As I learn about games and game design, one of my goals is to eventually create a game that people play against each other at a competitive level. One of the important aspects of a game like this, whether it is turn based or real-time, whether 1 vs 1 or team vs team, is balance.
Originally I was going to read several articles on this and summarize what I learned here, but this article will mostly cover what I learned from an excellent series written by David Sirlin on his blog at sirlin.net. I highly recommend heading over there and reading the original series once you are done here.
What is Balance
What most people think of when they think of the word balanced is usually fairness. That everybody at the beginning of the game has a roughly equal chance to win (not counting personal skill). And in most player vs player games this is an important aspect of balance. But as game designers, when we say a game is balanced we also mean that there is an large number of viable options available at high level play by expert players. And in the talk given by Ernest W. Adams that I summarized previously, he would say that a single player game is balanced if it is appropriately difficult or challenging.
Your game can have tons of options, but for balance, we only care about meaningful options. We want to avoid a game where only 1 or 2 dominant strategies emerge or where only a few characters, decks or builds (or whatever your game uses) are ultimately playable at a tournament level.
Symmetry vs Asymmetry
Games where opponents start pretty much identical lean toward the symmetrical side of the scale and are typically easier to balance. This is the simplest way to balance a player vs player type game. Examples of games that are on the symmetrical side of the scale include most FPS games and Chess.
The more diversity in starting conditions, the more asymmetrical a game becomes. Adams mentions that this makes games much harder to balance but at the same time makes them more interesting. Examples of this type of game include League of Legends, Magic the Gathering, and any other type of game where multiple classes, decks or characters are available as the starting option. In asymmetrical games, fairness is important to design. You want every option to have a reasonable chance to win in the hands of the right player.
Enough Viable Options
One of the harder questions when balancing a game is how to know when you have enough viable options. The first step in this is ensuring that moves and strategies have counters. If a move cannot be countered then it becomes 1 of those dominant strategies we talked about earlier and the game will quickly degenerate into people just using that move to win (not very fun). After a couple layers of countering you probably want to loop back to the original move so that the third or fourth layer of counter is just the original move, sort of like rock-paper-scissors.
One of the things to watch out for and avoid is chaff. Needless mechanics, characters and choices that do not add to the game and are ultimately useless. Don’t just add an option to add it, make sure it is useful. Explore the design space and see what works.
Checkmate and Lame Duck
When one player is doing better than another and they have reached a point in the game where it is impossible (barring a major mistake from the winning player) for the player that is losing to make a comeback, we call that a Checkmate scenario. And this is fine usually.
However, we want to avoid allowing the time from the beginning of a checkmate until the end of the game to be drawn out. This is a Lame Duck situation where one player or team know they can’t win but are forced to play out the game for a long period of time in the losing situation.
Playtesting Is Important But Hard
One of the best ways to find out if a game is balanced or not is to play it and test it. This can be hard. Especially if you have a large number of options. But play testing discovers things you never could have predicted.
When play testing you want to evaluate your playable characters, decks, etc. One trick to balancing games is to not rank your playable units linearly. Sirlin recommends using a tiered list with 3 actual tiers and 2 tiers that you want to keep empty. You want to keep things out of the “God” tier that must absolutely be played if you want to win and conversely you want to keep things out of the “Garbage” tier that must never be played if you want to win. The remaining 3 tiers should all be pretty close.
Getting language type feedback from play testers can also be hard because a lot of good players make decisions subconsciously. In order to balance your game you sometimes need to rely on the intuition of yourself and others. Be careful to get enough feedback that your decisions are not driven by just one loud incompetent person.
Most Important Design Tip
Probably the most important design tip that I got from this article was to try to start the design with mechanics and failsafes that self-balance. Sirlin used a fighting game called Guilty Gear as his example. It built in certain things like progressive gravity and combo breaking type abilities to keep characters from being able to completely immobilize and beat an opponent with no response.
They eliminate potential problems from individual characters with global design features. This is difficult but if you can figure out a way to do it, do it.
Various Tips and Tricks
My notes from the article have a few things that I though were important that won’t get whole paragraphs here. For more on these I recommend reading the original 4 article series on Sirlin’s Blog.
- Don’t balance the fun out of things.
- Counter matches (characters that counter other specific characters) are not good for 1 v 1 style fighting games.
- Avoid special cases and actually balance the game.
- Don’t set yourself up for failure by including too many characters or too much customization.
- If you add a new move, make it too powerful then scale it down as necessary (otherwise play testers won’t use it)
Conclusion
Balancing games is important and complicated. Perfect balance is hard to achieve because to keep the game interesting we don’t want it to be completely solvable. The more play testing you can do the better because it helps you figure out where the game is unbalanced. One of the best ways to balance a game is to have some global design elements that help ensure balance. While this is difficult, it will help tremendously if you can do it.
Sirlin offers a handy pdf summary of his article available here. This subject will be revisited in the future.
Go balance games.