The only time you can properly balance a game is when most of the game is done. Balancing your game ahead of time, before all of the gameplay is working and all the levels, if any, are made, can only be considered to be preliminary balancing. Preliminary balancing can be very valuable , however, and you will want to do it just to get your game fun as early as possible, with the full knowledge that you will need to adjust it later. That said, you cannot truly get a sense for how the entire game needs to function and how the difficulty must escalate over the course of the entire game until the game s content is complete. You can view your game as a collection of different systems that make up one large system. For a level-based game, each level can be considered to be a system in itself. Then, within each level, each combat encounter or puzzle can be considered to be a system. In order for the game to be balanced, all of these systems must be in place, since changing one system impacts how the other systems must be set up in order to achieve the overall balance you are seeking. At the same time, if you wait for all the systems to be final and some are running behind, you may well run out of time to get the game as balanced as you want. For example, on The Suffering , too many aspects of the levels were not completed until quite late in development, and as a result balancing was put off until it was too late to do it to the quality level we would have liked . Starting balancing a little earlier, before every last element of the game is in place, may be necessary to make sure you do not run out of time.
The time at which the game is largely complete and true balancing becomes possible-usually coincides with the time when the game is in full-on testing. This works out for the best, since balancing and testing are closely intertwined activities. Balancing often involves changing some settings in the game and then playing it to see if those changes create the amount of challenge you are interested in. For each pass on the balancing, both you and the playtesters should try to play the game. Then the testers can give you feedback about just how effective your efforts to balance the game have been and, combined with your own analysis of the game s condition, you can make more changes and iterate through the process again. People who can successfully balance a game by themselves , without the input of other playtesters , are rare. Often designers who attempt to balance a game by themselves succeed in balancing the game only for themselves, usually resulting in the game being too hard.
The best way to balance the game is to break down different systems into groups of numbers that can be easily adjusted and tweaked. For instance, suppose you were making a melee combat action game of some sort . If the players use a baseball bat in the game, that bat will have a number of different attributes associated with it, such as how much damage it does, how fast it attacks, how many times it can be used before it breaks, how much it costs to buy, how many hands are required to hold it, and so forth. Similarly, one can also break down enemy, player, and other system attributes into collections of numbers that can then be adjusted to vary the usefulness or challenge of that object. It is these values that you will continually adjust and massage in order to achieve the balance you are seeking. This of course assumes that at a basic level your game is already fun, with a sufficient variety of challenging encounters. If you wait until too late to get the game fun at a basic level, it may become too risky to make the changes necessary.
As you are balancing, you must be keenly aware of how the different values you change affect each other. You may change one weapon in order to make one combat situation a lot of fun but end up making another location in the game actually unbeatable. The more complex your game, the more impact the changes you make may have on systems you might overlook. As you are balancing you must fully consider every part of the game that your changes are affecting and make sure you do not break the game. The only way to be truly sure you have not thrown off the entire game is by testing it thoroughly. As a result, making significant changes close to your ship date is a nerve -racking experience. What if the changes you make break something that no one catches before the game is sent to the duplicator?
Of course, the method for balancing I have described above necessitates that the data that affects the behavior of the game s different entities be accessible and modifiable by the designer. This means that the code needs to be written in such a way that makes changing this information easy. This last point may seem obvious, but I have seen many engines in which changing information such as weapon statistics was far from easy to outright impossible. From the very beginning of the game s development, the programmers must keep in mind how the designers will go about balancing the game at the end of the project. If, instead, they bury a collection of magic numbers in the code, the game will become locked in a particular state, making balancing it impossible. Though balancing can only take place once the game is largely complete, the programming team must start preparing for that balancing from the very beginning of the project or effective balancing will be impossible . If the designer is to have any chance of balancing the game well, this balancing information must be broken out of the code through configuration files, level editing tools, or other designer-accessible formats.
When I wrote the first edition of this book, I included the following paragraph:
While balancing your game you should keep one rule of thumb in mind at all times: your game is too hard. Regardless of the type of game you are making or how talented your development team may be, by the time your game nears completion and enters testing it will be too hard.
Having just completed The Suffering , however, I can now say that this is not always the case. Following the your game is too hard rule of thumb, we managed to make The Suffering too easy, at least according to the press and many of the fans. This was in part because we did not allocate enough time to balancing the game, as I mentioned earlier. Nevertheless, I still think the above paragraph holds a lot of truth, since most of the time games are too hard when they enter the balancing phase.
Games usually turn out too hard because, up to the point at which testing starts, only the development team has been playing the game consistently. The development team has been working on the project anywhere from nine to eighteen months and during that time they have honed their gameplaying skills and have become quite good at the game, probably better than 90 percent of the players who will ever play the game. In order to keep the gameplay interesting for themselves, the development team has made the game somewhat challenging for themselves to play, which in turn means it will be too hard for 90 percent of the players out there.
The first comment testers will often make is, This game is too hard. As I discussed above, your first reaction will be to ignore this complaint, to chalk it up to their incompetence or inexperience with the game. They ll get better, you may say. And, unfortunately , that is true. If the game spends three months in testing, the testers will be just as good at the game as the rest of the development team. Then they too will probably stop thinking that the game is hard. It is entirely likely that the game will ship with the development team, including the testers, having no clue just how difficult it is.
As a designer you must be very careful to maintain an honest sense of how hard your game is, and during the balancing phase you must concentrate on making the game something that first-time players will have a reasonable chance of succeeding at when they first start playing. Always remember what the first impression of the testers was, and ask yourself if you have addressed the problems they immediately identified. If necessary, you should bring in new first- impression testers to see if the game is still too difficult.
Unfortunately, sometimes you may not always be able to make your game easier through balancing alone. You may have created a game design that, on a fundamental level, is hard to play. If you truly want your game to be something first-time players have an easy time getting into, you need to concentrate on this from the very beginning of your game design. My project Centipede 3D is a good example of how a game can become far more difficult than the development team ever anticipated. Attempts were made to balance the game to make it easier, but the gameplay was intrinsically designed to match that of the original arcade game. As a direct result, Centipede 3D did everything it could to make the players game short and fast paced. Unfortunately, players of home games want their games to last a little longer than what they get for 25 cents at the arcade. As hard as the game was in its shipping version, it is chilling to think that before it went into the balancing phase the game was easily ten times as hard.
When designer Jason Jones was balancing the Marathon games, he had an interesting-technique for making sure the game was not too hard. If he and other members of the development team could play through the entire game on its hardest setting using only the game s fist weapon, he figured that the game would be reasonably challenging for other players. Of course, other players get weapons far more powerful and easy to use than the fist, and they do not have to play it on the hardest difficulty setting. Jones handicapped himself in order to see how hard the game would be to normal players. Using techniques like this is wise. If the designer can win the game with both arms tied behind his back, other players will probably have a fair chance of playing it through with both arms at their disposal.
In the end, balancing your game is often more of a gut feeling than anything else. Some developers, such as Microsoft, have tried to develop more objective tests to determine how well balanced a game is, accumulating player metrics about the average time players spent in a level, number of times they failed a challenge before passing it, how much ammo they used up, and so forth. Though this data can be quite valuable, it can never fully replace design instinct. Though you may almost always assume that your game is too hard, there are few other rules you can follow to balance your game. You need to be able to see your game holistically, to understand how players who have much less experience with the title than you will play it, and to realize what will challenge them without being unfair or even cruel . Knowing how to balance a game is a skill that comes with experience, both from playing other games and from designing your own. In order to become truly skilled at balancing, you must do both as much as possible.