At GDC 2011 on Wednesday, notable game designers who have seen big successes talked about their games that weren't so successful, giving attendees lessons so they might not make the same mistakes. During the conference's lighthearted "Failure Workshop," Kyle Gabler with World of Goo developer 2D Boy showed off the studio's failed game Robot and the Cities that Built Him, a Flash game that that has robots blasting humans. Gabler said the game ended up being a waste of six months of work because 2D Boy was more in love with the concept of the game, and not obsessed enough with making the game fun. Six months into development, Gabler said, "We did what we should've done six months earlier… which is building a gameplay prototype. I don't know why we didn't think of it sooner," he laughed. He showed a prototype of the game to the audience. "The point of this is… aren't you so bored? We realized our game had never evolved [from the initial idea]." Gabler explained that the game just "wasn't us." Ultimately, he learned from the failure that "No amount of theming will save a bad idea, also known as 'you can't polish a turd'" And coming off of the smash success of World of Goo he warned that "trying to live up to a previous game is paralyzing. Don't bother competing with yourself." Instead, make a game that can't be compared to your previous success. PopCap sr. game designer George Fan, creator of another hit, Plants vs. Zombies, showed his 2001 failure, Cat-Mouse Foosball. A bit of a mix between Lemmings and Bomberman, the game was, in his words, a mess, and just wasn't as fun as he thought it would be. He dumped the game after one prototype. "There's just too much going on during the game," said Fan. "… It just becomes this not-fun experience that overwhelms your brain in a bad way." He remade the game just for the Failure Workshop. It was still bad, and he got big laughs from the crowd because the game was so ridiculous and manageable, with players shifting foosball rods trying to keep 2D cats from touching 2D mice before the mice got to the end of a long maze. But even though he can laugh about it now, back in 2001, it left Fan very disheartened. "I realized it was really bad and that made me sad. I couldn't believe how much disconnect there was between how much fun I thought it could be and how much fun it actually was." It was then when he actually thought about quitting game design. Of course, Fan continued, and is known for a beloved PopCap game in Plants vs. Zombies. He told attendees to work through the tough times. "Don't give up. If you love what you do, you will persevere," he said. "...I chose to learn from it. … Don't let it stop you, the same thing happened to me." He added, "I realized the only way you can get to the mindset of the gamer is to play your game yourself. Start prototyping the game as early as you can. You can never truly tell how fun a game is until you play it." Flashbang developer Matthew Wegner brought his studio's failure, Off-Road Velociraptor Safari HD, to the workshop. The game has players driving around in a jeep among jungle settings, running over dinosaurs. The game was a follow-up to an early Unity-based web game Off-Road Velociraptor Safari. The "HD" version was supposed to be a console and Steam downloadable title. But trying to go "HD" was a mistake, said Wegner. Even calling the game "HD" was a mistake, because that creates certain expectations with gamers and potential publishers. "It turned out we actually really hated working on this," said Wegner. "[Art director] Ben Ruiz would be working on normal maps on rocks for an entire day." The development process just became tedious. Wegner said that the studio just tried to play things too safe -- they had a concept that worked with the first version of the game, and just tried to pretty it up. "Our mistake was that we stopped designing for players," Wegner said. Designers should think about what does the player find fun. "Maybe we should make games fun," he said. Over the course of about a month, Flashbang totally revamped the graphics to make them more stylized, but the studio isn't really working on the game extensively any more, Wegner said. Stardock's Brad Wardell also took the stage to describe his studio's failure in Elemental: War of Magic. The company was coming off of big success with Galactic Civilizations II -- a very well-received strategy game that cost $600,000 to make and generated $11-$12 million in sales. The excitement from that game's success led the studio to increase the scope of its next title. "Let's make our next game much more ambitious!" Wardell said was the attitude at the time. Wardell upped Stardock's game development team from seven members to 18 for Elemental. Then it became a case of too many cooks in the kitchen. "There was no structure, no dedicated project manager," he explained. "No one owned any part of the game." At one point during the game's development, Wardell was at his office 100 hours a week, pulling all-nighters every Tuesday and Thursday. "I just figured I'd add 40 hours a week to my day job," he said. When the game launched, it earned an abysmal Metacritic score of 56 and was filled with bugs. He countered claims that Stardock launched the buggy game knowing that it was unfinished. But he said the studio was under the impression that the game was complete. Wardell said Stardock actually launched the game two days before the launch date because the team was excited about the release. "We thought it was done," he laughed. "We wanted our players to see the masterpiece that we put together!" But players and critics weren't pleased. Wardell said he encountered relentless "nerd rage." "I had people tell me that they hope I die of cancer," he laughed. Now Stardock has hired on staff that will help handle more production aspects of game development. Among some of Wardell's tips for other developers included: the project manager shouldn't be coding or doing art; the project manager should be the only interface to QA; the project manager must come down hard on scope creep; if the project manager isn't the lead designer, bring on a full-time lead designer; and consider bringing on an associate producer to help fill in the role. He laughed, "Unlike any of these games showed so far … this game shipped. It's at the store right now! You can buy it at Walmart!" Wardell is trying to make things right with gamers by releasing an Elemental expansion pack for free to anyone who bought the original game in 2010. "We're just going to bite the bullet, spend a couple million dollars - it was a pretty big mistake," said Wardell. The dynamic Chris Hecker then showed off his physics-heavy failure -- a rock climbing game. Hecker, who worked on Spore and is currently at work on his independent game SpyParty, said there were three things that went wrong with the rock climbing title: technology ratholing (doing tech that isn't necessary), non-game distractions and "lack of ass-in-chair." Hecker showed off a massive amount of code for the game, drawing laughs from the audience. Under the game's hood was also an extensive fatigue model for the climber. "I kind of rediscovered chaos theory on my own," he joked. "I realized when I had some distance on this [situation] that these three things seem like distinct independent things… but all the things went back to one core problem: I was scared of game design." Hecker said he would create a lot of tech problems himself just to distract or postpone him from doing any actual design work. Design truly can be frightening: "Design is hard, unpredictable, mysterious and unstable," he said. But now he's learned to approach game design more directly. "Everything I do is playable that day," he said.