GDD #2 – A Gruelling Design Problem: Keeping The Balance In Our Games

A Gruelling Design Problem: Keeping The Balance In Our Games

“Analyzing the balance as a problem.”

By Yigit Doruk

When we think about the critical aspects of a multiplayer game, I think the most common thought would be about “balance”. In singleplayer games, it would not be the “must-have” thing to do, but in a multiplayer game which is played by millions of people, -for example League of Legends- I think it is worth to do. I am not really into developing a multiplayer game nowadays but I think it is an interesting topic to research and I think it can help me as a designer because this pursuit of balance is a harsh design problem. In this article, I will explain what I learned during my research and transfer this knowledge to you.

In a nutshell, you need to make sure that all of the options in your game are fair among all skill levels.

First of all, what is exactly “balance” in our games? In a nutshell, you need to make sure that all of the options in your game are fair among all skill levels. (In general, it applies to multiplayer games, and I will talk more about multiplayer games during my article) 

If you spend any amount of time in any multiplayer game, I think you are familiar with these terms; “overpowered”, “unfair”, “cheap”, “underpowered”, and “useless”. These terms are used when an unbalance occurs in our games. So, we now have a piece of information on what is a balance but do we know how to apply the balance to our games, do we know our options?


Creating a balance is based on the design of the game. If you have a symmetric game where all of the players have the same starting conditions, creating a balance is not a hard task to do. However, in an asymmetric game where players have many different options to face off against each other, balance becomes a tough thing to create. Because developers need to make sure that all of the options are equally viable among players in the same skill level. It is good to remind again; “balance is a difficult thing to pursuit”.

It is good to remind again; “balance is a difficult thing to pursuit”.

What are our options to create a balance in our game? First, there are many different approaches to create balance but before we jump into the options, I want to say that balance is not all about numbers, player psychology is important too. You can create an ultimately balanced game, but your players wouldn’t probably have fun with it. When you want to create a balance you need to consider your players. Their psychology is more important than numbers. 


1- Trade-Offs

Trade-offs are based on advantage-disadvantage relationship. It is canceling out the character’s advantages with using drawbacks. For example, In a racing game (like MarioKart), if you have a really fast character you need to decrease its acceleration rate. But if you have a character with high acceleration, you need to make it slower. This is the core concept of trade-offs. Riot Games calls this method a “power budget”. They consider “advantages” as a “cost” but “disadvantages” as a “discount”.(For example, Vayne has high damage but she has low health) They say that if all characters are sitting in the same spot in power-budget, characters will be closer to being balanced. But how you can apply this method for unparalleled characters like heroes in Overwatch where we couldn’t easily calculate the power-budget for heroes as we do in League of Legends. Or how we can calculate the power-budget if we have tons of stats to consider? Are we going to measure all possibilities for our heroes might encounter and tweak them to create a balance or are we going to get lost in a pile of stats? The answer is no. As I mentioned before, balance is not all about numbers and variables. It’s all about player engagement right? We seek balance to create a more engaging experience for our players, so player psychology should be our main point. You might create a wonderfully balanced game without any leaks, but will it be fun and exciting for your players? If you stuck into mathematics more than your players when you want to establish a balance in your game, you will come up with an ultimately boring game which is smartly balanced…

What should we do then, how can we solve this problem, what are our other options to avoid this problem?

2- Counters

This is where we give our players to chance that they could negate each other’s moves and strategies with their preferred tactics and options. Like in a RTS game if Player A decided to spend his money on quickly rushing and attacking Player B and if Player B decided to go for a more defensive approach and spend all of his money on defensive units, Player A would be “countered” by Player B. The idea is, when you try to create a balance with using this “counter” method, you need to make sure that everything has a counter. Let’s return our example, the defensive Player B would also be “countered” by Player C, who is decided to be more economical and started to create tons of powerful units -which will smash the Player B’s fortress- with his ample amount of money. But if we creating more and more counters for everything, we could find ourselves in an infinite loop. For example, how we can counter Player C? Player C can be countered by player A, who is decided to be more aggressive. Just like in rock, paper, scissors game; everything has a counter, everything is a counter. This concept is really useful because you can ensure that no element is overpowered, it always countered by something else. Also, no choice is meaningless because it at least works as a counter to something else. This method encourages our players to become a more “multi-disciplinary” player, by using “mixed strategies” and it forces players to “switch tactics” during the game in a dynamic way. In class-based games, players are dynamically changing their classes during the game because they need to pick classes that respond to the team’s needing. They need to protect each other from weaknesses. If some elements(class, champion, item, etc.) are completely shut other elements down, they are generally described as “hard counters”. Like in TF2, Pyro is “hard counter” of Spy. Also, we have “soft counters” which is described as if one choice has an advantage over the others. Relationship between McCee and Tracer is a good example for his. McCree has an advantage over Tracer by his “Flashbang”, but his chance of winning is not %100. If we return to our “rock-paper-scissors” example, when you try to balance your game in the same way, It’s important to define what are the “hands” and what are the “throws”. Hands are our decisions before the game starts, such as, selecting our character or pick a race. And these things are locked when the game starts. You can’t change your character in the middle of the game in fighting games, for example. However, “throws” are the decisions can be pick during the match like; the moves, the units, and the strategies. Likewise, in Overwatch the whole team will be assembled before the game starts. You can’t change your players -if they didn’t disconnect from the game- and all of the team members can pick a hero to cover their weaknesses. This means that the whole team is “hand” and each player (hero) is “throw”. The throws are designed to be unbalanced against each other, to encourage counter-play and teamwork. Players need to choose different “throws” to defeat their opponents. On the other “hand”, “hands” are supposed to be balanced, so every hand should have access to all of the throws. For example, in any fighting game; If any character couldn’t Dodge, that character would be useless.

Now we understand what are our options to balance the game, like implementing trade-offs and counters, how can we make sure that our game is perfectly balanced? How can we measure our outcomes?


This is where collecting data comes very handy. Either you collect your data from millions of players or play-testers for your game; you need to collect this data from players. What data exactly? You may ask. First of all, we can gather “win rate” data for characters or races in our game. This is saying how often that character(or race, etc.) results in success. With this information, we can say if one character has a %50 win rate, it is perfectly balanced. Is this exactly true? If a character has %50 of win-rate, is it balanced? Can this information may lead us to wrong inference? Let’s think about the game with 3 characters. If Character A loses every match against Character B but wins every match against Character C, the “win-rate” for our Character A will be %50. So, let me ask again, is Character A balanced? When we are gathering data, we must be really careful. We can analyze each character match-ups against others by using “match-up charts”. But even this detailed charts couldn’t show us everything. We should consider the skill-level. Like I mentioned before in this article, our game elements should be equal and balanced among all skill levels. For example, a character may show it’s potential only by highly skilled players. In this case, these professional players can destroy every other enemy with this character. Whereas the low skilled players get killed again and again. Because of that, professional players are increasing the win-rate of that character but low skilled players are decreasing it and we came up with near %50 win-rate for that character. This rate is also going to mislead us because it is unbalanced among other different skill levels. So we need to make sure that we analyze the character’s (or other elements) “win-rate” and “match-ups” across all skill levels. But more than that, these numbers are not exactly telling us what’s going on in the game. As designers, we need to know which characters are people picking in our game.  We need to check that the character is well balanced and also fun to play. In addition to that, the character should not only useful in certain situations. For example, Symetra was only useful in certain situations before she got patched. The Overwatch team were aware of this problem and gave her 2 different “ultimate abilities” to pick from during the match. 

This where comes the player feedback is really important. The “pick-rate” of a character tells us how often that character is getting used. We might want to merge “win-rate” and “pick-rate” values of that character and analyze it to gain more information for that character and we can balance our character accordingly.

I think you might familiar with the term “meta” and “meta” is related to “pick-rate”. When the player community found the most effective character, card, strategy, etc. the “pick-rate” of that element increases dramatically and that “element” become a new “meta”. “Meta” can be created via forums, videos, and e-sport matches. If a professional team is wiping out their opponents with some kind of strategy or character combination, that strategy becomes massively popular among other players too. This “meta” is a wonderful opportunity for us. Meta can act as a self-balancing force. It forces players to switch tactics and their current way of play. For example, if one character dominates the meta and players are bored of losing against that character, they started to experiment. They try to pick different characters, change their strategies to kick that character’s ass. If they found some kind of “counter” for that character, meta might change again. The “meta” can keep our game fresh and can give the players who discovered that counter a real satisfaction. But we couldn’t rely on “meta” to keep our game balanced and fresh. While our meta is helping us balance the game, sometimes designers have to check for if any element is overpowered or underpowered and change them accordingly. If they spot any unbalance in a game element, it is time for a change. We need to find out what is “exactly” causing this unbalance in our unbalanced element. After spotting the real problem, we need to figure out what to do. Do we need to nerf? Or do we need to buff? Nerfing means making something less powerful, like reducing the damage of the character. However, buffing means the opposite, like increasing movement speed of the character.

This “meta” is a wonderful opportunity for us. Meta can act as a self-balancing force. It forces players to switch tactics and their current way of play.

You may think that nerfing the overpowered elements is the best solution, but it is not true. We can leave the overpowered element alone, but we can buff the elements that counter that element. There are lots of reasons why you should choose buffing over nerfing but if you want a basic explanation here is why: “Players don’t like the nerfs.” This is all simple, if you nerf your game elements frequently, players will think that nerfed element is useless and boring. But if you encourage your players by buffing the other elements that may counter the overpowering element, now players believe that they can have a chance to kick the unstoppable character’s ass. By executing this method, you can maintain balance without upsetting your players with continual nerfs. But this is not always the true solution, some cases you need to use the nerfs to keeping the balance or sometimes you need to get rid of that element entirely if it becomes unbearable and messes with your balance entirely.

So, we detected the unbalanced element and tweaked it with using “nerfing” or “buffing”, how could our players know this change that we made in our game. How could they know what is changed or what is still the same? How could they come up with new strategies to adapt themselves to new changes? Also, if players might comfortable with any character, if we change that character’s stats (or even how they supposed to work entirely) suddenly, they are going be shocked. Because of these reasons, we need to inform our players before we make those changes. These regular changes for the games are called “patches”. So every time we “patch” our game, we need to make sure players will get information about that patch. We could use “patch notes”, “videos” or “blog posts” to communicate with our players and explain to them what changes are applied to the game with this patch. These patch notes are so important that players could alter their strategies with only reading the patch notes and without trying to figure out if that change is applied to the game our not. Also like I mentioned before, we need to make sure that our game is balanced among all skill levels. Most multiplayer games rely on skill(or level, basically experience) based matchmaking systems. But how we could do that on more accessible, party-style games? We might want to add some supporting factors for inexperienced players to give them the chance to stand against experienced players. Like death-streak mechanic in Call of Duty MW’ s multiplayer mode where the players get extra bonuses if they die again and again. Some kart-racing games are using the same helping hand too. The last player in the track gets bonus abilities to knock down other players to be able to get into the race again. We can also implement other mechanics like giving the game some kind of luck factor to keep the balance among all skill levels. Like in battle-royale games, the starting items that you find are based on luck. An inexperienced player might find good loot to keep them on the game and have a chance to kill other players, whereas the experienced player might find poor loot but still have the chance to be #1 with his abilities in the game.


Keeping the balance in our games is a harsh task to do and also really interesting and arduous design problem to solve. We could implement some balance options to our games to solve this problem like applying “trade-offs” to our game elements. Also, we could make these elements to “counter” each other with using “rock-paper-scissors” foundation. But keep in mind, even if you think that you balanced your game very well, you need to test these options in action. You could analyze the “pick-rates” and “win-rates” of these elements to decide what elements are considered as “over-powered” and what is considered as “useless”. These rates might also show us what is current “meta” of our game. We could know more about our game’s “meta” by reading player community blogs, youtube videos, and other social media posts.  In addition to that, we need to use “nerfs” and “buffs” to change the values of our game elements, but keep in mind that “buffs” are better than “nerfs” in most cases. Player psychology is a really important thing to talk about too. You need to inform your players about your current or upcoming “patches” (Updates for our game which is full of new nerfs and buffs) via Youtube videos and blog posts. In the end, make sure that your alterations to your game to create balance are equally viable among all skill levels. Also, don’t forget that each time you add new features or new game elements, your balance might be screwed over again…


                              See you in my next post, take care 😉

Spread this blog post to the world 🙂

Close Menu