This is a short article describing our experiences and what we think went right and what went wrong during the development of Caribbean Pirate Quest.

Caribbean Pirate Quest is a pirate themed adventure puzzle game. It combines a smooth three puzzler gameplay with an adventure treasure hunt. It is Spell of Play Studios second completed and released game.

Project Inception

The first embryo of the three puzzler was written in C++ in 2005 under the name "Color" with simple graphics and no adventure elements, during the summer 2006 we decided to complete the project and to make a pirate theme for the game.

The decision to focus on a simple casual game was spawned by the fact that we had just abandoned a bigger more ambitious project. In this sense Caribbean Pirate Quest was much a second choice and not a thing that really motivated us. This is probably not the best way to start a project.

A first real version was released under the name Pirate Quest in autumn 2007.

Level balancing

Even in a simple game as Caribbean Pirate Quest there is a lot of balancing issues. We have spent countless hours playing through all levels, making them harder or easier depending on feedback. The big problem with balancing a game is that you have become an expert in playing it yourself. What you find fun might be to hard for the average casual gamer. We have tried to test the game on a lot of different groups of people from children to elderly. As a result, the game is much easier and not much challenge for us but the average casual gamer should not be stuck for hours.

One thing making the testing harder is that we randomize which level a player gets to play. This means we have not a clue which levels are played before another.

In retrospect we could have created three different difficulties and tried to automate collection of testing data. We could have divided the levels into categories and randomized inside the categories instead of having a large amount of randomness.

Slow Development

As mentioned Caribbean Pirate Quest has been under development for a very long time. Such a simple game should not take that long to develop and in many ways Caribbean Pirate Quest can be regarded as a dated casual game. It is definitely not as complex and pretty as many other new casual titles.

Our graphical artist abandoned the project during early autumn 2007, this was a big blow for the project. We now had a project with a lot of programmers art and decided to continue the project but on half speed and try to find art resources elsewhere. Most of 2007 was really spent fine tuning the core game play and trying to balance the levels. During autumn 2007 we got in contact with a new artist who agreed to complete the project. We quickly replaced the programmers art and made the theme a bit more colorful and "Caribbean". A more complete version was released during 2008.

One problem was when we realized the first artist did not upload all original graphics resources and since his hard drive crashed it was not possible to retrieve them. This meant we had to redo many parts of the graphics.

Maintaining the policy that all original assets must be on the server is also a good thing, and also that artists need to document what fonts and resources they use.

Eye candy

As mentioned before we had a lot of problems with graphics and this also shines through in the final product. There are indeed some wonky parts and lack of art direction in the game.

In retrospect, we should have created a simpler graphical style focusing more on effects.

Adventure Game

The adventure part of Caribbean Pirate Quest is what separates it from most other puzzlers our there. This is also the part that has been most appreciated by gamers. However, the actual searching for treasures is a bit of a contradiction to getting a high score, that is you can finish the game with a low score if you play in the smartest way. This is indeed not a good thing.

We should have focused more on the adventure part of the game, added more gameplay in this part and made the storyline the motivator in the adventure game.

Misc

Prepare early for localization, use unicode, fonts and art resources should be prepared for scaling and replacement.
Settle for a graphical theme early, we have changed the theme and elements of the theme a few times in CPQ and since artist time is expensive for us this is bad.
Keep a strong vision and only one project leader who defends that vision throughout the project.

Finishing the game

Not all games with a wafer thin amount resources get finished at all. We are very proud that we have finished and shipped one more game.

Change

Testing, testing and testing. We tested on every friend we could think of and listened to their feedback. We tested on people who where not used to playing games to find out if the help/tutorial worked. We released the game to the public early and often. We embraced change, if something was bad we changed it quickly. We killed our darlings if they no longer made sense in the game.
We focused on the things the player spends most time doing in the game. The game play in CPQ is smooth and well tested.

Voice acting

We did all voice acting ourselves, there are a lot of good resources for Pirate talk on YouTube and lots of dedicated pirate sites. Since we did not have access to a studio we built a improvised studio from cardboard boxes and borrowed a good microphone. The sound quality was good enough and we have got a lot of feedback on the fun voice acting.

We could have done a lot more voice acting, it is cheap if you do it yourself, and also quite fun.

Daniel Toll,
Tobias Olsson