Battleships Forever

I’m Sean Chan and I created Battleships Forever. I’m Singaporean and when I started making games I didn’t have any programming background. I’m rectifying that now but I started this project with just a Diploma in Communications (specializing in TV production).

I’ve always liked creating games. When I was younger I not only built and played with Lego, my friends and I had our own game system built around using Lego as units and the tiles on our floor as a movement grid. Instead of just playing Battleships during classes, I used to draw little maps of the Mario games that you played by tracing your way through.

I think we’ve all played games which had so much potential but somehow or other the developer messed up something crucial. Well, I got tired of letting that happen and decided to do something about it.

I started off making custom maps for Warcraft III; I made the original Tank Commanders map over there. After awhile I realized that my game ideas were growing too large to be implemented in custom maps so I moved on to Game Maker.

It’s the motivation behind your work that really sets you apart. If all you’re chasing is dollar signs, you’re going to end up with some pretty mainstream stuff. On the other hand if your only concern is making a great game, then you’ve got the indie spirit nailed.

This is a good thing. Consider what YouTube and cheap camcorders have done for films. Lowering the barrier to entry is pushing and expanding the medium at an incredible rate. Not only does that mean that we see more fresh ideas from more people, it also allows a wider audience to appreciate the subtleties of game design.

The gamer audience is still a little underdeveloped in this area. If you hear two buddies talking about a game, you’ll hear them rave about how “awesome” it is. They might quip about a couple of features but generally the only way they can describe a game is by using a superlative. We just don’t understand what makes a good game yet.

Game development is like the wild west of new media (which is already pretty new territory). It’s exciting to be on the frontier of a new medium where anything goes and everyone has a right to stake a claim. On top of that, games are an active form of entertainment, that’s what really sets it apart from more sedentary pastimes.

It seemed like a good way to get some publicity about the game. I didn't really think I would get nominated; just entering the game into the IGF got me a fair amount of attention. For example there is Jonathan Mak's quote about playing the game with umm, toilet paper by his side.

The US$95 entry was a little steep (especially when converted to local currency) but I figured that it was worth it for a game that I've spent more than two years on.

I won’t mention names, but I was casting about for a good space ship game to play and the ones that I found had some pretty nasty flaws. Mostly they use too many menus and sub-menus, basically they struggle with presenting the player a usable interface for controlling their ships. I wanted a game about star ships duking it out and I didn’t want to have to wrestle with nested menus.

Battleships Forever is built on the concept of having as little abstraction as possible. If your ship is supposed to have a turret, it has a turret. That turret is actually represented there. If it is traversing to port, you will see it rotate. If it gets destroyed, you will see it getting blown up. If your ship is meant to be vulnerable from the rear, then there’s a connecting wing strut that takes the whole wing with it when blown off. There are no abstract “armour” values and definitely no “to-hit” chances.

Performance has always been a bugbear throughout the development process. Battleships Forever is built using Game Maker. While I think Game Maker is a fantastic tool for creating games, there’s no denying that it’s slow. Slow like pants. I constantly revise code to optimize the game as much as possible.

Not designing the game architecture with multiplayer in mind. At the time I thought that it would be completely beyond me to do decent netcode. It still is, but if I had planned for it I could at least get someone to help me implement it now. Sadly the whole structure of the objects and the game really doesn’t lend itself to multiplayer implementation now.

Custom ships and the core game are kept separate. The Sandbox is its own game mode and you can't get custom ships in the campaign missions. You can choose to enable custom ships in the skirmish modes of the game, but that invalidates your highscore for that session. The customs ships are more of a way for players to use the game engine to play out their own fantasies. The Ship Maker has shown me just how much people like to create their own ships and it has really driven the popularity of the game up. When I see players posting on forums about Battleships Forever, more often than not they use a screen shot of their own custom ship instead of a screen shot of the normal ships. This is what allows players to identify with the game and become invested in it. They're not just promoting the game, they're promoting their ships, on a very personal level.

I do have plans to do a game where customized ships are integral to game play, but that's for a future project.

Having separate objects for every ship component has forced me to write very efficient code for these objects. I've gone back and revised the code for the ship components themselves countless times. The key was to keep as much shared data with the ship as possible. For example, each ship in the game has its own list of targets sorted by distance so when a turret needs to pick a target, it goes through the list of targets that the ship maintains instead of seeking a target on its own. In this way I minimized the amount of calculations required for each individual component.

Collision checking is hurdle and I fudge a lot of things just to keep things running smoothly. I avoid running precise collision checking as much as possible, which means that most game objects have simple collision rectangles so you get a few odd instances where a projectile hits something that isn't really there. The difficulty lies in deciding when it was acceptable to fudge something and to what degree I could simplify things without making it obvious. That meant lots and lots of play testing to weigh the amount of processing it cost to get a certain effect against the effect it had in game. When in doubt I just went with whichever option was cooler :)

Uhhh, I’m pretty much solo on this. But what really keeps me going is the encouragement from my players. It’s good to know that people are enjoying your game. Other than that I’m a pretty haphazard worker. I might get an idea at midnight and I’ll end up coding till 6am before crashing.

I started the project about June 2005 while waiting to be conscripted into the Singapore Armed Forces. Development wasn’t continuous; obviously there are constraints on your free time when you’re a soldier.

Once I was through with my basic and vocational training I got posted to the Navy and that’s when I had enough free time to get working on Battleships Forever again.

How much time remains? I don’t know. When I first started the project I expected to complete it in 3 months. Boy was I wrong. I’m aiming at completing it sometime early next year though.

Game Maker. I use Excel spreadsheets to aid in play balancing and Photoshop for some graphics.

After making a bunch of custom maps for Warcraft III I began to feel the need for something with which I could make more extensive games but still be simple enough for me to be able to produce something quickly (and working alone). I'd done a basic course on Java by that time so I knew that writing games on a full-size language like that was more than I could handle. I looked at Blitz, DarkBasic and even Torque. I chose Game Maker for two reasons. First of all it had a drag and drop interface which I thought was interesting (I was familiar with the concept because of Mind Rover, does anyone even remember that game?), turns out that I didn't like the drag and drop interface in the end (I write my games in GML script) but that was one of the pull factors. The second reason is that Game Maker is more focused on doing what I want to do; Easy and fast creation of 2D games. I didn't need flashy 3D stuff, I just needed an easy way to make and publish games. Game Maker gave me exactly what I was looking for.

Don’t look down on the little guys that use Game Maker!

It's the combination of tactical challenge and pretty explosions. Blowing stuff up is disturbingly fun; I made sure that you have plenty of things to blow up at all times. I'm sure many players have had a ball simply by spawning ships and immediately blowing them up in the Sandbox. On the other hand the game also offers a deep tactical challenge that will test your RTS skills to its limits. In that way, I think the game appeals to a really wide range of players.

I wanted to. Badly. But yeap, didn't have the cash (still don't!). I did go to the exhibition though.

I'll answer the second part first. I'd like to work on a MMO at some time because there is so much unexplored potential in MMOs. There are gameplay mechanics and situations that would only work in MMOs. To date, the only FPS game that I've played that really made me feel like I was on the front line of a planet-wide war was Planetside. Other FPS games, even titles from the Battlefield or Tribes series, don't quite produce conflicts of the same kind of scale as Planetside. Also, there is beauty in the player-run organizations that are born in MMO games. The kind of corporate intrigue and backstabbing that can arise in EVE Online is absolutely mind boggling.

That said though, MMO games suffer from a lot of issues that need to be addressed. I tried World of Warcraft when it was in the open beta stage. I gained an intense dislike for the game when I was forced to travel between multiple towns simply because the designers saw fit to deny players access to certain services (crafting merchants, bank access) in the outlying towns. In some places you simply can't get what you need and as a result you're forced to trudge to another town that does. Was I enjoying that? No. I hate games that force the player to perform a series of "make-busy" tasks that have no value on their own. It's a very heavy handed way to stretch your content to get maximum play time. I believe that game play should reward a player with access to even more interesting game play and the process of getting that reward should be rewarding in itself.

I believe that a lot of these issues are rooted in the fact that early MMORPGs had to keep things simple in order for their server to handle the load. Technology has moved on from there but game design has not. Too many MMO games out there try to copy established game mechanics without realizing that the game mechanic itself is flawed because it was designed for technology that is a decade old. Games like Tabula Rasa or Planetside really push the limits of interactivity in an MMORPG and it's a real shame that not many other developers have the same kind of vision.

Is Singapore crazy about MMOs? I don't know really. I know some people who play various MMORPGs but I'm not really in touch with the whole scene.

I’m not sure really. The state of the game industry in Singapore leaves a lot to be desired. It’s a strong possibility that I might have to migrate to another country in order to make a living making games. I guess I’ll see what opportunities present themselves at GDC and roll with it.

In terms of what’s my next game, well I have a fairly sized stockpile of game ideas. I’ll probably pull something out of that and make that. Sometime in the future I’ll come back and do a multiplayer sequel for Battleships Forever.