GDC 2010, Day 2

Come with me to day 2 of the Game Developer's Conference. But first, here's some things I left out from yesterday.

Jaime Griesemer Again (Bungie)

Jaime Griesemer made a point yesterday that when gets feedback, he doesn't like hearing solutions, just problems. He's ok with "I don't like this" and he's even better with "I don't like this because [of X]" but he's not hot on "this should be changed to that." Often these solutions are not feasible. Sometimes they have technical problems, sometimes they cause other even worse problems in some other area of the game, or whatever. He says don't discuss solutions with playtesters, do that with other designers.

I've found this to be good advice from both sides of the coin. I've also heard lots of "change this to that" pieces of advice that can't work, but the real message from the player is that SOMETHING is wrong, so figure out a better solution. On the other hand, the playtesters I work with these days have a close relationship with me. They have learned a lot about my ideas and methods and are often able to provide good solutions. Even with that, there have been many times when there most valuable feedback was identifying a problem that I then puzzled over to find a solution. I mainly bring this up because Sid Meier said exactly the same thing today...but one thing at a time.

Another point he made yesterday was about ignoring balance feedback in some situations. He was saying that if the people giving the balance feedback (aka, the people complaining) realize that you can easily change a number somewhere to change the game, then they will complain over all sorts of things. Maybe a strategy or weapon or move or something is pretty good, so they complain rather than explore the game more and find counters. And yet many of these complains go away the moment the game is in a more fixed form, like when it's actually burned onto a disc and changes would be hard. At that point, many of the previous "complaints" go away and those players learn to overcome whatever challenge by actually getting better. Obviously you have to be careful about when to ignore or not ignore this kind of feedback, but I've noticed this same phenomenon.

Rob Pardo Again (Blizzard)

I remember a few more things Pardo said yesterday. "Don't make players read a story." He limited quest text to 512 characters on purpose in World of Warcraft, not for a technical reason, but to make quest designers keep it to the point. He said that players should be able to get the gist of the story by only reading the objecting and actually doing the quest. The quest text can then enhance, deepen, or further explain things, but it shouldn't be necessary for understanding the basic story.

He said one place they really failed at that was Diablo 2 quests. In that game, you talk to the quest giver and they launch into 2 to 3 minutes of monologue about all this story stuff. You sit there with no form of interaction. The quest is really just "Kill Andariel" or whatever though. He said that's a fail. But he said World of Warcraft: Wrath of the Lich King succeeded here, specifically the quest chain for become a Death Knight. (I agree, and so does basically everyone else.) In this quest chain you steal a horse that you then turn into your Dreadsteed. You get quests from Arthas himself at first, but we see how your alliances change. Just playing through it all and reading almost nothing gives you a great sense of what Death Knights are all about.

Another point was that players will choose the shortest path, so make sure the shortest path is also the most fun. In EverQuest, the fastest way to level up is to fight the same monster over and over a million times, standing in like the same spot. In World of Warcraft, the idea was that quests were the fastest way, not standing in one place killing the same monster. He pointed out that even the most boring of all quests "Go kill X bears" or something is a big improvement because at least you finish that, turn it in, then get another quest to go kill some other monster, maybe with some other abilities, and that is located somewhere else in the zone. It gets you to move around at the least. Better still are quests that are more creative and fun, and that give enough XP to be worth doing even for the optimizer players.

Pardo also said that Blizzard is known for polish, but that polish isn't something you do just at the end; you have to do it all the way through. I think that's another way of saying you need to iterate and iterate and iterate. He showed a top down map of Arathi Basin (the battle ground in World of Warcraft) that was really low res and pixelly, like something you'd see on an Atari 2600. He said that was the design document for the battleground, ha. (It was remarkably accurate!) Then he showed a screenshot of the earliest playable version of it. It looked ugly of course, but it was playable. They could move the bridge or the flags or whatever and try different things. They did this through its development and it turned out to be one of the best battlegrounds with the LEAST overall development time.

The contrast was Silvermoon City. Silvermoon is a HUGE city, bigger than they had done before. It was so big that they had to break it up into several sections, each built in isolation. It was so unwieldy to connect up all these sections and actually play it as a whole that they very rarely did it. I think he said they did that only about 2 times in a YEAR (oh my). As is no surprise, Silvermoon turned out to be unwieldy to actually navigate as a player, too. It just didn't have the continuous iteration and polish (as a complete, continuous city) that Blizzard usually does. Pardo said that they now call this "Silvermooning" and are very careful to avoid any situation that prevents them from doing many, many iterations on something.

Another point he made (he sure made a lot of points, btw) is that he has to create a culture where his employees "show their work early." He says if you work on something (maybe a map or a character or programming a feature, whatever) you don't really want to show people when it still has obvious problems. It will make you look stupid. But the alternative is to keep working and working in secret, building up to some kind of "big reveal." At this point, you're too invested. If you've been working on something for three months, he said, and then finally show it, you aren't looking for feedback. You are looking for a pat on the back. But the only way to make things good is many, many iterations. He encourages his employees to show each other even their very early work and to give each other suggestions on whether that work is going in a good direction or not, or how it could be improved.

Ok, now let's start GDC Day 2 for real.

What You Need to Know About Casual Games 2010

This was the worst named session in all of GDC. Or...was it the best named session? Maybe the worst because it showed almost no casual games and almost nothing from 2010 (the games are mostly from 2009 I think). Or maybe it's the best named session because it seeks to redefine what a "casual game" even means. I thought it meant lame match 3 stuff, peggle, and facebook non-games. Apparently to Juan Gril and Nick Fortugno, it means a bunch of awesome experimental web games. Sweet.

Juan and Nick took their first minute or two talking slowly and not saying much, tricking me into a state of relaxation. Then they proceded to launch a 50 minute thrill ride of games and game-explanations. Slides and movies of games rapid fire with their commentary along the way. These bastards spewed so much information out of their firehoses that I had to actually take notes on something for the first time in like 10 years. Now I'll rapid fire it at you (no links to the actual games though, sorry no time.)

Closure. The only nominee to this years independent game "nuovo" award that was also nominated in another category (music I think, and I think it won that). A black and white game that takes place mostly in darkness. If a platform is illuminated, it's "really there." If it's darkness, you fall through it and die. So it's about making sure the light is in the right place at the right time. This is a good blending of theme (light vs dark) and mechanic (light vs dark).

Time Fcuk. A simple looking platform game, but the trick is there are "layers" (like in photoshop). You have a button to switch which layer is front (real) or back (not real, but shown in gray). I thought it was especially clever to put a different "gravity item" on each of the two layers, so that switching layers actually switches which direction gravity goes. This game was inspired by the black and white game Shift, where you can press a button to sink through the floor to the other side, basically choosing whether to move in the "positive space" or the "negative space" of the map.

A Mazing Monk.

You, Me, And Cubes.

My Kingdom of the... (um, this is crazy fast, I don't even know what they said about these or what the rest of the title of that one is)

The next few games they recognized for the visuals.

Machinarium. Single click interface, you start as a robot head then must build yourself before you move on to other areas. They mentioned it took 3 years to develop.

Blackwell Convergence. A series of LucasArts-style point-and-click adventure games done by a single developer. They said the story-games are pretty well written, and that there are new and interesting mechanics in addition to the classic stuff. You are a detective, and your partner is a ghost. The ghost can move independently, so it's like you have two characters. The ghost cannot interact with the environment, but he can pass through barriers to get to places your main character can't get yet.

Enlightenons. A hidden object game that looks surreal. (?)

Alabaster. A text adventure game(!) that is somehow good or something. There was some explanation about why it's ok that it's a text adventure game in this day and age.

Aubergine Sky.

Beggar. Explores that life of beggars, making a political statement with gameplay.

Flower. Beautiful of course. Juan cautioned though that if you cry because it's that beautiful that you should "get over it."

The next set of games are multiplayer.

Restaurant City. It's a slow sim by Playfish (on Facebook). The clever thing is that you make your real friends into employees. Like make your boss the janitor or whatever. They said people forget about Restaurant City because Cafe World came later with pretty much the same features and so on, but is way more successful. Although the presenters really liked the multiplayer emphasis of Restaurant City, they speculate that Cafe World is more successful possibly because it chose to focus more on cooking than on the multiplayer aspect. (Ironic for a "social game." Newsflash: social games are mostly single player games where you pass some tokens around. That's what Daniel James said two days ago, at least.)

Castle Age. It has a "call to arms" button that tells your friends to fight the monster you're fighting. This means they can fight it on their own time and help you out. It's basically an asynchronous "raid." (Seems like I should actually investigate that when I have more than 0 time.)

Monaco. I don't even know what this is. A top down, cooperative stealth action game (!??) that won the independent game award yesterday. Everyone who mentions it says how amazing it is, and they don't give any more details or even screenshots that explain anything.

The next games are arcadey.

Bit.Trip Beat. Very interesting idea. You have a pong-like paddle along the left wall and balls fly at it. The catch is that there's music...and the balls are coming in beat to the music. It looks like you're actually travelling to the right, and all this junk is coming at you that you must deflect. As soon as I saw it, I thought "this is like Rez in 2D, plus Pong, plus really hard difficulty." Then the presenters said that it reminds them of Rez (see, I told you) and that it's really hard. They thought that it captured the concept of Rez like almost nothing has since. They also said it was one of the most successful games they know at having them enter a "flow state" while playing.

They stopped to make the point that "casual" games can be hard, they just can't be complicated. Guitar Hero is hard, but it's not hard to figure out what you're supposed to do. You can go from easy songs to hard ones but it's still a "casual game."

Learn to Fly.

Fig. 8

Cyclomaniacs. (no idea what they said about these.)

The next games are about management of resources/time.

Swords & Soldiers. A 2D RTS on the Wii. It's side-view, which is an odd choice for an RTS but they said it really works well and makes for a great casual game. Also, the UI is in a bar across the top of the screen. They said that the Wii has technical difficulty with handling pointing at UI elements all over the screen, so it was a very good decision to put all the stuff you point at in a strip at the top.

Plants and Zombies. A tower defense game that simplifies things even more than regular tower defense. The enemies don't walk in mazes, they walk in straight lines. Everyone loves this game apparently.

Cocoanut Queen.

Drop 7.

Spider is a great touch screen game on iPhone. Great that you swipe and the spider jumps and leaves a trail of web. Feels great, looks great, and you don't have to cover up the action with your finger to make it work. You can swipe anywhere, not just on the spider.

DrawRace. You draw the path your car will go (top down view) with your finger. Then the car actually races around the track against other cars. Careful, because the speed you draw determines the speed your car will go. If you have it go too fast, it can't make your turns though. They say it's awesome and fun and is a great way to do a game that works against other players even with ridiculously bad network connectons.

SmokeScreen. An ARG that simulates social networking with a fake facebook-like site, fake IMs and so forth. Teaches the dangers of stalkers and giving identity to strangers and so on. Educational and the presenters said pretty useful because it's so literal. The privacy settings page in the game looks exactly like the real one on facebook.

Sid Meier

Sid says he is a logical guy trained in math and computer science and it took him a long time to learn that he has to think in terms of psychology a lot in order to make games.

He apologized for many of his sins. He said Civilization started as a real-time game, but that was a mistake. The simulation marched on with or without you, and you felt like a viewer. When he changed it to turn-based, it felt like YOU were causing all these things to happen, and shaping the history of the world. Good thing he figure that out, he says.

He thought the "Rise and Fall" idea could be cool. Your civilization rises, then falls to the brink of disaster, then rises up again even higher than before! Exciting! No it isn't, he said. Everyone hates it and loads a save game when the fall starts to happen. As a result, he now knows to make it "rise then rise, then rise some more."

He thought it would be cool to have random bad things happen, like they did in real history. Plagues and earthquakes and floods and so on. No it isn't, he said. He said you have to be really careful with anything related to randomness, and if it's an event that hurts the player, they have to know why it happened and how they can avoid it in the future. He actually talked quite a bit about the idea of allowing the player to plan how to do things better next time because you want there to *be* a next time, and getting players already thinking about playing against is great. Replayability is very important.

Remember how Pardo said that players will take the easiest route, so make that a fun route? Sid basically said the same thing when it came to saved games. He said many players would save before every single battle and load if they lost, so that it ensures they will win 100% of the time. Though this may be optimum play, it's also totally boring and pointless. He said that you have be very careful not to give the player tools to make the game totally boring and pointless, and that you have to "save the players from themselves" often. He then changed it so your save included the random number seed so that you'd just lose again if you loaded. Apparently people found some way around that, but at least he tried.

Also, he wanted to limit when you could save to help address the problem more. So he made it so you can only save in ports. He said this restriction seems to make sense in the fiction of the world, and it masks things a bit that the real reason for the save-in-ports rule is to change the player's save-load behavior. I actually disagree with this and I wrote an article on save game systems for Game Developer Magazine about that (also on this site.) I think the game designer does not have the right to tell a player when he can save. He should be able to save ANYTIME. The trick is that you don't have to save the exact gamestate, save whatever stuff makes sense from a design-standpoint. If my friends are knocking at the door, I have to stop right away, not when I get to a port or something. So let me save, and put whatever you deem necessary in that save file, but at least let me turn off the game immediately.

Anyway, this goes to Sid's theory of "The Unholy Alliance." He said he should trademark that phrase, and regrets not owning the rights to his famous "games are a series of interesting decisions" line. He's joking. Anyway, the unholy alliance he means is between the players and the game designer. Each one can ruin the whole thing at any moment. Players can turn the game off or hack it or whatever and designers can make levels impossible or whatever off crazy thing to ruin the player's fun. Each CAN ruin things, like the "mutually assured destruction" by nuclear attack between the US and the Russians in the cold war. But each side decides not to ruin it, and instead we work together to have a fun experience.

The most amusing part of Sid's lecture was about the role of randomness in his games, and watching him react to various illogical feedback he gets. He's learned to just go with it and accept that he should match player psychology rather than mathematical correctness. He started by showing two sides of a battle in Civ Revolutions. One side has a rating of 1.5 and the other has a rating of 0.5. That means that one side is 3 times more likely to win than the other. In four battles, we would expect, on average, for one side to win 3 of those and the other side to win 1.

So he said people would complain, "I lost but I had the 1.5." Sid explains "yeah but it's a 3:1 thing. so sometimes you will lose." The complainer is still grumbly. He ges more grumbly when it's 4:1 and he still loses. The player explains that at some point, his win should be guaranteed because it's just so big of an advantage. Sid says ok what about if 4:1 mean you just always win, but we keep the 3:1 as it is. The (rhetorical) player says ok.

Then someone complained about a different battle, this time it was a 3 vs a 1, rather than a 1.5 vs a 0.5. The player said "look, I have this big 3 and the other side has this small 1, but I lost." Sid scratches his head saying, "but we just covered this. It's 3:1 ratio." The player says "yeah but look, it's a large number 3 written there!" Sid also observed a case where the player had the 1 and the enemy had the 3, but the player won. He asked "does this bother you?" The player said "no, it seems fine. Even with a smaller number I still win sometimes." Sid wants to cry I think.

He somehow gets people to be ok with this whole 3:1 thing. Then someone complains, "hey I lost a 3:1 battle." Then Sid says "We JUST talked about this like five minutes ago, and it was ok." The player says, "Yeah but this is different. You see, I lost a 3:1 battle, then right after that, I lost *another* 3:1 battle." Sid says right, it's because it's random. Streaks happen. Each event is independent of other events." The player says "right, but you see I lost two in a row. That doesn't seem fair." Sid says he changed "randomness" to work how players expect. Now each battle DOES take previous battles into account, and if you failed before, you're less likely to fail here. Notice that Sid Meier and Rob Pardo both ran into this same problem with randomness and both solved it the same way. (Dear Kongregate.com: maybe Kongai should use that same method. It's hard to ignore that two gaming legends have done this while we still have the same complaints about bad streaks of randomness in Kongai, shrug.)

EDIT: Oh, and one more thing. Sid gave another example where a player complained that he lost in a 20:10 battle. Sid said "Ok, but that's just 2:1. And in a 2:1 battle, you'll usually win, but sometimes you'll lose." The player said "but it's *20* to 10. I mean TWENTY. That's way more than 10. It's 10 more than 10!" Sid said "right...10 more..." I don't remember if changed 20:10 to just make the player win, but he probably did as he cried himself to sleep.

The Connected Future of Games

This was a panel discussion with Ray Muzyka (BioWare), Brian Reynolds (Zynga), Jason Holtman (Valve), Min Kim (Nexon America), and Rob Pardo (Blizzard). N'Gai Croal moderated.

The Valve guy mentioned that it's very important to allow people to be a fan of THE GAME as opposed to a thinking of them in really isolated terms like "they own the PC copy of this game." On the one hand, a fan of the game can do a lot more than play the game, like watch movies online about it, or engage on the forums, or buy related merchandise, or whatever. Another angle is that they are now launching Steam on the Mac along with Mac versons of all their games, so you can "a TF2 fan" or whatever, instead of "owner of TF2 on PC." Play it one PC computer, then go home and play it on your Mac, or whatever.

Incidentally, a couple more Mac things I noticed. There were a lot of laptops around me during Sakamoto's keynote yesterday. I counted them and found 7 Macs and 2 Windows machines. Also, all the presentations at the boring V-con thing were done with Windows machines. All the presentations as the indie game summit were done on Macs. People who read my website seem to think Macs are some rare thing, yet they are dominant and pervasive amongst the professionals around me who are doing interesting things.

Anyway, back to the panel. The Valvue guy also talked about the Portal 2 promotional stuff, the ARG or whatever you would call it. He said many people asked him which marketing firm they hired for that, but he said they did it in house...with their game designers. It's a game and they have game designers. He said it would probably suck if it came from marketing. He stressed that game designers know their audience well and that they know games, so it makes sense for them to design something like this. He said it's important to let game designers think beyond the game itself, and into other areas like this.

The Zynga guy completely agreed with this, and went even further. He said the design of all the social mechanics that surround a game are more important than the actual mechanics inside the game. This is certainly true for Zynga games.

The BioWare guy talked about which content to make free and which to charge for. He said the right way to think about it is WHEN you want your revenue. If you charge for more content, you could get more revenue now. But if you give out free content, you get more players to play it, you keep them happier longer, and you build your brand. He said this way, when you do actually sell something later, such as DLC or even a completely new game, people will buy it because they trust you. He said some people will buy BioWare games sight unseen now because BioWare has built up that much trust. So it's possible that giving away some content for free is actually giving you lots of revenue 2 years from now, or 5 years, or 10.

The Valve guy added to this, saying that giving away free content in updates can sometimes even make you more money NOW, too. Whenever Valvue does an update to TF2, like another map or changes to a class or something (all free), they see a big jump in number of players. They get lots of players who haven't played in 6 months or something, all flood into the game after an update. The effect of this is that all the friends of those people hear that everyone's playing TF2 now (via steam or facebook or a dozen other possible ways of knowing what your friends are doing) and this means a lot more people buy TF2.

Pardo talked a bit about Battle.net, and how doing their own service is a ton of work, but allows battle.net to be deeply customized for their games. Someone asked if they would offer battle.net as a 3rd party service that non-Blizzard games could use. Pardo said they've thought about it, but they just don't know if they could do a good job of it. He said it's hard enough to get things working right with their own games and supporting someone else's sounds really hard.

Pardo also mentioned off-handedly that Starcraft 2 will have a map store similar to Apple's App Store. I didn't even realize this, but apparently it was announced quite a while ago. Cool. (This was maybe the 5th time at GDC that someone's business plan is "just copy Apple's App Store.")

I thought Pardo's most interesting and cutting comment was in response to the question "What would you like to see out of the next generation of game consoles that would entice you to release games on a console?" He said "a better certification process." Amen to that. Non-game developers don't even know the half of it, but there a lot of restrictions. Pardo rattled off a half-dozen things off the top of his head that he can't do on a console now. What if they want to update a console version of World of Warcraft? And...a lot of times? Each one goes through some long cert process that makes it infeasible? What if an update causes an unexpected major problem that they need to hotfix? This seems not even possible with current console cert requirements.

Pardo said when Blizzard releases a new raid boss or something, his designers all sit in a room together watching the first players fight that boss LIVE. They watch closely and sometimes change the numbers during the fight! (Note: the changes don't affect the current instance the players are in, only instances created after that point.) Anyway, there's all sorts of things he can't do on consoles. Honestly, it looks pretty sad compared to all that battle.net can do. It seems like console makers should bend over backwards to let Blizzard do whatever they want, honestly. And that raises the question of why everyone else can't do what they want, too. I have thought about developing a fighting game, but it seems infeasible to do well on a console because such a game REQUIRES many balance patches to be good, yet consoles currently make you pay a ton of money for each patch. Meanwhile on Mac and PC, you can patch whenever you want, 100 times per day if you want.

Prototyping Gears of War

This amazing session showed many actual prototypes from Gears of War, showing us which things were scrapped, and which went on to be in other more sophisticated prototypes, and how features evolved into the final game. Too bad it was full when I arrived and I was turned away. Damn. I went to this instead:

Game Design Challenge 2010: Real-World Permadeath

Eric Zimmerman (GameLab) hosts the game design challenge every year. He gives some screwball topic to three contestants ahead of time, and they have to present a game design that satisfies the challenge. Then the audience votes (by applause) for the winner. If you win, you get to come back next year. If you win a second time, your prize is that you don't have to come back a third time.

This year, the challenge was to design a game involving actual death. That is, the actual death in the real world of a real person. Zimmerman explained that games have always trivialized death. Pac-man had lives, and ever since we've been dying and dying and thinking nothing of it. But death is a powerful force. Can we design a game that centers around the real thing, and not a fiction about death?

The contestants were Jenova Chen (That Game Company), Kim Swift (lead designer of Portal), and a two person tema of Heather Kelley and Erin Robinson. That means 3 of the 4 are women. At least there's some progress on the gender issue here.

The team of Heather and Erin presented first. Their game was about the process of making a will when you die. This is usually depressing and horrible, so they wanted to make it fun and uplifting instead. I forget who was speaking here, but she said in her family, they passed on many objects from generation to generation. They're labelled on the bottom, saying who originally bought/created them, when and where the object is from, and a note on why it's significant. Anyway, the game they propose involves creating a catalogue of information about your own objects, as a first step. This helps to preserve your family history, and also the game will use this information. They proposed as scanning interface to get info into the game, and mocked it as an iPhone app. I thought about how you'd do this years in advance of your actual death, and how iPhones might not make any sense in the future (replaced by other technology), but whatever.

Anyway, after you die, your family gets to take a fun quiz about you and all these items! In many cases, grandma didn't really know exactly who to give a certain thing to, and the game decides who it goes to by figuring out which player knows the most about item and why its significant. For big ticket items, this could reduce fighting or resentment, because it will hopefully go to who deserved it most, they said.

At this point, I thought about Heather's voice is unusually low for a female, but not in a bad way. It has an interesting quality to it, and I found myself listing to how she pronounced her words, and the tone of her speaking. After I while, I realized they were still going on about this multiple choice game you play after your grandma dies.

Then Jenova Chen presented. He said he thought first about which platform his game should be on, and because it's about something real rather than a fiction, he though it really belonged on Facebook. (Are there any sessions were Facebook is not mentioned? Btw, Sid Meier talked about Civilization of Facebook, and I didn't even go into that.) Anyway, Chen said once he decided on the platform, he just "went with the trends" for the name, and called it DeathVille. He even had a black logo. He said that an earlier talk taught him that if you make things pink on Facebook, that more people will click on them, so he showed us a new logo he made that day or something. It was pink and said HeavenVille.

Anyway, his idea is that on Facebook, you can get a feed for all sorts of stuff, why not a feed of all the people in the world who are dead? He showed a map fo the world with big red spots showing where the most people died that day. We could see how many people died and were born in any given day, which he finds very interesting information to gain perspective. He says that in his lifetime, he has talked to less than 6 people who later died, so he feels he lacks perspective on death and something like this would help give a sense of scale.

Then he showed going to the facebook page of a dead person, and how you could "collect" dead people like friends sort of. Then he thought about the value of those dead people. He was quick to point out that when someone dies, we don't really care how much stuff they owned. We care what they gave back or what they contributed. It's really hard to rate that. He said with more time, he might use a sort of "Hot or Not" voting thing to determine this, but given his very limited time to prepare, he used google search results instead. He showed how many results each of these people had: George Washington, Albert Einstein, and Michael Jackson. Of course MJ had the most, but this seemed like it wasn't fairly representing the rankings of value. Then Chen thought "if a guy has been dead 1000 years but we still talk about him, that should count for a lot." So he decided it was better to multiply the search ranking by the number of years that person has been dead. This put George Washington miles above MJ, so that seemed right.

Then Chen wondered how high he would rank on this, so he googled his own name (with quotes around it) and was disappointed how low it is. Then he did Shigeru Miyamoto and Sid Meier, also higher than him of course. Then he did various girls he worked with, who really blew him away. Then Jade Raymond of Ubisoft, also way higher than him. He then noted that for none of the males he googled, but all of the females, the results page shows pictures at the top, meaning that most the hits were from image searches. He said something like "Hmm, I wonder what that means..." then moved on.

Anyway, the idea for his game is that you can buy and sell these people as a stock market of the dead. George Washington is going to be a pretty stable stock. MJ is going to be a lot more volatile. He might go up or down a lot in a year depending on if he still gets lots of coverage, or if it drops off sharply soon. Chen said you really want to find rising stars and buy their stock before they die. (Later, a question from the audience asked if he thought about possible exploits of insider trading amongst assassins. He said he hadn't had time to work out those details.)

Chen then proposed a new idea here, by using a mockup of moderator Eric Zimmerman's DeathVille page to illustrate. Chen imagined he bought Eric because he thinks Eric will be valuable when he's dead. He'd like that stock to go up, so to help it happen, he can add more information to Eric's page about why he's culturally relevant. In the example, he put links to the four books Eric has written, and links to the games he's worked on, and so forth. This earned Chen lots of DeathVille coins, and it also will hopefully inflate Eric's search rating. Really, it's a kind of wikipedia that bribes people to fill in information with external rewards. (That's my wording, but I couldn't resist, given Chris Hecker's talk yesterday.)

By the way, Chen had the audience laughing the entire time. It was a happy treatment of death.

Next, Kim Swift (lead designer of Portal, yeah really) explained her idea. She said that she's been called morbid before. Then she talked about ridiculously morbid things. She thought about if she was going to die, how would that be. She has settled on cancer or a heart attack being most likely for her. She thought if she had 2 months to live, how would she spend it. Then she realized her game should be grief counseling for those who are diagnosed with a terminal disease.

Her game takes place in an office, then in a neighborhood, then I forget where, at home I think? Finally, in the hospital. You play a character who will die in two months, and you play it day by day. You have one meter for energy you an expend that day and another for karma. You use your energy to do helpful things for other people so you can make a difference in the world (increase your karma). There's all sorts of mini-games for helping coworkers sort paper or helping friends get the right ingredients to make spaghetti sauce, and so forth.

Each day, you have less energy available than the previous day. This simulates you dying, and becoming weaker. Remember how i listened to Heather's voice and found the tone of it interesting in the earlier presentation? I listened closely to Kim's here, trying not to hear specific words but just the tone. My suspicion was confirmed. Even the emotional tone, regardless of specific words, made me sad and made me want to lie down in depression. Her description of having less and less energy each day as you get closer to death was not really cold and clinical, it was depressing. My energy meter was depleting just sitting there.

When you're in the hospital, almost dead, with hardly any energy left, your last mini-game is to use the dual analog sticks (while viewing a closeup of the character's mouth) to make a smile. This way, your last act is a positive one. Again, her tone of voice made this depressing somehow.

I wondered who should really win this. The multiple choice thing about the will certainly shouldn't, but Chen and Swift left me really torn. If this were a comedy competition, Chen should win. He was the most entertaining. But did his game really address the game design challenge? It mostly did. Swift's depressing presentation made me want to die a little myself, but was that the point? Actually it wasn't, the goal wasn't to depress us, she said the point was to do grief counseling for those who have to deal with their impending death. Maybe her idea is closer to the truth of a "serious game" that makes a difference? I also wondered, what if her entire notion of what you do when you are about to die isn't right? I further wondered if that is even a valid complaint. Maybe we could still vote her the winner even if her particular method of counseling wasn't good because the point was to get in the right ballpark of a game that helps people who are dying at all. Or maybe there was nothing wrong with notion of dying anyway.

A guy from the audience then asked this exact question, about what if he wanted to die totally differently. He was very unsatisfied with her answer and as he walked away from the microphone he said "that is a very North American way to die."

Well whatever. Depressing truth or happy stock trading game? I abstained from the clapping vote.

Zimmerman summoned contest alum Steve Meretzky to the stage to assit with judging how loud the clapping is for each contestant. It went like this:

Heather/Erin: A lot

Chen: Even more.

Swift: Hmm...same as Chen? Possibly barely less? Or not? Close call.

Zimmerman and Meretzky conferred for a long time while we were all sitting in silence. One guy called out from the audience, "Seriously, we're DYING here." Zimmerman said "touche." He turned back to us and announced that the third place went to Swift (wait, what?), the second place to Heather/Erin, and the first place champion is Jenova Chen.

And that ended the second day of GDC 2010.