When good games go bad, devs share lessons learned from failure

Making games is hard. People fail, often, and sometimes they’re willing to talk openly about it in the hope that maybe someone else won't make the same mistakes.

Game developers Chaim Gingold, Paul Kilduff-Taylor and Joe Mirabello did just that today at the Game Developers Conference, and developers new and old may appreciate some of their stories of trying -- and failing.

“I’d heard of the post-release blues, back when I worked on larger projects,” said Mirabello, a former triple-A dev whose first indie project, Tower of Guns, hit big in 2014.  “But I didn’t think it would affect me. I thought it mostly applied to other people, who had pet projects.”

Mirabello cautioned fellow devs that that's not the case; that he did experience post-release blues, but not in the stereotypical way. Instead of being too attached to Tower of Guns, he was excited to move on but too unattached to all the ideas he had after launching it.

“So I did what a lot of developers do; I retreated back, into supporting my game,” said Mirabello. “That focus probably helped make the game stronge,r but it was still be avoiding the problem." 

The problem, for Gingold, was that he was paralyzed by fear of failing after his first game was such a success.

Sophomore jitters can drive you into a tailspin of prototyping, reworking, and overscoping

“Because this was not my first project, I didn’t feel ilke I could show anyone anything that was worse than what I’d already shown," he said.  “There were expectations now, from others and myself, and I wanted to show the world now that I’d improved.”

To motivate himself to move forward, pick a new project and stick to it, Mirabello set himself some clear project goals:  work on something completely different; make something small, quickly; practice using new tech; and make something that looked nicer than his first game, Tower of Guns.

Mirabello started where, he admits, lots of indies do -- prototyping a puzzle-platformer.

“I never intended this to be anything other than a little filler project, while I thought about what my REAL project would be,” said Mirabello.

Eventually he scrapped that, and started prototyping a game about remodeling kitchens, even as his own kitchen was being remodeled around him. It was okay, but not something he felt he could ship (“would anyone want to play a game about kitchens?”) and pressure was mounting to announce a new project, since months had passed since Tower of Guns’ release.

“I was sleeping terribly; I was feeling sick often. I was basically in a full-on softmore slump, at this point,” said Mirabello. “I began crushing myself, as developer. As a former AAA dev, as a successful indie dev, I was SUPPOSED to know what I was doing!”

Then, his wife became pregnant. Mirabello knew his life would change in nine months, so he set a new goal: finish something before his daughter was born. He made a new prototype, one about a purchase order fulfillment ‘bot running around a cavernous warehouse in first-person, trying to fulfill enough orders to satisfy its corporate overlords.

“I tod my wife I’d be ready to show this protoype off within a month,” said Mirabello. He then built a quest tracking system, a solid save system, and…”I was going to be ready to show this game to people….in TWO months.”

Then he added a few more things, and bumped the reveal date back another month. He'd fallen into a "tailspin" of expanding his game's scope so he wouldn't ever actually have to show it to anyone.

“While I was doing this, I knew I was doing it; I STILL did it,” said Mirabello. He’s been in games for ten years, and he reminds other developers that he knew how important it is to get projects out early, get feedback, and focus on core “hooks.” Even so, he still couldn’t do it -- even as he watched himself spin out.

“My confidence was shattered at this point,” said Mirabello. “No matter what compelling points any of these prototypes had, none of them was enough to keep me moving forward.”

The process of rebuilding his confidence been pretty slow, admits Mirabello. His daughter was born, and in the wake of that experience he slowly started to focus his game dev efforts and is now working on a new project with partners.

“For the first time in years I have coworkers, and that’s already helping to create an ecosystem of shared confidence,” said Mirabello. “Teammates, by their very presence, offer daily vaiidation of what you’re building together.”

What did Mirabello learn? Mostly, that it’s good to fail while you’re a solo dev with low overhead; he only lost time and confidence, but gained a new perspective on his abilities, some new technical expertise, and most importantly, he’s learned to recognize the tailspin.

Indie developer Chaim Gingold shared a similar tale of lessons learned through failed prototypes, but where Mirabello seemed to overscope his prototypes out of a mix of fear and ambition, Gingold says his great failure was always being too ambitious -- and not finding that critical core gameplay loop that makes a good game shine.

"The year was 2007, and I had been working on Spore for four years,” said Gingold. “While I worked on many parts of the projects, the editors were my baby. And they were done.”

Focus is key

But he felt an itch to move on, to get out from under the studio structure and explore the (then much rarer and more mysterious) life of an indie developer. So he started, he says, with a game that came to be called Pocket Kingdom. It was a one-man project, in 2008, that aimed to be an ambitious online world-building game.

“Turns out, that was a big mistake,” said Gingold. “But onwards I plunged, deep into the wilderness.”

He felt he was an expert on prototyping, and failing fast; it was something that had helped him design the Creature Creator and other great editors on Spore, and he thought it was a strength that would carry him through a successful indie debut. But there was too much to do, and he couldn’t focus on prototyping and mastering one system because after throwing together one mechanic he’d had to move on and immediately start prototyping something else.

“I wasn’t prototyping; I was making the actual game,” said Gingold. “It was going very slowly, TOO slowly."

The thing he was missing, Gingold realized, was that the project had no “center” -- it had no deep, core system or premise. 

“A very smart designer friend of mine told me I needed to focus,” said Gingold. “It was a big realization for me.”

So he made a little iPhone game, finished it relatively quickly, and felt good in a way he hadn’t for some time. Also, he says, it taught him how much work it is to actually finish something.

“Finishing any project is really hard; it’s hard to have a focus,” admitted Gingold. “It’s hard to finish a project, even when it’s working; you have to decide when you want to stop, when you fix bugs….it’s exciting and exhausting.”

To finish a project that’s not happening, says Gingold, you have to find the energy to pull the plug.

“That’s really painful,” said Gingold. “I want to talk about the silver linings of this failure -- it’s how I find my way in the world, how I keep it together.”

As Gingold limped away from this tricky post-Spore period, he thinks he learned how to focus better, how to zero in on the simple, attractive core of a game’s design. He applied some of those lessons on his recent project Earth: A Primer, but he thinks he (and his fellow devs) still have lots to learn from studying the remains of failed projects.

“They’re buried, in the graveyard behind our houses,” said Gingold, and they're worth revisiting. One day, he says, he'll go back to that post-Spore period and try to salvage more of the good work he left behind.

Of course, you can also learn a lot from the failings of game that actually ships. In the wake of launching Frozen Cortex, Mode 7 Games' Paul Kilduff-Taylor took the stage to talk about what can happen when you trust yourself too much -- without considering whether your game's core concept is appealing (and adequately communicated) to your players.

"If you have a conceptual failure, you're going to be fighting against it your entire dev cycle"

“When conceptual failure goes wrong, it can go very wrong indeed,” said Kilduff-Taylor. “If you have a conceptual failure, you’re going to be fighting against it your entire dev cycle.”

Mode 7 came to prominence in a big way after the release of its 2011 tactical strategy game Frozen Synapse.

“It kind of made our name, as a studio,” Kilduff-Taylor said. "So when it came time to make our next game, we wanted to take the core of Frozen Synapse and change it.”

So Mode 7 announced Frozen Endzone (pictured), a sci-fi sports strategy game that swapped out Synapse's snipers and assault troopers for quarterbacks and receivers.

“However, the response to this was...not ideal,” said Kilduff-Taylor, pulling up a slide full of negative comments seemingly written by Frozen Synapse fans. They seemed to think Endzone was a football game, rather than a tactical strategy game in Cyberball clothing.

“This really indicated to me that the one thing we’ve gotten critically wrong with the game is, people thought it was a sports game,” said Kilduff-Taylor, even though Mode 7 meant the sports trapping to only be fluff around the game’s core tactical gameplay.

“Thinking back to all the games I really liked, and what we’d done with Frozen Endzone, we had this problem where there wasn’t a core fantasy at the heart of it. Nobody really wanted to be managing a futurustic sports team, with tactical strategy stuff,” said Kilduff-Taylor.

He then brought up a slide of Rocket League, and suggested it to fellow devs as an example of a sports game that de-emphasizes the fact that it’s a “sport” and emphasizes how fun and over the top it can be through things like images (animated or otherwise) of oversized cars leaping through the air.

Frozen Cortex ultimately did okay, review-wise, but Kilduff-Taylor says that wasn’t enough. The game was hampered by its confusing concept, in his estimation. So Mode 7 went back to the drawing board, and started work on something with a crystal-clear concept: Frozen Synapse 2.

With a trailer for the game on the screen, Kilduff-Taylor suggested his team got significantly more mileage out of a trailer that shows actual gameplay as early as possible and makes it clear, right away, what the core concept of the game is and what notable experiences a player can have in the game. It's something he recommends to other devs, to help ensure people can zero in on the exciting bits.

“When someone does something in your game, it has to be notable,” said Kilduff-Taylor. “If you can’t tell an exciting stoy about something someone did in your game that ties into the core fantasy of your game, then nobody’s going to talk about your game.”