won’t be.

My degree is actually in creative writing, so that means I was trained in being an entertainer who was in control. Authors have more control than almost any other creative field. It’s awesome.

When I started out way back in text MUDs, I spent all of this effort doing things like adapting the sagas of Beowulf and the Nibelungs into one connected giant quest with amazing writing and all this immersiveness and all the rest of that.

What did players do with this? Can you guess? No, they couldn’t skip it. You couldn’t move through it without it. But they could break it. They broke it.

That’s the thing. Breaking things is actually a huge part of how children play. That’s how everybody plays, by messing around and breaking things. You need to learn to embrace this.

On Ultima Online, we had a bug. As we were populating the world, we needed some things to ship on the CD and be immutable static objects, and we needed some things to be streamed and updated dynamically from the server: dynamic objects.

We went and built the map, and there was just a little checkbox toggle for when you were building and laying things down. We needed the candelabra that you could light the candles on to be dynamic, but the walls needed to be static, because otherwise players could pick them up and walk away with them.

It was really easy to forget to tick the checkbox, and a few times we built entire cities one way and had to delete them all and build them the other way. But it turned out that we left an awful lot of stuff dynamic by accident, including little holes in the water where we had put down dynamic water tiles.

So players moved their mouse over every pixel on every screen in the game, found the water that they could pick up, picked it up, put it in their backpack, and then would fish out of it.

Because the game was a simulation. Every object knew how it worked. It ran off of an abstract resource property simulation system and off of an artificial life model for all of its AI.

This turned into a rares market, where people started eBaying these objects that you’re not supposed to be able to pick up. eBay had only come out like a couple of months before UO shipped, right? UO was also the first game with a reputation system of any sort, because rep systems had only come out on the web about two months before we shipped. Everything was being invented at once.

So we didn’t know that these rare objects would start selling for a thousand dollars on eBay. eBay was for Pez dispensers, not for virtual objects. Pretty soon people started, “Wait, digital currency, could that be a thing?”

We didn’t know that would happen. But that led to gold farming, as people started making a job out of gathering rares, and then they would take the rares, and then the companies said, “Wait, we should just sell the rares directly.” That is how microtransactions were invented, because we forgot to lock down the water tiles.

So from control to total loss of control. But you actually need to embrace it. You need to lean in on that, because players will do it anyway. Players will speedrun. Players will make chess problems. Players are what have kept chess alive by making chess problems, not us.

Players end up driving all of this because we change them when we make games, and we end up changing the world, at least in small ways.

Any of you anime fans? Yeah. Any of you see Serial Experiments Lain? Okay, so this is an anime where basically it’s a cyberpunk anime. The internet comes alive, virtualized people, whatever, and she gets sucked in.

And of course Sword Art Online is one where players of a virtual MMO get sucked in, and it turns out the world becomes real.

The first time I saw Lain and the characters said, “Oh no, there are PKs around,” I almost fell out of my chair, because I was the one who gave us the term PK. When I was on Ultima Online, we needed a way to explain to players the difference between PvP and PK. These were terms that had existed in the MUDs, but game players didn’t know what they were, so I popularized the term.

In fact, a whole bunch of the terms that people use today I am personally to blame for. The word sandbox, me personally, right? But that’s what happens. We end up changing the world by accident in these random little ways.

When MUD was first created, Richard Bartle came from the wrong part of England. He had the wrong accent. He didn’t go to one of the fancy schools. So when he and Roy Trubshaw made MUD, they wanted it to be egalitarian. They didn’t want it to be a place where only the upper class could succeed.

That is why you can advance by levels in MUD, and why anybody could become a wizard, because he felt everybody should have a chance.

You’ll notice here that every player has two names, the real name and their character name. I want to point this out: Richard was playing Polly. People said, “Richard, you look like a girl,” and he said, “Online, anyone can be anything. I’m doing it on purpose so that people understand what virtual worlds are for.”

Think about the ambition in that, that he decided to do that in 1978.

We were arrogant and young. We knew we were changing the world. We did things like write declarations of the rights of avatars, and we got laughed at mostly. Of course I wish today Facebook had this, because we have no real rights on Facebook and Twitter.

But sometimes the change that we want to see doesn’t really happen, because change is kind of unpredictable, and the ways in which you impact the world can be kind of random.

Some of those things, like the cool “hey, let’s do online reputations,” turn into the horrifying Chinese Sesame Credit system, where they rate whether or not you are politically in line enough, and it affects things like your scholarships and whether or not you can travel within even China.

It turns into things like Cambridge Analytica. Our analytic systems for avatars turn out to work really well for all kinds of online profiles, and now we have metric systems that are affecting elections in the course of world history.

We have taken all of these tropes of online games and turned them into our internet, mostly without the rights documents and the idealism. We can’t always control the way in which these things happen.

That’s because we need to remember as we work that if we only create based on the other games, we’ll just end up self-referential and not addressing the world as it is.

I’m sure that everybody knows that Miyamoto is a huge fan of gardening, because his games are all set in gardens. All of them. And he clearly loves mushrooms especially.

We all know that Will Wright is fascinated by robots and by how things work and little agents, because every single one of his games has been, in a way, about human robots.

They both draw inspiration from things outside of games, and I’ve tried to remind myself to do that. Being a fan of games is not at all the same thing as being a maker of games. Many of the most powerful games we have come about because we look outside of games for inspiration.

Ultima Online was actually also inspired by robots. This here is a cockroach robot. It’s scared of the light, and it was an experiment at MIT that was being done in the late ’80s. Put a photo sensor on the robot, put a bunch of them on Rollerblades, and then turn the lights on in the lab, and they all run and hide in the corners.

That is how I decided to put artificial-life AI into UO.

Star Wars Galaxies was inspired by reading that I was doing about social graph systems, about the ways in which humans cluster and form networks. That is how I realized, wow, our online games are really terrible if all we have in there is people who kill each other. Why can’t people have a full range of human experiences? Shouldn’t they be able to do other things?

Players were falling in love and getting married and having online weddings and all of that, but they were doing it in spite of the game system, because the game system was telling them, “No, go kill more orcs. Go get more levels.”

So instead, we put in leveling systems for dancing, for going camping, for raising pets.

UO was the first time there was a crafting system in a video game. It was the first time there was character customization in a 2D video game. And on Galaxies, Bethesda beat us to market with 3D character face shifting using morph targets by a few months, and I was really mad.

But it was about enabling more forms of human experience, and it came because I was reading about how society actually works.

But with all of that, you can look at the real world, you still have to remember—and this is a tough pill for me to swallow, and it’s one that I’ve only really acknowledged to myself really recently—because if you’re in this for the craft, if you’re in this for the art of it, you want to make things that leave a dent in the universe, as they say. You want to make things that change the world.

And it feels sometimes like any of the ways in which I managed to change the world in small ways were mostly by accident, and any of the ways that I wanted to change the world didn’t happen.

When I’m honest with myself, when historians a thousand years from now, assuming we don’t cook the planet with climate change, assuming they look back at the history of video games today, this is probably the only thing that will be in the encyclopedia entry.

I mean, how many of us get to make chess? How many of us get to make Tetris?

Leaving a mark on the world that profound is really hard. You can dream and aspire to it as an artist, but that’s why it’s important to love the process more than the product. Because if you put your hope in the product, time will disappoint you.

Most of you don’t remember what Ultima Online was, and it’s gone. It’s still there. You can go play it. It’s twenty years old, and nobody wants to log in. Galaxies is shut down. You can’t play it except players are trying to keep it up as a museum, which is something people try to do.

There’s Randy again and his programming partner Chip Morningstar. Their little dent in the universe was making Habitat, the first graphical online world, in 1985. They and their other programming companion Doug Crockford had moved on to invent a bunch of technology along the way.

Doug is of course best known today because one of those technologies he invented to solve virtual world problems was JSON. In fact, large quantities of JavaScript in general. Chip is now on the JavaScript committee.

So they left a small dent when they made Habitat. But Habitat today runs in a museum. And what good is an online world, an online community, when its entire population is the people who happen to be at that museum kiosk at a time? It’s about as alive a city as Pompeii.

We can learn things from it, but it can never be what it once was, because an empty city is not the same thing. There’s some sense in which preserving the work that we do is impossible, because the players are more than half of it. If the players aren’t there, then it’s not there anymore.

So instead we have to think about how we stack game on game.

I was given this cool award, which is awesome. Yay. You know, I can actually—I’m a legend now because the top award says so. But on my flight here, I bumped into a guy in the airport and he said, “Oh, so what are you doing?”

I said, “I make games.”

“Oh yeah, what kind?”

I said, “I make MMOs.”

And he spontaneously said, you know what, he didn’t know who I was, but he said, “Star Wars Galaxies is my favorite game of all time. It’s how I met my girlfriend. It’s where all of my friends came from. I tried to play the emulator. It sucked. It just wasn’t the same anymore. The game’s really dated,” and all the rest of that. “But you gave me a decade.”

You know what? Screw the award, because that in the end is the kind of legacy you leave behind.

That said, it’s been a while since I got an award, so I’d love another one.

The only way to really advance that way is to share the learnings. I tried calculating, and I think this may be the 70th unique talk that I have given. A lot of these talks are not small. A lot of these talks are year-long research projects, inventing things like game grammar, trying to determine whether or not we can do social graph networking in online games, trying to figure out what the structure of business cycles is or whatever.

That’s my warning that I should start wrapping up.

And, you know, writing the books.

In the end, sharing the learning and helping others along this journey may be more important than the games you leave behind, because the games, again, will probably be meaningless, and the people and the culture you build is what can last.

In the end, I said I’d leave you ten lessons, but I ran out at nine. Then I realized, well, the tenth should probably be there for everything I just told you is really just like the mud that they’re digging up from under the buildings at Bryggen. I’m probably not saying that right. But they dig it up in order to raise the buildings and keep the buildings around, and then they go look through the mud.

There’s actually a little sign next to all the broken pottery that says, “What we find usually isn’t of archaeological significance, but it’s still good to know where we came from.”

Sometimes I think maybe that’s what really matters. Our destiny is always, you know, if people stand on the shoulders of giants, then that means that your destiny, because you’re all giants, is to get stepped on, and somebody else stands taller because of you.

Ultimately that means that what you should do is ignore me and stand on top of me and squish me. That’s what should happen, because that is how we actually get anywhere and how games will continue to evolve.

At the same time, no, please don’t do that.

Instead, what I think I would urge us to do is: get yourself an emulator cabinet at your place, at your home. Go back and try playing the games starting in 1973 and realize how far we’ve come and how much we stand on.

Go back and look at people like Roy and Richard, who had that ambition to make the online world a more idealistic place. The people who thought that they could do a graphical MMO on a Commodore 64. Think of people like that.

If you loved M.U.L.E., think of Dani Bunten Berry, literally one of the greatest game designers of all time, and most of you probably have no idea who she is. The first head-to-head modem games, pioneering couch co-op games, some of the first pacifist games ever made. M.U.L.E. sold 30,000 copies, and Dani died in 1998 saying, “Nobody on their deathbed ever said, I wish I’d spent more time with my computer.”

Because she understood that games are social. Games are about the people. What we do is about the other people.

In some ways, as much as I say invest in the craft, ultimately the games are about the people.

So look around at the other people here. Build your scene. Get stepped on, built on, whatever. And try to remember these lessons twenty years from now when you go, “Oh, he was right.”

Thank you.

[Applause.]