by Richard Garfield

Summary: Perfect game balance is a myth. Game balance is about improving play for a wide range of potential players and often involves costs and benefits.

What Is Game Balance?

The term "game balance" is used in a lot of ways. The way I think about it, these days, is that "balancing a game" is tweaking the way a game is played in order to make it better in some way. This might include adding cards, changing cards, modifying rules, or changing some of the numbers, such as the victory points associated with certain things. One can also address balance by changing the metagame, like how long you have to make a move, how many points constitute "victory", or what the ante is. Really, any change to the game can be "balancing" it as long as you don't pass that fuzzy point where it is no longer the same game.

Balance is often about making things fair. If going first or playing a particular side in an asymmetric game is an advantage, it might be leveled with a balance tweak. It is also used when a particular strategy is thought to be too powerful or weak, or when some part of the game doesn't impact the play enough.

The term "balance" is somewhat unfortunate because it implies there is a single correct place for the final game. In fact, tweaking games in this way is exploring a game space, and while some of the tweaks may lead to objectively worse play, some will make little difference, and some will be better for some players and worse for others.


Spectromancer

My favorite story of game balance comes from a digital game I co-designed: Spectromancer. That game had a small number of character classes that players could purchase, as well as one class that was free: the Cleric.

Naturally players complained that the Cleric was underpowered; we expected that — it is free, so it must be underpowered, right? But we did not choose it to be free because it was the weakest; we chose it because it was a flexible class that showcased a lot of the mechanisms in the game.

Since Spectromancer was a digital game, we could easily collect data and check the claims. We did that, and what we found was quite interesting. The Cleric was underpowered...for beginners and intermediate players. The Cleric win percentage was running about 45% for beginners and 48% for intermediate players. For experts, though, it was one of the choice classes, at about 55%. When you think about it, this result isn't surprising; the Cleric was chosen for flexibility, and flexibility is skill testing.

On the other hand, the Necromancer was widely believed to be overpowered. It had the opposite statistics, though, running about 55% among beginners, 53% for intermediates, and 45% for experts. This is because Necromancer has some great powers, but when you know what you are doing, you can play around them.

Our complete analysis showed that of the six classes, 2-3 were favored for beginners, 2-3 for intermediates, and 2-3 for experts. No class was favored for all levels of play, and no class was a disadvantage for all levels of play.

Balancing for the Best Players

There is a school of thought that preaches balancing games for the top-level player. If we had subscribed to this, we would have had to weaken the Cleric and strengthen the Necromancer.

Would the game have been better then? Maybe for the expert, who then might be able to play the Necromancer without feeling at a disadvantage. However, since many players already felt the Necromancer was overpowered and the Cleric underpowered, for them this would have made that imbalance even more extreme. It would have made the game worse for them.

Advocates of this approach would argue that this is temporary, and players would grow to appreciate the balance for experts. There are two problems with this approach. First, players won't hang around to become experts if they aren't enjoying the game. Second, and I think more importantly, many players didn't evolve to an expert level of play despite playing for a long time. They lacked either the interest or the capability of playing at "top level". In this case, changes would have driven off players who were having a good time with the game by playing the way they wanted to play.

This philosophy of balancing for the best players has lead to some designs that are difficult to get into. Sometimes I feel like the fun has been designed out of them — and this feeling has occasionally been confirmed when I talk to some of the playtesters whose eyes sparkle at how much fun they had developing the game, at finding the broken combinations and exploring crazy game situations. It can make me feel like I missed out because that process sounds more fun than what I find in the final game!

This is sometimes not a philosophical decision, but a consequence of using iterative game design with the same group over time. It is always a good idea to regularly introduce new players or play groups when game testing to make sure you always have fresh eyes and are not discounting their opinion because they don't know the game as well.

This is not to say a game designed this way is always bad; it might be excellent for an expert at the game. In fact, sometimes this approach is entirely appropriate; it depends what your audience is — but it is a mistake to think that it is the single correct approach or that it comes with no cost.


A concrete example of problems from balancing for higher level play comes from Artifact, a digital game I worked on. This game still has a small devoted audience, but it failed to sustain a larger community in part, I believe, to mistakes in balancing. Some of the cards were considered completely broken by the community, but we knew that, for good enough play, they were not. We had years of data that confirmed that — but we would have been better off weakening these cards despite the fact that in some technical sense they were not in a bad place. The cards were balanced for the expert at a significant cost to the beginner.

Many Dimensions of Balance

I have presented one dimension upon which a designer might have to make balance choices: player skill. There are, actually, many to consider. Every different play style is potentially something to think about with regard to balance. For example, many battling games I have played have focused on game play that is good if most players are attacking. This might make a great experience for many groups, but if the game allows defensive play and enough people decide that defensive play is strong — or they simply like defensive play — the game might not work at all. The design breaks down not because it was balanced for inferior play; it breaks down because it wasn't designed for a broad enough range of play styles.

An example of balancing along another dimension that I ran into was with Bunny Kingdom, a drafting game. With Bunny Kingdom, you draft cards to expand your empire and objective cards that contain all sorts of weird ways to additionally score. At first, the objective cards were given points that made them more of a tie breaker than a primary goal; the spotlight was on the empire building. However, some players liked the process of figuring out how to get the most out of objectives more than they liked building an empire. At one point, I did a revision in which the points given to the objective cards were increased a lot, and while there was some grumbling from the players who liked to focus on empire building — and thought the game should be principally about that — overall I felt like the range of players who liked the game was larger, and I embraced this new standard.

Note that with this example, it is not balancing for the expert or the casual player; it is balancing for the different ways that players play the game and finding a way for them to coexist. Regardless of which way a designer went with this issue, a crop of experts would grow around it and play appropriately to what the game ended up being. That doesn't mean that expert play was ignored — as the broad vision of the game settled down, certain cards came up that in expert hands could dominate play and these had their points nerfed — but I tried not to nerf them to the point they were worthwhile only if you were a very good player. Finding a value that neither dominated top level play nor seemed useless to casual players was not always possible, and that is, for me, where the true balancing act of game balance lives.

Another balancing story comes from Robo Rally. The newer versions have enough new stuff going on that they could be considered different games, but the design exercise was one of re-balancing based on changes to where I thought the game should be.


When Robo Rally was first made, the balance was set for my interests at the time, which were centered on the racing and programming. My favorite part of the game was making programs that could tolerate all sorts of disruption. The option cards were, to me, spice rather than a central feature of the game. Many players didn't view it that way and commonly introduced variants that would get more option cards into play. When playing without these rules, players often had to face the decision: "Should I do what is fun or try to win the game?" Of the several changes introduced by newer versions of the design, one of the most important is that all players will get access to option cards over the course of the game, even if they are completely focused on making correct plays.

Balance on Demand

The idea that there is an ideal balance based on the best players misleads not only designers, but also players. Inexplicably, these players often believe they are among the best long before they have played enough to qualify even as a beginner. I have had industry professionals offer balance advice for games designed over the course of years literally five minutes into their first play! I find this shocking — but the disrespect it shows to the designers, developers, and playtesters pales in comparison to the disrespect it shows the whole field of games, where a single game can absorb lifetimes of study and exploration.

An interesting phenomenon of modern game culture is the belief that publishers should adjust balance to the players rather than the players adapting play to the demands of the game. Since each play group will have its own ideal balance that will change over time, this often doesn't even make sense. Following this player pressure can lead to games made for the noisiest players rather than accommodating a broad community. I believe this is at least in part a consequence of digital games in which balance can be — and often is — adjusted on a regular basis to address player concerns.

This is not entirely a bad thing, but overreliance on adjusting game balance means that when players have challenges, they won't try as hard to overcome them — and monumental game achievements will be removed from the table. For example, one season in Magic the card "Necropotence" was dominating play. Many players wanted the card banned, but the team at Wizards of the Coast who manages the banning decided not to ban the card. The world championships had a field full of Necropotence decks, but the championship went to a secret weapon, a Finnish deck (I believe) called "Turbo Stasis". Had Necropotence been banned, this mountain wouldn't have been there to climb. If players can rely on a publisher adjusting their game in this way, why would they invest a lot of energy into solving a game-related problem? They expect the game to change, but their play to remain the same.

It is worth noting that the team that chose not to ban Necropotence did not know about Turbo Stasis. They had respect for the depth and complexities that games can have and wanted to give room for players to surpass them. Players should be given that opportunity. A game in which the designer is the best player in the world is really more like a puzzle. Give players the opportunity to go further than you!

There are two ideas here that appear to work against one another to some degree. I am advocating listening to and considering all the player feedback, and I am also saying you should be stingy with balance fixes on demand. As designers and developers, during the process of making a game we should be accommodating and experiment to find the best place for our games — but once finished, as players we should respect that balance, at least until we have put in some significant time.

Customization

So players shouldn't make balance adjustments if their playgroup doesn't like a game's balance? That is not my belief at all! Whatever balance adjustments players want, they should make — it is their game. The thing that I think needs to be curtailed is players pushing their balance adjustments on others through a central authority. When players wanted Necropotence banned, they were saying, "No one can deal with this, so fix it." But the fact was that THEY couldn't deal with it. (Many playgroups made balance changes of their own by disallowing certain cards, and that is great!)

As long as all players know what is going on, you can't really go wrong with playing around with game balance on your own: at worst you try something new and learn something about the game; at best you find a way to play that works better for your group. Actually, at best you come up with something that works well for lots of groups, and this variant either becomes an alternative way to play or an official way to play. That is what happened with the Commander format of Magic play that was more casual and multiplayer.


I re-balance games all the time. For a long time I played Lost Cities with a scoring that made it riskier to start an expedition. With Carcassonne, I made the end of the game uncertain because I didn't like memorizing key tiles. With Tichu, I simplified the scoring and focused on the part of the game I liked the best: going out. With Magical Athlete, I removed the Assassin. None of these changes were prompted because I thought the designers made a mistake; in fact, it is a sign that I admired the games that I bothered to tweak them to fit my playgroup's ideals at the time.

I have heard concerns from both players and publishers that being more flexible with the official balance will lessen serious play and lessen interest in tournaments, yet many games are played with all sorts of variations that players enjoy, and they maintain and grow serious play. Texas Hold 'em became the dominant way to play Poker, but I don't believe its success was weakened by Poker being played all over the world in countless variations. Even in digital games where it is often difficult to play variants, you find cases like Starcraft in which there is serious competitive play, but some people play their own rules that prevent attacking early to change the balance to their preference.

Game balancing is exploring a region of game space, and there is no perfect balance for all players, so it is a good idea to empower players to find their own balance by tweaking the game in ways that make it better for them. Of course, a player can always play any way they like, but some games make that decision more acceptable by suggesting variations of play or inviting players to experiment with house rules or removing certain cards from the deck if, for example, they don't like "take that!" in their game.

Metagame Balancing

It is sometimes possible to balance a game from the outside rather than the inside. Metagame balancing attempts to leave the game alone and instead change the way in which it is played. In Chess, this might mean playing several games, while alternating who goes first or adding time pressure. In Gin, this might be playing to 50 or 200 points rather than 100. In Go, it is komi, the points granted to the player who goes second to make up for the disadvantage they have.

This approach is often fuzzy. Is playing to 100 points a rule of Gin or a metarule? Also, even though the spirit of the metagame balance is to balance the way the game is played rather than the game itself, metagame balance can change the game radically. Speed chess is very different than Chess with no clock.


One interesting thing about the komi bonus in Go is that it has been going up over the years; it began as 2 points and is at 7.5 now. If players are getting better, this is to be expected as better players can leverage their advantage more — but that suggests that the players who aren't as good are suffering under a higher komi than they should. The mathematician Elwyn Berlekamp suggested that players should bid on the right to go first using komi, which neatly adjusts the balance to the level of the player. Alternatively, some games use the "pie" rule — the second player is allowed to swap with the first after seeing the first move — to encourage the first player to make a move that isn't too good.

Both Berlekamp's idea and the pie rule suggest a meta-meta balance in which the players set elements of balance as a part of the game. This sort of balance was used with Keyforge as one mode of play is to bid chains — a form of disadvantage — for the right to use a particular deck.

Conclusion

It is amazing how many different types of players can engage with games in so many different ways. Ideally designers think about balance that encourages breadth as well as depth. As players, we can take ownership of our play experience and find the balance that best suits us. Strong game communities are cultivated when balance is understood to be something that can be adjusted to allow lots of different types of players rather than seeking the one true balance that defines us all.