We all know that Twisted Pixel's 'Splosion Man is a great game. But do you know how it was made, what went into it, or what mistakes were made during its development? I didn't either, until I attended Mike Henry and Sean Riley's postmortem on the game at GDC Austin. There, they dished out the story behind it all, with lots of 'splosions and meat throughout.
Hit the jump for a summary of Mike and Sean's thoughts on 'Splosion Man's journey from an unpolished project to one of the most critically acclaimed games on XBLA today.
Mike and Sean's opening slide bore the words, "Making an XBLA game in six months.. what were you thinking?" Looking up at it, the guys laughed and responded, "Just don't try to make a game in six months. But if you have to.." So began the postmortem for Splosion Man, Twisted Pixel's latest and best selling game.
First, "Things That Went Right" were highlighted. The most important thing was not clinging to dying mechanics. Doing so is a good way to waste a lot of time with something that will never work. As they got deeper into the development process, a lot of old, unworkable ideas were scrapped in order to make their work more efficient. Another area where they did well was with the tools they used to create the game. These programs did a lot of the grunt work and allowed them to focus on what they had and polish it.
In the prototyping stage of development, the main philosophy that Twisted Pixel had was "ugly and quick". In their words, "Have a good game before a pretty game." Not only did this help them work out the kinks in the mechanics quicker, but it was also a good way to inspire creativity. They loved to see how the designers worked with what little they were given. Some pretty ugly pictures of the old version of the game were shown. 'Splosion Man was in his familiar shape, but all one color, the levels were barren, and the game's robot enemies were just grayed scientist models with extra shapes drawn on.
Focus was another thing that they feel they did right making 'Splosion Man. "Think only about what's most important about the game," they told the audience. In their case, the most important thing was polish. Mike and Sean both admitted that having that as their focal point was kind of silly when working on a six month model, but they felt that not having enough polish was the "biggest trap to cause the game to fail".
This meant making a lot of cuts on ideas. Originally, they wanted to have 75 levels for each mode, special power meters, power ups, and several other things that would make the gameplay more than just normal 'sploding. But the further they got into development, the more they started to take away from the game. They shaved down these extra components so that they could work more with the core mechanics.
They used priority lists to help them figure out what was important enough to keep in the game and what was not. Beta testing was also a huge help. The guys think back to one of the latest additions made to 'Splosion Man, the Way of the Coward option. They saw that a lot of players were dying a lot during testing, and knew that it was important to find a way around it without compromising the game's preexisting design. As many of us who have played it know, the game can get very hard, but it's not because it's designed badly.
While playing a video of a very early version of the game up on the big screen, they mentioned that they did a really good job with making this first demo. They were able to fit an example of every single part of the game into it, from simple 'sploding puzzles to moving platforms and barrels that propel 'Splosion Man across the map. It wasn't very pretty looking, but it was a great display of every type of gameplay present in the final build. They were especially proud that they could fit everything into a single demo, because "if you can't, your game has too much going on".
Then they finally got to talking about "Things That Went Wrong". Unsurprisingly, the devs said that developing 'Splosion Man was mostly remarkably smooth. But among the bad was the pressure of time. According to Mike and Sean, the folks at Twisted worked 90 hour weeks, and way too hard to get the game out in the allotted six months. "It was partially our fault, because we were so passionate about the game. But even with passion, people can still be burned out." To make these hours more manageable, they formed core hours for workdays so that people could make their own schedule. Each person was required to work during core hours, but anything beyond that was left entirely up to them. They were even given the option to work at home those extra hours so they could still spend time with their families.
They also had some problems with networking. They had to retrofit 'Splosion Man into Live online play, and if they had not gotten an expert to help them out, they may not have been able to make the online multiplayer work. Not focusing on that at the beginning was definitely a strike against them, and something they will know better than to do next time around.
On a similar note, next came "Things We Wish We Had Known". They had some regrets with the way they programmed some objects and how they handled testing. They believe that things would have worked out better if they had started testing the game earlier. Too many improvements went in post alpha and post beta, including the aforementioned Way of the Coward and the countdown timer for co-op mode. There are also a lot of things they feel "went in broken and stayed broken".
Going back to the cuts they had to made, they lamented the fact that they did not know how painful it would be to take things they would have really liked to be in out. Better death animations, better boss functionality, and better tutorials would have been in the game if time were not a factor. But they felt like they would have been unable to polish the game with so muh extra stuff in it, so again, they stripped it down to what really mattered. It seems their biggest regret is making themselves so pressed for time, but they seem to have learned their lesson.
To wrap things up, the guys took a few questions from the audience. The first person asked why the six month schedule was chosen in the first place. Surprisingly, they answered that they wanted the game to be ready for XBLA's Summer of Arcade. Statistically, it's the time of year when Microsoft puts the most effort behind marketing Live games. It's also set during a great time of the year to release a game; it's before the Christmas rush, and in a dead zone where not a lot of other new games are coming out at the same time. Twisted Pixel wanted to get 'Splosion Man in so that it could reach more people.
The next question was asked by an audience member who was curious about the cuts made to the game. He asked, "Was there ever a point where you thought about abandoning multiplayer to focus on the main game?" The guys responded that even though they did think about it every now and then, they stuck with it because they wanted to try and establish themselves as multiplayer game makers. They admit that there was a lot more to them at first, but again, a lot got cut out so that they could focus on the cooperation mechanics. "We scoped it down, but wouldn't cut it out completely."
An audience member was curious about how long it took for Twisted Pixel to develop the tools used to make 'Splosion Man. They explained that they had the basic layout generation tools already made from The Maw. The only two tools they had to create for 'Splosion Man helped with Live gameplay functionality and 2D level generation (as The Maw was 3D).
And with that, the 'Splosion Man postmortem came to an end. Having a bit of insight as to how the game grew from such simplistic ideas and builds made for an incredibly interesting talk. It was also refreshing to see the team come to terms with all of the mistakes and things they wished they could have done for the game, but overall, it's very apparent that they are still very proud of the final product. With good reason! 'Splosion Man was such a great game to only be Twisted Pixel's second, and to only have six months of work behind it. They are sure to use the lessons learned to continue to grow as a company.
[*].disqus.comto your security software's whitelist.