Ian Schreiber on Game Balance
How do, mortals?
This appears to be the final entry in my notes before I hit the stuff I've already blogged about.
After this, I probably still have a few more things to post on, so I'll get any of those out of the way, but after that, I think we're done.
Balancing, according to Schreiber, is best left until after you have a good set of core mechanics.
Balancing means different things with different types of games;
- Singleplayer - to make sure the challenge level is appropriate.
- Asymmetrical mutliplayer - to find out if one starting position is easier than the other.
- Strategy - to find out whether one strategy is more powerful than another.
- Game objects - to find out if different objects have the same cost/benefit ratio.
When balancing a singleplayer game, it's important to note that the audience's skill level will increase as they play the game. The way the level of challenge increases over the course of the game is called pacing or skill-ramping. Finding the level of challenge appropriate for the audience's level of skill can only really be done in one way: play-testing.
One problem about the audience's skill level is that, when plotted on a graph, the number of people in relation to the level of skill will form a bell curve; the highest number of people with have a middling skill level, and a lower number will be more/less skilled at the game, meaning that you can't reach all of them with one difficulty ramp. The best thing you can do when balancing is to try and hit as many members of your audience as possible.
One alternative measure is to incorporate support for those players who are more or less skilled than your target audience, through methods such as multiple difficulty levels, handicaps, and alternative rule sets.
Asymmetrical multiplayer games involve the two (or more) players or teams of players having different starting positions and/or resources. Balancing this means making sure that, while the teams are different, neither has an unfair advantage over the other.
Truly symmetrical multiplayer games are rare; even Chess, in which both players have more or less similar starting positions and resources, is asymmetrical to a point because one player gets to make the first move. Of course, that doesn't mean that Chess is imbalanced, it's just not perfectly symmetrical.
The general rule is that the more asymmetrical a game is, the more playtesting is required in order to ensure it's truly balanced. A good way to check for balance is to quantify the resources into a comparable system, such as numbers. This isn't always possible, however, and the harder direct comparisons are, the more playtesting is required.
Strategy games, or games in which different strategies are used against another player or the game itself to achieve victory, have a gaping potential problem: a dominant strategy. Once a single strategy is found to dominate all other strategies, that strategy will be used by all players that know it, and the game will cease to be fun or interesting. This is what happens with Tic-Tac-Toe (also known as Noughts and Crosses); once players know of the winning strategy, the player who goes first automatically wins the game, provided they know the strategy. If a dominant strategy is found, the best thing to do (to the best of my personal knowledge) is to remove it from the game.
If a strategy game has several potential winning strategies, however, it is best for these to be well-balanced to make the game more interesting. Te ensure this (and to find any dominant strategies), the game will require a great deal of playtesting. Note down any strategies that seem to be used more than others, and which strategies seem to win. Remember that players may well choose the easiest or simplest strategy rather than the most optimal.
Game objects can be anything that have an effect in a game; cards in a trading card game, units in RTS and war games, weapons, armour, and so forth in RPGs. Try to ensure that any and all game objects have a similar cost/benefit ratio. When balancing a game with game objects, you want to prevent two things:
- any game object from being so powerful that it becomes a dominant strategy.
- any game object from being so weak that it becomes a false choice.
According to Schreiber, there are three ways to balance game objects: transitive, intransitive, and fruity.
Transitive relationship: find the cost curve of the game, and directly alter the item's cost or benefit to match the correlation of the other objects.
Intransitive relationship: similarly to rock-paper-scissors, have one object defeat another, and be defeated in turn by a third object, which will in turn be defeated by another (potentially the first object).
Fruity: when the game objects simply can't be related and compared as in a transitive relationship, being just too different, the only way to gain balance is through excessive play-testing, tweaking, and reiteration of the object.
Next, Schreiber tells us three general game balancing techniques:
- Mathematics (which is difficult and can be wrong)
- Instincts as a game designer (which can be inaccurate)
- Playtesting (which relies on the quality of the testers)
And finally, he gives us a few tips and techniques.
- Be aware of the different objects and systems in your game and their relationships. (What is the aesthetic core? Look at the interconnections.)
- Make one change at a time.
- Learn to love Excel, as this has a huge number of benefits, such as keeping/organising lists, task management, and collecting and manipulating statistics.
- Use the rule of 2 when altering game statistics; if something is too small, double it; if something is too large, half it.
- Balance the first turn advantage; rotate who goes first, give the disadvantaged players some kind of support, or reduce the effectiveness of early turns.
- Write down your own game design rules as you learn them.
This reading overall made another great deal of sense (how about that, all of these wonderful, informative pieces of reading filled with great advice!). Game balancing isn't something I've put a whole lot of thought into, but this has reminded me that it's very important to perform to ensure the game is both fun and... well, balanced.
Well, that's about that, mortals. It's been a good day for this blog, I must have written and uploaded at least three or four new posts.
I apologise if the previous one and this one have some bad grammar or spelling mistakes; the spellchecker seems to be broken. I'll see if I can fix this once this is posted.
[Edit - this has now been fixed. Still don't know what the problem was.]
Praise the Emperor!
