The Constant Quest for Equilibrium

Balancing an MMO, and World of Warplanes in particular, is a constant process that aims to create equal conditions for all players, while maintaining variety and attractiveness of the overall gameplay. Today we want to begin a new series, looking at how we see balance in our game, how it changed in the past and how it will evolve in the future. The first part of this series, written by Battle Balance team lead Anton Muskeev, illustrates how many factors the balancing team takes into account when trying to fine-tune the planes, shooting mechanisms, physics and many other aspects that make up our game.

Let’s first establish what balance is to us. Here in the game balancing department, we are responsible for adjusting the values of all aircraft as individual subjects, including all of their possible configurations, for all types of players in all our supported regions. Each configuration of each aircraft consists of multiple combinations of modules, weapons, equipment, ammunition and crew skills. When looking at all these factors, you receive a somewhat oblique average stat value that is difficult to decipher. In these cases we turn to more obvious stats that usually paint a clearer picture. However, it’s not always what you’d think. For balancing for example, it’s not necessarily the win ratio – the stat that most people think about when aiming to determine the overall efficiency of a plane. When you look at it closely, you realise that the win ratio is largely dependent on the team setup that you happen to be playing in: the types of participating planes, their tiers and their classes have a very strong influence on the outcome of your battle. At the same time, these factors can vary strongly dependent on the time of day, the number of people playing, if there have been new planes introduced into the game recently, how the matchmaker works and so on and so forth. If we had the ability to pair up each player in teams where for each participant there were an absolutely equal counterpart, flying the same plane, on the same configuration with the same crew skills, then, regardless of the flight characteristic of the plane or the player’s personal skill the win ratio would always hover around 50%. In fact, it would be a bit lower, because you have to account for some of the battles that end in a draw.

That’s why the win ratio is not a perfect factor to look at when trying to determine the game balance, which is why we need to look elsewhere.

Let’s look at what balancing of a plane is. Each plane is subject to balancing. Let’s take a simple example in which every player has access to only one configuration for each plane. For arguing’s sake, let’s pretend that everybody is bound to only use Premium planes without the ability to strap on additional equipment, special ammunition or use sophisticated crew skills. Everybody starts out with the same conditions. Here’s where already a lot of factors come into play that disturb the precious equilibrium. All the planes are matched into separate teams, consisting of players with different skill levels. Each team battles differently. Every battle involved planes of different tiers. After each battle, the players receive rewards based on their performance. Using these rewards, players can unlock new planes and receive access to new parts of the game. We have many planes in the game that are split into tiers, branches and nations. And all of this needs to be calibrated in a way, so that nobody is at an inherent disadvantage. That’s the job of the game balancing department.

Let’s start at the highest level.

Ideally, players should be able to research and purchase newly unlocked aircraft from different branches and nations at the same pace. The player is supposed to be able to comfortably reach new branches and pursue them to the very top, staying invested in the process all throughout. To achieve this, the player needs to receive a balanced amount of XP and credits for participation in battle. We need to make sure that our economy model applies as planned. That means: the higher the tier, the more XP is required.

Consequently this means, that the economic balance and researching balance depends on the speed at which the player receives XP or credits.

Another benchmark for balancing is how well the necessary population across tiers is met in-game. In other words: If there are enough players in the game available for the matchmaker to be able to assemble battles for each level. Most MMOs have the majority of players playing at the lower tiers, while fewer, albeit enough for proper matchmaking, play on high tiers. In the case of World of Warplanes, this is, unfortunately, not entirely the case. However, it is also important to note that players won’t climb to the higher tiers, if they don’t see a reason for it. That’s why in MMOs progression unlocks new features and options for those, who choose to play at the summit of what’s possible. In World of Tanks for example, players bear with the unforgiving Tier X-economy because, in return, they’ll be rewarded with the possibility to play on the Global Map and Clan Wars.

Ok, let’s assume that we have enough players on all levels. This is where the matchmaker comes in. It is tasked with assembling balanced teams consisting of equal numbers of planes of different tiers and classes, based on the overall level of the battle. It is also important that the matchmaker assembles the classes according to the proportions that we had in mind when designing the battles, in order not to have too many one-sided teams but rather a controlled level of variety. That means we’re looking at the balance of tiers and classes when dissecting what planes ultimately end up in a battle. If there are specific player groups present, as is made possible by the Flight feature, the matchmaker needs to account for that balance too, trying to keep the same number of flights per team. That means that flight balance is also a factor that needs to be considered when looking at the overall balancing picture. Additionally, it’s important to know that, on average, different types of players appear in the player pool at different times of day. Battles played against regular players during lunch time are significantly different to battles that are made up of hardened and Flight-connected ace-pilots, who stalk their prey during hour-long night hunting sessions.

Let’s continue. The balancing of teams happens based on three main criteria: tiers, flights and classes. However, when assembling the teams, it’s not, as one might think, all about only planes but also about the map and the chosen game mode. The spawning points, elevation of the map, positioning of clouds, objects and AA guns – all of this also influences the main goal of each battle mode: winning. Depending on the rules of the game mode and its inherent rules (destroying opposing planes, capture positions, availability of re-spawns, repair or power-ups), the path to victory can be quite different. That’s where the balancing of the game mode and the arena or – more simply put – the map comes in. The main objective of this kind of balancing is to ensure equal chances to achieve the game mode’s goals regardless of the setup of the remaining team. These goals can be the collection of superiority points, destruction of enemy planes, ground objects or, ultimately, the victory and the acquisition of the appropriate amount of game resources (credits, xp, etc.). Balancing also needs to take into account that all classes need to be equally useful and important on each of the randomly-selected battle arenas.

Now it’s time to look at the planes themselves. Don’t forget that, for the purposes of understanding, we look at each plane with only one, primary configuration. If the matchmaker doesn’t use completely ‘mirrored’ teams, the players flying some of the aircraft of a particular class can have an advantage over the players of the other team due to the superior flight characteristics of their specific aircraft. To us, that’s admissible as long as all the planes remain equally useful to their respective teams (based on the role they play) and as long as the planes still meet the expectations of the players piloting them. The next point is very important too: There’s absolutely no need for a plane that nobody flies, just like there’s no use in a plane that everybody flies. This part of balancing is something that we have little control over. It’s usually the players themselves who determine the usage quantities of each plane. By choosing your plane, you vote for what’s popular. At this point it is our job to give the players the freedom to choose from a wide variety of aircraft, while at the same time avoiding that everybody picks the same. The popularity of a plane is also a benchmark for how well it is balanced. Usually players pick their planes based on what plane offers the best compromise between its weaknesses and strengths, their personal preferences in terms of the nation of the plane, and what other planes follow an aircraft higher up the tech tree that the player is trying to get to.

Going even deeper into the subject, we of course have to acknowledge that aircraft can have multiple module configurations from stock to top, the ability to install additional equipment pieces, ammunition and even the capacity to train the pilots that fly them. And, lastly, you also have to take into account the player himself, who can be more or less skilled at the game, have different connection speeds to the server and be in different physical states (one being fit and ready and some other playing tired and over-exhausted). Apart from you, there are 29 other of such players in each battle that the matchmaker assembles. All of these factors lead to a very varied situation at the beginning of each battle, which only continues to diversify as the battle progresses. We think that that’s good – because change and variety is inherent to all life and also to games.

So with all this in mind, how do we manage to balance planes in the end? We look at a multitude of factors and draw our conclusions from that. As such we observe different configurations of planes, groups of players with different skills, and assess from that the popularity of a plane, how much damage it’s capable to dish out, how many planes or ground targets it can destroy, how many superiority points, XP and credits players earn on it, and many other metrics. Additionally, we also take into account our community’s subjective opinion on planes, which shows through their feedback from the forum or other media, super tests, special balancing tests or the personal feedback of our game balancing specialists during their obligatory sorties on each aircraft that needs adjusting or qualifying. We almost never rely on average values of one key metric to assess a plane and always try to look at multiple key metrics together to determine a warbird’s final efficiency.

To sum all of the above up, it’s clear that balance is something very complex that is not very tangible and only reveals itself under very careful examination. Pretty much every feature of a game impacts the final balance of it. In the case of video games, where the subjective experience sometimes even overweighs the underlying statistical and technical reality, even the visuals and audio of a plane can influence, how a player perceives the efficiency of a certain aircraft or its armament. In World of Warplanes, the game balancing department acts as a last line of defence, capable of preventing (or affirming) changes that could potentially change the eco-system of the current aircraft power structure in-game. In most cases, we’re able to do just that. Sometimes, we miss something and need to make corrections. We’ve made many adjustments since release already, but there’s still a lot of work to be done. Ultimately, it’s not only the perfect balance that we’re after. We aim to make the game fun, comfortable and as varied a gaming experience as possible.

In the future posts of the series we will answer the most popular questions that our players asked us via the forums and other media. Until then – see you in the skies, pilots!