The Last 10: Going From Good To Awesome
Benson Russel of Naughty Dog talked about how they polish their games. As he said this, everyone in the room thought to themselves "Blizzard...Blizzard..." But actually this was about Uncharted 3, and not Starcraft.
Benson explained that polish is not something you just hope happens, that it has to be scheduled and planned for. He showed a diagram of a normal company's production schedule. It has some pre-production, a very long production phase, then alpha, beta, and ship. Then he showed Naughty Dog's version of this, with the same total length because the point is the relative lenghts of each phase. They have a much shorter production phase and a longer alpha phase and beta phase. There is even an additional "hands off" phase after beta and before ship.
The shorter production phase ensures that any core mechanics are figured out even sooner than they sometimes are at other companies. He also said this kind of scheduling requires a "hard alpha," like you can't half-ass it. The alpha really does need to have everything there in a rough form and the entire game playable from beginning to end. Once it's all stitched together enough to play it all through, and everything in place (at least roughly), alpha becomes the polish phase and for them lasted 4 months on uncharted 3. He said they would like it to last a bit longer next time. I think beta also lasted 4 months, but I'm not sure on that. The "hands off" phase is actually after the polishing is done. At this point, it's only the QA department and programmers playing it, and the only purpose is to find major showstopping bugs. On Uncharted 3, during this period they actually did find a very difficult to detect loading bug that would have caused the game to crash on 25% of the Playstation 3s out there.
To show us what "polish" means, he showed several example videos of Uncharted 3. He had videos showing a bug, and then showing the same scene when the bug was fixed. It was often small stuff, but that's his point, really. That you can't let a bunch of small bugs drag down the overall feeling of the game, it all adds up. He had to play most of these bug videos twice so we could even *see* what the bug was. In one, the main character is in a tunnel and sees rushing water coming, turns from it, and runs. As he turns, there is an animation glitch (over in a fraction of a second) before he runs. In another example, the scene starts with a stationary camera, then the camera moves in to show the action. The bug is that during the part where it's stationary, there is accidentally one frame (yeah, one frame) where the camera moves forward. In another example, an NPC throws an enemy to the ground, but the enemy is then right under his feet, which doesn't really look right or make any sense. The corrected version has the enemy fly a couple feet to the side, so he's not directly under the guy who threw him. Another example showed that one scene where the main character falls into a new environment has the background ambient lighting set to a wrong value during part of the fall. And yet another example showed a scene where for just one frame, the screen was pure white for no reason.
So yeah, they fix all this stuff during alpha and beta. And they even have a person on the team, not the QA team but like the real team, whose job it is to look for these polish issues. They have a fairly high ranking person do this because it's not just about spotting the issues (though apparently he is great at that), it's about making a judgment call about how to fix or what to fix. There is often the issue that fixing something might be really easy or really hard (is it worth it?) and the other issue that fixing something might be low risk or high risk (is it worth the possibility that the fix could fuck up other parts of the game?).
Benson also explained that as alpha and beta go on, it is (intentionally) harder and harder to get changes approved. At first, it's like a free-for-all. Everyone fix everything they can, go! They monitor bug counts and everyone does their best to keep their bug counts down to at most X bugs, set by their project managers. By looking at these bug counts, they can see if any particular team member is overwhelmed, and maybe needs help fixing stuff, and if another team member doesn't have many to work on. Also, seeing the rate of fixes helps them estimate if they are on track to ship on time or not.
So at some point in all that, they institute a rule that from then on, they have to be more careful about making lots of changes because ship date is getting closer. So you have to get the approval of one of three executives (including my friend, game director Justin Richmond, haha go Justin) to change something. A bit later on, you need two approvals. After that, you need all three of them to approve any change. This is just good sense, if you've ever been on a software project before. They are reducing their risk screwing things up at the last minute this way.
BURN THIS MOTHERFATHER! Game Dev Parents Rant
In the annual rant session hosted by Eric Zimmerman, developer's break out of their prepared scripts to tell us passionately what's bothering them. I think this is an important part of the conference overall, because we get under the surface about what's really going on. Except...we totally didn't this time and it mostly sucked. I am more upset about three things before breakfast than half these people were about whatever they are supposedly "ranting" about.
Graeme Devine started us off just fine, at about 80% or so on the rant-scale. He is sad that the world of game development that we're passing on to his daughter sucks. He blames us, the veterans who shaped this industry. He says a good thing we have going is that our indie scene is actually doing interesting stuff, and reminding us that it's not all first person shooters or whatever. He says it's good that the IGF celebrates indie games and that it actually gets the winners a publishing deal on xbox. We *should* be encouraging our new blood in this very way, so that's great! What isn't great is when those same innovative newcomers go on to make their second game. At that point, they are given the same shitty treatment that the rest of the industry gets, which he incidentally thinks sucks too. He says in our industry today, if you want to make something on the scale of a couple million dollars, people are so paranoid about risk that they're only willing to back FPS, RPG, and RTS. (Editor's note: uh, I don't know about RTS. I mean that ship has sailed and Blizzard won. But yeah I get the point.) He's mad that interesting projects are exactly the ones that deserve a couple million dollars, not the things that actually get funded.
Then there were really disappointing non-rants. One about how someone's kid plays games too much and how she tries to set limits like no games mon-thurs, or tries to withhold food, or unplugs his internet connection. She wonders what external rewards she can use to get him to do real-world things, given how hard the competition is from rewards inside World of Warcraft like new gear that she's competing with. Uh...what? How about it's great that your kid plays games and he's learning a lot from that actually and that he might turn out just fine. I played games and didn't have a mother busting my ass about it all the time. Also, if you think external rewards are what he needs, rather than internal motivation, I don't even know what to say. But if you want a passionate rant, then it should have been me eviscerating her entire presentation.
Then there was some boring thing about some boring thing.
Then it was Jason Della Rocca's turn. Note that this session started with Eric Zimmerman setting up the mics, and casually mentioning that he was glad Jason was up there with him, because Jason is a "superior masculine sample." Come to think of it, Jason is quite manly and charismatic. Also, he's done enormous good to our industry, has helped me personally, and is generally wonderful. You know what he's not wonderful at? Ranting. He is sad that we all got a pretty good game literacy education because we started with simple games like Pong and then games got more and more complicated as we got older, but now kids might start out with random iPad games that are kind of too hard for them. Why Jason thinks that some babies possibly playing overly complicated games is the most BURNING issue he can "rant" about is beyond me. After the session I asked him what's up with his "rant" being like a semi-boring thing he's barely sad about, but not really. He said it's "because he's Canadian."
Then there were some mysterious mystery ranters. I totally forget the first guy's name, so sorry. I've seen him speak before, too. Anyway, this guy fucking RANTED. I mean he cared and cursed and talked intensely. I was actually so excited about that, and thinking about how I'd write about him knowing what rant means, that my mind drifted and I forget what he talked about. I guess that didn't work out well. I think it was various useless bullshit distinctions that make him mad like "what is a game?" and "wait, this isn't a game," as if expanding the definition to include things outside of *your* games somehow cheapens your work. He says we decided on "video games" as an overarching term a long time ago, and get over it. He probably said a lot of other stuff like that.
Chris Hecker guest ranted too, at about 75% rant-intensity. Hecker talked about the three-way between the press, game developers, and game players. The core thing he's mad about is that we keep making the same games over and over again. I'd like to interject that even though we already had RTS games, I think it's fucking awesome that Blizzard made a polished one in the form of Starcraft 2. And in fighting games, there *still* are hardly any that are really good enough in my opinion, so it makes sense to keep going. Innovation in entirely new genres is great, but so is refining current ones. So this kind of rant, I just don't know about it. I mean I guess there's a problem, but it seems overstated.
Anyway, Hecker explains that all three parties cause this problem. He showed a review of some FPS with Black Ops in the title, and "the good" section was like super long. "The bad" section was "campaign too short." Hecker thought "the bad" should say "also you bought the same fucking game 6 months ago." He faults players too, for *wanting* the same game over and over again. And he faults developers the most, because while the other two parties count a lot, only developers can actually *make* other kinds of games. So it's kind of up to us to show how good other kinds of games are. And he says the problem here is that great developers throughout our industry's history have very often said "I make the kind of games I love to play!" Hecker responds "please want to play some other fucking types of games." I forget if "fucking" was in there, but it should have been.
Jade Raymond of Ubisoft I think might have gotten a misprinted invitation to speak at a "quiet and polite pontification session" rather than a rant session. She calmly mused about the world issues we faced in the last year, like some war somewhere, and the occupy movement, and internet censorship laws, and so on, and how we really should use games to speak about some of these issues, but mostly we don't. She mentioned several games that missed opportunities to join in this dialog, including one of Ubisoft's own games (I think Splinter Cell?) In that one, there is an interrogation mechanic in the game, and she explained how it could have been used a "frog in the boiling water" thing where in order to be successful at the interrogation, you have to do more and more terrible things each time, and eventually you do some monstrous thing that doesn't fully hit you until afterwards how terrible it is, because you've been conditioned to go along with it all. Actually that does sound interesting. I think the content of her presentation was fine, though I think I'm not doing it justice here.
Frank Lantz ranted at about 60% intensity, but his actual subject matter was ranty enough to get full credit. He faults game designers for not having enough ambition and for thinking way too small. He pointed to metacritic and some other chart site and said it's common practice to measure success that way. Like if you're top of those charts, that's success. But Frank thinks it's thinking way too small to even accept the model we have of churning out new games and sequels to those games and content packs to those, as if it's all this consumable, throw-away stuff.
How about a game that someone designs a city around? How about a game that you can see from space, because it's affected society that much? How about a game that lasts 500 or 1000 years. He says there ARE games that we build cities around (as he showed a picture of sports statdiums). There ARE games that are over 1000 years old that we still have (Chess, Backgammon, Mancala). There are games that so mainstream that they permeate culture and that even presidents play (many pictures of Poker). And we're content with top of metacritic? Really?
Frank showed a fictional man and woman from the future, in their spacecraft discussing games. The woman says how she thinks the games of the 21st century were really great. She thinks they showed all sorts of themes about what we liked, and they were fun, and there was all the running and jumping and shooting. They were just great. Then the camera pulls back, and she says "it's your move" as we see her move a piece on her Chess board. Frank says that it's entirely possible that "Chess will *lap* us."
Frank's son then "ranted" about how he doesn't like dialog trees in games, except actually he does. Frank's son is totally cool, but this is another example of how a "rant" session should include actual rants. Like, things people are actually really fucking passionate about.
I went to lunch with Frank and his son, and we bumped into a guy he knew from Riot at the mustard table. I hate mustard. Also I told this guy that his game tramples on the spirit of competition, and that it's shameful that it takes dozens of hours of forced grinding to have "real" character instead of the gimped ones you get when you pay. How about paying gives you a real character with no grinding?
He seemed confused. Then he said they have to do it that way because they need to have a matchmaking system where people of the same skills are matched, and it wouldn't be good if bad players were matched against the best players. I said that's an invalid argument, and that skill-based matchmaking systems are entirely figured out. They exist in chess, Starcraft 2, and a hundred other games. There is no problem in matching players of the same skills, and if there was, preventing people from buying a full character wouldn't be any kind of answer to that. "Next," I said.
He said ok but LoL is a team game and so it needs different matchmaking. I said that's also no argument because that kind of matchmaking is also solved. Microsoft's TrueSkill system expands Elo to include team-based rankings, and it's already in use in Halo. This is a solved problem and again not a reason to prevent people from buying characters. "Next," I said.
He said he'd rather not debate it, so I said ok and we went our ways. Frank later said that seemed mean, I mean that was just some guy, why say that stuff to him. Frank is ok with a lot of things though, too many in my opinion. Frank works for Zynga now, by the way, which kind of says something here. I said I thought it would really improve the entire world of competitive games if Riot employees were regularly aggressively argued with at every conference, and that if players did it too at every turn. Instead of popularizing the idea that a forced grind is ok in a competitive game (which is against the spirit of competition), maybe they could be told that's fucking bullshit. In fact, if the tides were so strong that lots of people told them that, we would have improved the entire state of competitive gaming. So how about that? You know what that is? It's a fucking rant. Why is my lunch conversation at the mustard table 10x the intensity of half the "rants" in the rant session that promised to burn down the house? I mean seriously. And while we're on that subject, why is there an entire table full of just mustard?
Experimental Gameplay Workshop
After a week of Facebook, iOS, social, monetization, clones, and "I almost died making my game," this is what reminds everyone that our industry is still cool. A bunch of crazy people showed their crazy stuff.
Steve Swink fucking blew me away. Note that Steve blew me away two years ago too, with the beginnings of his game Shadow Physics. I was sure that "Portal, Braid, Shadow Physics" would be a common phrase. For reasons possibly unrelated to game design, Shadow Physics didn't work out though. He cancelled it and started on something else called Scale. So yeah. "Portal, Braid, Scale." I asked if I can be his co-designer or something. I have always wanted to make Gödel, Escher, Bach: the game, and maybe he's doing that, sort of.
Scale is a first person game (oh hmm...maybe should be 3rd person) where you have a gun thing that can shrink or grow anything in the world. Yeah, like anything. So you can get on a rock then grow it to make it big enough for you to reach the top of a house, or whatever. You can shrink a entire landscape with a huge river, so that it's so small you can jump over the river. If you encounter two pressure plates on the ground that must be held down, but only have one rock, you can make it a huge rock that pushes them both down. You can shrink a house down so small that you can carry it around, bring it somewhere, then go inside it...and find a crack that if you shrink yourself down small enough you can go explore. You can grow a small butterfly so large that it becomes a flying platform for you. You can do a ton of other things too.
There is a conservation of mass in the world, so in order to grow something, you actually have to shrink something else. This constraint makes it actually work, as there are limits to how big you can make things. Sometimes you have to think a bit how to even find enough stuff to shrink in order to grow some other thing enough.
There's just so much potential in Scale. Aftewards Steve asked if I knew Maquette, I said I didn't, and he was surprised. He called over Hanford Lemoore, creator of Maquette, and he gave me a demo of it. It's weirdly similar, and yet not. Inside the level, there is a small model of the level. When you change things in the model, it changes the bigger level and vice versa. In the main level, there is an organge block that's big and blocking your path. You can go to the scale model and pick up the orange block there (it's small in the small model) and move it so it doesn't block your path. You can even...carry it from the model into the regular sized world.
That starts to get mind-melting when you realize that after you do this, if you return to the model (after setting down the orange block) that there is still an orange block in the model. Why? You picked it up and put it somewhere else, why should there be an orange block in the model? Because inside the model, there is another smaller model of the level. And when you left the first model with the orange block, another smaller version of you left the model's model, carrying the orange block to the, uh, model. So basically both these guys are making Gödel, Escher, Bach: the game. I think I will try to help them if they let me.
Yeah so back to the experimental gameplay workshop itself, rather than private demos afterwards of last year's entries, lol.
The game Storyteller is a very unusual and interesting puzzle game. It shows a few panels, like from a comic book, but they're empty. You have characters and objects you can drag into the frames. You are supposed to tell the correct story. The first example is like "Jim died" or something like that. You have Jim and a tombstone. If you drag a copy of Jim to panel 1 and panel 2, that's not right. If you drag Jim to panel 1 and the tomstone (but not Jim) to panel 2, then the game knows that Jim died because he appeared in a previous frame but now there's a tombstone.
The examples got more and more complex. You can drag boys and girls with hearts over their heads and they are in love. If one is not in a later frame, they are sad. There is like regular Jim and a different "object" you have that's "Hero Jim." There's some other character, regular and villain. Heroes and villains don't fall in love, so they have different logic from the regular versions. There's an old woman character who can turn a regular character into a hero. If you start with regular Jim, add a girl so they are in love, next frame have the old woman turn Jim into Hero Jim, then next frame put Hero Jim and the girl, they are both still in love. Why? Heroes can't fall in love I thought? Actually normally they can't, but each frame inherents states from the previous ones so because regular Jim loved her, Hero Jim still does.
There's a pretty interesting set of mechanics to make puzzles, so I think this game has a lot of potential. On a purely logical level, it could offer fun challenges. On a fun level, it's often pretty funny how the story you're supposed to tell unfolds, and the game sometimes puts speech bubbles and narration on the comic frames that reflect what's going on as you solve it, and that's even more funny. It's fun to watch someone else play.
Glitchhiker is a game that was created at a game jam by 6 people in 48 hours. The game is now dead, permanently, and this session was a remembrance of it. It was an arcadey game that intentionally looked slightly glitchy, and your goal was to score at least 100 points. This is kind of a masterpiece in a way, because it shows how you can take a mundane thing (some generic running / shooting thing to score points) and make it a very meaningful, memorable, impactful thing.
Why is it any of that, you ask? Because the big idea here is that the game itself had lives, not the players. The game had a pool of 100 lives, and anyone who played it was using up one of those lives by playing. By scoring 100 points, it gave a life back to the system. By scoring 200 points, it gave another life back. This is "massively multiplayer" in the sense that even though multiple people can be playing their own copy of the game, it's still pulling (and refreshing) that same pool of 100 lives. As the game got lower and lower on lives, it got more and more (intentionally) glitchy, including graphics, sound, and even gameplay bugs becoming more frequent.
Players were told that when it reached 0 lives, the game would be dead and would never be playable again.
The "never playable again" part is actually for real. I missed the actual tech thing behind this, but it was something about a randomly generated password encryption thing that prevented the developers or anyone else from ever accessing the game again after it died. They thought if anyone were ever able to play it again, it would not have the same meaning, as it would just be resurrected over and over. Dead means really dead to them.
The entire point of this, I think, is to look at the way players experienced this game. To play it at all is a RESPONSIBILITY. Some people said they didn't want to, because if they failed, they would have taken away some other person's chance to play it who might have enjoyed it more. People who did play it and failed to score 100 points felt guilty. One excellent player was able to get 200 points multiple times, and played it continuously for 6 hours in an effort to help the game survive. By the way, the game lasted 6 hours, and it was killed by a drunk Canadian.
Keep Me Occupied. This is an arcade game physically built, designed, and programmed by some passionate people who supported the Occupy Oakland movement. One joked that it would be impossible to explain such a big social movement to the audience in only a few seconds, but he did a pretty good job of it. He said imagine the people who make the rules for a system are also the ones who the rules benefit. He showed a picture of Notch making Minecraft and another Photoshopped picture of Notch excitedly holding up an enormous check for tons of money from the mega Minecraft tournament victory.
Anyway, they are passionate about the social change behind this movement, and told us briefly about other social movements in our history. They have almost always involved large groups of people physically coming together into the same space, as that kind of face to face communication and camaraderie is extremely powerful. The pictures of those gatherings look a lot like the pictures of Occupy Oakland.
One particular movement from history was notable to them, the labor movement. They showed a picture of "our grandfathers and grandmothers" holding up a sign that says "8 hours of work, 8 hours of leisure, 8 hours of rest." They pointed out that the 8 hours of leisure they were fighting for...that's us. That's what we do, is provide that leisure. Our industry is only even possible due to the social change those before us fought for. So they said they felt they owed it to those courageous people to do something about our current social problems. They would make an arcade game and put it in the common building the Occupy Oakland was using. This building provided free food, medical attention, library services, and leisure as a break from the intense goings on. They thought that if they could do their part to get more people in there, it could result in those people realizing there are real and valuable services offered in that building too. Though police apparently raided or destroyed the building or something, the physical game they made survived and was there on stage.
The designer said for the game design, he was inspired by the board game Risk Legacy. In that game, you apply stickers and other changes to the board as you play, so that future plays have different starting conditions based on your past plays. The "Keep Me Occupied" game has a similar idea. It's (on purpose) incredibly simple so that even non-gamers can play it. It's just a map of doors that each need switches held down to open. So you and another player work together (in the spirit of Occupy Oakland) to get farther and farther. I think there's a time limit, too.
The catch here is that when you die or lose or time runs out or something, you STAY there forever, even for all future players. So even if you completely suck, you can end up holding one of the switches for the first doors, forever, helping all future players get farther. It's really a poetic take on what cooperation means.
It was also interesting how they physically got the arcade machine where it needed to go. There was a march to bring furniture to the common building, as a kind of statement, and they were part of it with their arcade machine. This was actually really difficult because it's heavy and unweildy. It got lots of attention, and people played it as it went on the march. (There was some battery or something that powered it? Not sure.) Strangers offered to help them push it, seeing that it was heavy. At one point, the layout of the path meant they couldn't go that way with the arcade machine, and had to take an alternate route. Strangers helped them navigate that route, helped them rejoin it, and the whole experience of this was meaningful to them. There's something striking about them experiencing the spirit of cooperation as they transported an arcade game that itself is about cooperation to a social-political gathering that is also about cooperation.
Then they played the game on stage, but due to technical difficulties, they couldn't have the video on the big projector screen and on the screen the players could actually see. So they had the audience split in half (half for player 1, half for player 2) and provide directions to them with pointing our hands. This was yet another example of cooperation. Frank Lantz whispered to me that it's ironic that it would be far more efficient if just one person directed them, instead of this giant mess of people that is obviously doing a bad job and confusing them.
Twirdie. This is a golf game. It is the most interesting golf game ever. You type in a word, and then the distance your ball goes corresponds to the number of times that word was said on twitter in the last minute. I'll let the awesomeness of that just sink in for a moment.
So you type like "Obama" and it goes really far. They typed in "GDC" and it went only 9 meters, lol. (So 9 tweets.) That's actually the point, right there. They said that at first people use words that are from their own little worlds, but they very quickly start thinking globally, they start thinking what matters on a big scale, what might be going on in the world right now.
The two developers played an actual game against each other on stage. One opened with "Beiber" and said "you guys may laugh, but this guy has been a powerhouse all throughout development." He said it's interesting that you want to open with your strong words--your Beiber's, your Kardashians and so on--but then you want less popular words so you don't overshoot the hole. One of them used "ninjas" when he was very close to the hole, as he said it's a real solid word that always has low, but not too low number.
They said it's interesting that there really is such a thing as mastery in this game. You really can get better and better at it. And by getting better, you are understanding this data set (the set of all tweets) that is so large that you wouldn't think you could understand it at all. So this illustrates how a game can teach you something on a very deep level because it's all about you interacting with the system over and over and learning from each of those interactions.
Another hilarious note is that in playtesting, someone brought their 6 year old kid, who played the game 2 hours straight and beat the parents by the end of it. He was a super star tester and was one of the best players of the game. His dad used "driving" and the kid said "people don't tweet about driving, duh." Ironically, later in the presentation when the developers played each other, one chose "traffic" and it didn't move the ball nearly enough. No one tweets about driving, come on.
The audience was completely engaged in this game. *I* was completely engaged in this game. It's really fun and interesting. They said often in playtests, groups gather together to help each other play, like "hey SXSW film festival is huge now." Or "it's tax season now" or "MLB will be big right now for sure" or "Try BP, or oil or oil spill or ducks" after the oil spill disaster. Somehow sharing knowledge about the world was fun. And even a 6 year old kid who didn't know about any of these world events and who doesn't use or read twitter found it fun to master the system anyway. I want to play this game.
And that is all for GDC 2012. I'm so exhausted from this entire week.