[This Gamasutra-exclusive postmortem covers the Eidos-published, Deadline Games-developed PSP exclusive Chili Con Carnage, which debuted in the U.S. in February 2007.]
Hola Amigos, Como Están?
I always find it difficult to write a post-mortem. There are so many experiences throughout the development of a game that it’s difficult to single out specific events or issues that you feel represent the essence of what went right and wrong. Nevertheless, looking back at the development of Chili Con Carnage there are definitely some lessons to be learned and remembered. I’d like to take this opportunity to share them with you.
The whole premise for making this game was that at the end of Total Overdose (released in fall 2005 for PS2, Xbox and PC) we felt we stood with a game that was fun and filled with good ideas but hadn’t managed to exploit the full potential of the core game mechanics. That, combined with the soon to be released Sony PSP, made us decide to make an adaptation of Total Overdose for the PSP. It was quite important for us that it was an adaptation and not a port as the game had to fit the handheld platform to achieve the greatest play experience.
We decided from the outset that we wanted to focus on the gameplay and create the best action game for the PSP. Furthermore we wanted the project to take no longer than a year so decisions were made very early about the ambitions with regards to new graphics, environments, story and music. In essence we felt that we could make a new and fresh game experience re-using many existing Total Overdose assets without compromising the quality or the goal of wanting to create the best action game on PSP.
So on we went, porting our proprietary technology to the PSP while developing the new game in our flexible PC development environment…
What Went Right
To be honest, there were a lot of things that went really well when developing this title. Somehow, most elements just seemed to “click” and it was pretty smooth sailing along the way. I’ll try and emphasize the areas I feel went best.
It’s a fairly clichéd point to always emphasize the team but we really had a good constellation of guys on the project. We had a new team leading the project with some Total Overdose “old timers” in supporting roles. This gave the necessary fresh look on the Chili feature set and made it easier to kill any darlings that the old team may have had.
Although the majority of the original Total Overdose team were working on their next project they were still so passionate about the old game that they were happy in assisting us whenever we had questions or problems and came with a lot of good advice and ideas that helped make Chili the game it is today.
Last but not least our own QA team was also great help when developing features providing us with feedback and statistics from copious amounts of play tests. I really urge anyone who has the possibility of using a QA department in the development process to do so.
Great Base Material
The original Total Overdose team had so many great ideas and had so much drive to hone them that it made our jobs on Chili so much easier. We could basically cherry pick the best features and focus on them instead of trying to cram the entire PS2 game onto the PSP. This of course reduced the risk of having to invent a completely new game from scratch as we had our “feature prototype” in the already released game. You can say that we had the rare privilege of “finishing a game instead of shipping it”.
As you may or may not know Total Overdose and Chili Con Carnage both have their tongues firmly stuck in their cheeks and that just gave us so much leeway in how outrageous and silly we could get. We had a great time blue-skying the game and many design meetings felt more like attending a comedy performance than actually designing a game and it definitely made everyone feel really good about the game we were working on.
Simple Camera And Controls
Our proprietary Kapow Engine is highly flexible and puts a lot of power in the hands of the designers and not just the programmers. Our tech department assured us that the existing feature set would be available on the PSP once they were done porting the tech. That meant we could basically start game development from day one on the PC instead of having to wait for certain features to be available.
This in turn meant that we had time to experiment and create many control and camera prototypes trying to home in on the optimal configuration for the “challenging” PSP controller layout. Combine that with the awesome work of our QA department and their play testers we had the perfect foundation to reach good if not excellent controls and camera on the PSP.
You could say that this is true for every game, but the fact that we could find a fairly well balanced control scheme very early in the development process meant that playing the game, testing it, iterating it, testing it again became so much more enjoyable for everyone and therefore the team was probably more likely to go through more iterations (thus achieving better results) than had the control scheme only been nailed down later in the project.
Game Format That Works
Our decision to create an adaptation and not a straight port of Total Overdose meant that our core feature set such as level design and sizes, enemy behaviors and core gameplay mechanics were all tuned for short play times, quick progression and instant gratification. The whole gameplay experience emerged into this snappy, frantic, instant fun type of affair as opposed to a lot of other games where transport, story and pace all work against the action principle and just slow things down.
It’s a format that I’m happy to see more and more PSP games adopting. That means that our decisions were right at a time where ports to the PSP were more common than new games or adaptations.
Everybody from tech guys, game programmers, level designers, graphic artists, animators and the sound designer were very passionate about the game and where we were taking it. On top of that the leads were very good at structuring the work load for their respective teams and that combined just gave a lot of confidence in the direction, scope and schedule.
My lead artist told me that he had clocked a total of 50 man-hours of overtime for his team on the project. The scenario was the same for all the other crafts groups and I’m happy to say that we were really good at scoping our ambitions and adjust deliverables to minimize the amount of overtime needed on the project instead of vice versa. The next goal must be to reach zero overtime.
As a funny side note I can tell you that the team was so dedicated that once when I came back from vacation they had made an entirely Danish version of the game, complete with voiceovers and whatnot, all in their own time. They had had so much fun doing it that the whole experience was a great morale booster. Unfortunately the Danish version didn’t make it into the final game as the QA time needed wasn’t scheduled and budgeted for (we can thank the Hot Coffee scandal for ruining impulsive easter eggs forever).
What Went Wrong
As always, nothing is picture perfect and we sure made a couple of mistakes when developing Chili Con Carnage. Here’s my take on the biggest ones.
Game Too Short
We were too confident that players would “get” the game and our desire to make players perfect their score and combo-strings. Reading reviews and comments it’s quite obvious that the majority of players don’t have that mentality and would much rather prefer a “play once” game experience. Maybe it was something in our presentation of the levels or we didn’t make it obvious or desirable enough to increase your scores but the end result was that many players felt the game was too short.
It’s also quite obvious that the additional game modes we included did not count towards extra play time and as such the time spent on those could have been spent on increasing the “story” mode of the game.
Last Minute Filler
As part of an attempt to prolong the game experience we added the challenge missions quite late in the process. As such they did not turn out as great as we would have hoped. The challenge missions were part teaching the players how to get better scores hopefully enticing them to go back to previously played levels to maximize their scores and part breaking up the normal levels with small mini-game like experiences.
It was a difficult balance between requiring players to complete the challenges and letting the players decide whether they want to play them or not. In the end we decided to let challenge missions be optional, which I still think was the right choice, but it didn’t change the fact that they still felt tacked on.
In order to reduce scope, save resources and manage to bring the game to market in a year we may have cut too many corners relying too much on the fact that the “distilled” game experience would be enough to satisfy the player.
Not Enough Unique Art Content
It was decided from the get go that we weren’t going to invent a completely new graphical style. We had these great over the top characters we felt we could bring back, a great texture library we could lift from Total Overdose and a lot of props we could recycle. We chose to focus our artist time on creating new end-bosses, new levels that fit the handheld platform and control scheme and worry less about creating new unique art content.
The reasoning behind this was that there probably weren’t going to be too many players that would own both Total Overdose and Chili Con Carnage and therefore there wouldn’t be too many that would feel “cheated” about recycled graphics. Obviously we hadn’t accounted for the reviewers that had all played both games and as such had a difficult time distinguishing the two titles as they more or less looked the same. We should probably have spent more resources on creating a new main character and supporting cast.
To reduce risk in developing a multiplayer game mode, which we hadn’t done here at Deadline Games before Chili, we decided to go for a non-synchronized ad-hoc game mode that would be easier to implement (no infra-structure mode enabling players to play over the internet). The result was the fiesta game mode where people would play on their own arenas sending enemies and “bad events” between each other to try and achieve a set score first. It was a great and fun game mode but it was a concept very different from other multiplayer game modes that players are used to and therefore it was probably misunderstood.
Ultimately I think we could have spent the time better extending the single player experience instead of introducing a multiplayer mode where people had to have their own copy of the game each and be close enough together to player via the WiFi ad-hoc mode of the PSP.
No Control Customization
We felt that we had found the ultimate control configuration for the game. In fact we have two configurations that the player can choose between, but eventually I don’t think that was good enough.
The PSP ergonomics are so “difficult” that every player has his or her preference as to how the button layout should be and I think the game experience may have been better for a lot of gamers if we would have allowed them to customize the controls to their liking. It is something that would maybe have taken a total of two man-weeks to implement (what with the tutorial text and everything) and probably would have made the game better for some players.
Game Premise Changes
What you don’t know is that Chili Con Carnage started life under the game premise Total Overdose: Exaggerated. We were to take the original locations and present them as three different versions of the original Total Overdose story. Each version would be an even wilder exaggeration of our main character’s retelling of the events to his colleagues, friends and babes he was trying to pick up. Level design, game flow, story script, cut-scenes etc. were all developed with this in mind.
Quite late in the development process Eidos came back with some requirements from sales and marketing that meant that we had to change the whole game premise. Not something you need just five months before Alpha.
From a sales point of view the changes made sense and as such the changes needed were actually good but needless to say that we had to hustle to write a new story, record the new script, completely change our cut-scenes and alter the game flow to accommodate this. Ultimately I think it took away some of the time we could have used for even more polish or an even longer single player experience.
At the end of the day, once the last piece of code has been entered, the last pixel drawn, the last vertex of a model tweaked and the game is on the shelf I have to conclude that developing Chili Con Carnage has been one of the best development experiences in my professional career. There were many ups, few downs and just a lot of fun and great challenges making this game. Looking at the reviews it looks like the majority out there feels that a lot of the fun we had carried over to the end result.