Excit: Post-Mortem

There is an interesting custom among game designers to write so-called “Post-Mortems” about the games they made. So far I know, filmmakers, musicians and writers do not handle their work in such a scientific manner. I find Post-Mortems an excellent tool to establish a systematic and reflective approach to design, especially game design. So for a change, this time I would like to talk about a game I made: Excit.


One of the more eccentric Excit levels.

Excit is a very simple and quite small flash game. It is not a pinnacle of game design but it was a very interesting project nonetheless. I’ve learned a lot about game design through Excit and I would like to share some of the findings:

History
In 2005, an advertisement company I was working for as a contractor approached me with a project. The task was to create a viral marketing tool for a quite successful business software company called MIS AG. At that time, they were thinking about “something with animation” but I quickly got the idea to make a flash game for them. They agreed to finance the development of some idea sketches and I got my colleague Daniel Renkel on board. Together, we came up with A LOT of different ideas although some of them were pretty vague. We presented them to MIS and they decided to go for a clone of Roadblocks. I think they favored that idea because it was the only one where there already was something to play with. We had a lot of original ideas but it is that much harder to sell them when you compare them against a game that is already out there and is obviously fun to play.


Before developing the game, we had an excellent opportunity to try a lot of different game ideas. Here is a (bad) photo of a mind-map we used in the process.

So we developed the game over the next couple of weeks. Basically, I did the programming and most of the visual design. Daniel did most of the level design. We also got Julia Zaadstra on board to help with some of the graphics and the balancing later on. She also came up with the name.
We delivered the game on-time but unfortunately, it didn’t get released. By the time it was finished, the company we did it for got into some incomprehensible business mumbo-jumbo. We already noticed that when we had to change company’s logo in the game 2 times mid-development. After two years or so, the company didn’t really exist anymore – at least not in the original sense. We tried to contact them but we’ve failed to reach anybody. It put is in a quite funny position where we made a game and got paid but nobody played it. I can tell you, having money is fine but it is not the reason why you are designing games in the first place.
So in winter 2007, we’ve decided to release it on our own. It was not a big deal. Basically, I uploaded the thing to a friend’s server, got the “Hall of Fame” working and we posted in on 3 forums. I was expecting maybe 100 people playing it. I was wrong. Now, one year later, we had 2,5 Million players. The game turned out to work pretty well as a viral tool. Too bad the thing it advertises doesn’t exist anymore.
We had some people approach us about licensing and re-branding Excit for them. As I will explain, we had to turn down a lot of the requests. But recently, we had a interesting opportunity again and so we gave in. I re-wrote the game in Action Script 3.0, Daniel designed 30 new levels and we will soon release a new version. This new release is the reason for the Post-Mortem.

What went right: The Power of Animation
As I have already explained in a short post on Game Design Scrapbook, one if the main strengths of the game is some detailed, juicy animation. Here, we picked up where Roadblocks left off. Roadblocks already had some nice squishy “animations”, especially of its “main character” – the ball. We liked that very much so I intuitively started out with implementing such special effects at the very beginning of the development. Even before most of the features were implemented, we already had our own squishy spreadsheet cursor bouncing all over the place. Because the animations were in so early, we tweaked and balanced them to death. I remember spending a whole day just trying various algorithms for the particles effects for the teleporter animation.

What went right: Overdoing
Actually, we used this very anal approach for pretty much every element of the game. We made several different versions for the visual design of every special block. We made sure it was both – easily understandable and in line with the game’s overall look & feel. We also dropped various blocks from the game because we figured out that they didn’t contribute to the game experience.

Here is a selection of various concepts for the visual elements in Excit. Click on the images to embiggen.

I have also spent countless hours searching for the perfect sound effects. They had to fit to the animations and the overall look and feel but shouldn’t sound annoying or obtrusive. I now frequently realize that this is where many flash game developers think they can cut corners.
I guess our approach was that if the game idea wasn’t so innovative, all we could do was to make the experience as smooth and polished as possible. Here, the old Mythbusters slogan paid off: “If it is worth doing, it is worth over-doing”. We knew the gameplay is fun and rewarding. So we were certain polishing it would make it even MORE fun and rewarding.

What went right: Unique, Meaningful Theme
One thing that most of the review sites and blogs immediately picked up when they wrote about Excit was it’s theme. We envisioned a spreadsheet look pretty much from the start. It was part of the whole idea of developing game for a business software company. Let’s face it: business software is not something you would normally associate with fun. But we could boil down their products to one thing: they were writing software to do things that most company solve with standard spreadsheet programs. So the slogan “Escape the Spreadsheet Chaos” was a perfect fit to illustrate what they offer.


This is an early photoshop mock-up of Excit. Although the final game does look a bit different, the main features remained exactly as envisioned.

The spreadsheet theme had two advantages. On the one hand, it gave us a clear target to work towards when developing the interface. Every decision we had to face was quite straight-forward: just make it look like a spreadsheet. Obviously, we still had some tough decisions because it wasn’t a spreadsheet after all. However, the clear goal already established some parameters to work with so we didn’t have to re-invent the wheel every time.
The second advantage was that it is a unique theme. It sets Excit apart from other games like it and makes up for a great opener when people are writing about the game. It makes Excit immediately recognizable and it is the theme which created follow-up interest and allowed us to re-release the game now.
The important thing about it is that is not ONLY unique. It also has a purpose. It is not just some arbitrary, pretty “skin” on an otherwise old game idea. We think this is actually an important advantage of Exit over it predecessors like Roadblocks or Orbox (or most puzzle games anyway), where the graphics are often nice but arbitrary and meaningless.

What went right: Meaningful Level Design
The one responsible for the game design was Daniel Renkel. He is actually the one who proposed including a Roadblocks clone among our initial proposals. Daniel was pretty enthusiastic about the game at that time and he put a lot of that enthusiasm into the level design. It shows in the way each level has a distinct identity. The levels are not just random challenges, they differ from each other in visual style, pacing and in the way they play out. Many of them are centered around a unique idea. For example, Level 29 features a set of “rooms” closed with door blocks. There are some keys scattered around the level but the player will soon realize that there are less keys then doors. So the players must determine which doors need to be opened in order to solve the level. Showing the “rooms” as fenced-off areas is technically not necessary, but it helps to understand the concept of the level and to distinguish it from other levels.
As a contrast, a game like Orbox had some excellent, challenging levels but they almost exclusively consists of featureless arrays of scattered blocks. The levels don’t seem to have a distinct identity. I think it is this meaningful level design which makes Excit slightly more engaging and sets it apart from the competition.


Level 29 is a good example of meaningful level design in Excit. Each level has a distinct identity.

What went wrong: Balancing
While we are at the subject of levels, there is also a lot wrong with them. One thing I was never quite happy with is how the difficulty curve of the game plays out. We decided to present every special block in the game with a very simple tutorial level. Because we didn’t want JUST tutorial levels at the beginning, we thinned them out by adding some “normal” levels in-between. Because there are 6 special blocks, the tutorials are over around level 15. At this point, the “real” levels begin and the difficulty curve suddenly takes a steep step upwards.
Judging the difficulty of the levels turned out to be very difficult. For example, we decided to move some levels towards the end of the game because even if they weren’t that difficult, they certainly just looked daunting. This is especially the case in level 30. In other cases, there were levels where players suddenly were stuck even though the solution was staring them in the face. Level 21 is one such an example. In order to solve it, players must enter the same teleporter twice from two different directions. For some reason, most people don’t consider it and get stuck. Much to the bewilderment of Daniel, I had to move this particular level way further towards the end of the game than he initially anticipated.
Re-making the game and designing new levels, we found out that we still haven’t mastered the art of correctly judging the difficulty. I’m afraid the new version is still not well balanced.


Although it isn’t that difficult to solve, level 30 earned it’s place simply because it looks extremely intimidating.

What went wrong: The Release
When we initially released the game, were were completely oblivious about how the flash game business works. We didn’t implement a sitelock, we didn’t implement Mochiads, we didn’t implement Mochibot, we didn’t even implement AdSense. We didn’t expect much traffic anyway. A day later or so, I was stunned to see that our “Hall of Fame” got a huge amount of entires. Looking at the server statistics I realized that we completely exceeded our expectations. I put in AdSense just to cope with the bandwidth costs. A few days later I also implemented Mochiads and Mochibot. Still, we missed a lot of the hits at the most crucial time and some of the versions without Mochiads still got out to various pirate portals.
Later on, we ran into some more problems because our Flash movie is over 700 pixels in width which is the maximum for posting a game on Kongregate. It took a bit of work to get the game on Kongregate and the result isn’t very smooth either. Those may be little things but they sum up and we certainly gained a lot of experience on how to properly release a flash game.

What went wrong: Small Features With Big Impact
Although we took so much care into polishing the game, some areas still fell short. One thing we never were satisfied with was the scoring system. It was flawed from the beginning. The game is not a typical game for a high score board so getting the “Hall of Fame” to work was a nightmare. Typically, you would have 30 levels. Because it is difficult to solve them in one session, we implemented a password system to jump straight to a latter level and continue playing. However, this was in conflict with the idea that you could post your highscore after level 30. If players used a password to skip some levels, their score for the levels they skipped wasn’t registered. Players, who solved everything in one go, would have a MUCH higher score than players, who used a password. The problem got even worse when some achievements were added at Kongregate. One particular achievement required players to finish the game in one go and even collect some of the elusive MIS Logos. To top it off, because of some mysterious connection issues with the Kongregate system the site sometimes fails to register some of the player’s results making that particular achievement a very frustrating experience. We received a lot of complains from players for not implementing a cookie-based saving system as most modern flash games do.
But that is not even all. The way our scoring system is set up, we don’t even have a single score. Instead, we sort the database of results on 3 different values (MIS logos, moves, time). While in some ways simpler to balance, it is less transparent for the players. It also actually prevents us to connect our game to most of the High-Score APIs out there, like Mindjolt’s for example.
They seem like small details but they ruined the experience for a lot of players and made us miss some important opportunities.

What went wrong: Inflexiblity
The game was written in Action Script 1.0. In 2005, Action Script 2.0 was still “the new thing”. At least for me it was. I was pretty inexperienced in flash development. I learned much of what I know today about flash actually while writing Excit. As a result, the code of Excit was pretty convoluted. I used no OOP what so ever. The code wasn’t even in a separate file. It was embedded directly in the *.FLA and scattered among different frames and invisible off-stage movie clips. So naturally, changing anything was a difficult thing to do, especially after two years, when I lost track of how everything works.


The original Excit was coded quite chaotically. The code is scattered among various frames and off-stage movie clips. I lost track of how it all comes together. This was the reason why I decided to re-code everything in Action Script 3.0.

So when people approached us with requests to license a modified version of Excit, we had to turn them down simply because we couldn’t handle the changes it would require. Since it was all in Action Script 1.0 we had even some problems implementing helpful tools as most of APIs are designed for Actions Script 2.0. Together with the heavy MIS AG branding, we navigated us into a pretty inflexible position where we had a cool game but couldn’t develop it further. This is one of the main reasons why we decided to completely re-write the whole game in Action Script 3.0.

What went wrong: Mis-communication
One of the most frustrating experiences in the project came from mis-communication (pun intended). We already had our share of some serious internal problems, especially at the end of the project. To add insult to injury, the communication with the client was even more broken. To be fair, the project had an excellent start. The fact that they even funded the concept phase was amazingly professional. We would like to have some more projects that begin so well. However, as the game came together, the client started suggesting a lot of changes which we weren’t very happy with. One particular thing I found quite annoying is how they insisted on having their own people do all the in-game text. We had to take all the text from the game into a Word document, mail it over to them, get it re-written and implement it back. And this wouldn’t be even half as bad if the people who wrote the texts actually have any idea what the game was about or how it even looks like. We ended up having whole paragraphs of text cut down to a singe sentence, leaving holes in our layouts and bewilderment on the player’s faces. Initially, we even had cool tips and hints to EACH single level – they were all cut. It was a very stupid setup. I can’t even blame the person writing the texts – they couldn’t leave the texts as they were because somebody paid and expected them to mess around with them after all.
We couldn’t even argue with them because all communication went trough the advertisement company. Finally, it was our first flash game so we didn’t really have much experience to back our arguments with authority. And in the end, there were even some financial problems so we had to fight for over a year until we were fully paid.

What we are going to do differently now
The project was a great source of experience. I’ve learned a great deal about Flash game development. I think it is a great platform with some unique advantages over other gaming platforms. I wrote some about it in a recent article. I realized that it might be even an interesting business model and I would like to try some more Flash projects – possibly even make a living with it one day.
Re-doing the game was great because we could finally get rid of a lot of problems that have been taunting us for a year now. It made me understand the unique qualities of the game better. If we are lucky, it might create more possibilities for us in the future.


One of the things that never worked right was the scoring System. We were glad to fix that in our recent re-make.

On the other hand it is somewhat discouraging to see how some of the problems we had seem to re-occur. So when we made the new version, we still had a hard time getting the balancing right. Also, while the communication works a lot smoother then with our old sponsor, the clients we work currently with exhibit somewhat similar behavior. We find ourselves again and again in situations where clients second-guess our decisions and ask us to do changes in the game we KNOW will make the game worse. In the future, we might simply turn to flash portals as clients as they seem to have more respect towards our game design experience.
Finally, I would like to thank everybody who gave us feedback on the game. We are extremely happy seeing how many people seem to enjoy the game. For those who don’t: we will certainly try to improve! If you don’t know Excit yet, try the original one here. The new version is available here. It is in German but that shouldn’t be a problem as the game isn’t text-heavy. You should try it especially if you already played the original as it features brand new levels!