Balance Q&A: Part 1

Our players have a lot of questions concerning various aspects on the game, and naturally balance is one of the hottest topics.

1. Balance and inequality. Players define these two terms differently in-game. What is for you balance in an asymmetrical game with a multitude of different planes?

I have written about the intricacies of balance in World of Warplanes in the first part of our Development Blog series on balance. In short, I think one can say that balance is a reasonably-stable state of the ecosystem of the game, whose metrics correspond to our pre-defined mathematical model: players receive for their battles an expected amount of XP and credits, the battles themselves are being played out with the right variety of expected planes, all maps are being played on equally often without any one being selected more than the others and so on and so forth. At the same time, balance is achieved when there’s also enough variety of gameplay situations. That means that the player has meaningful choices that he can make: what branch of aircraft to research, what modules to unlock and what equipment or ammunition to use, etc.

2. Do you think that an unbalanced and asymmetrical game process makes the game more interesting overall?

Would you enjoy a game where you’d have to play against an exact copy of yourself, which would imitate your every move the same way you do? It’s specifically the asymmetry, variation, choice of possibilities, enemies and situations that require you to constantly adapt, change and improve your playstyle that stimulates players to enter battle over and over again in order to strive to get better.

A good game must give a player a wide variety of choices, tactics and opponents, keeping you on the edge of your seat and giving you opportunities to grow and improve as a player. In World of Warplanes, the biggest choice is that every player can decide for themselves how they want to contribute to their team’s victory: take out ground targets or go hunt planes in the sky. There’s also a choice between a broad selection of warbirds, each with their own distinct character: some like tight dogfights, some like to “Boom & Zoom” in vertical combat, some like the distinguished approach of Attack Aircraft where you need to choose your targets upfront and also think about what to do next once you passed your destination. A wide variety of different planes ensures that the player is confronted with many different gameplay experiences, which is always good for a game. If you’re looking for an adrenaline rush, you can, for example, take a seat in a fast fighter, which is bound to leave you with your heart racing after a heated battle. However, if you’re more looking to cruise, you can always play an Attack Aircraft, which will let you approach things slowly and methodically, taking out one ground target at a time.

An asymmetrical team setup only means that each side needs to play to react to the given situation, playing to the strengths of the planes that they’ve gotten matched up with, in order to seize the opportunity and take home the win. There is no one particular pattern, the following of which would lead you always to victory. Every battle needs to be new and fresh and not just a copy of your previous encounter. In our case, it is our goal to enable the biggest possible range of gameplay experience for the player, while allowing him to always have a meaningful impact on what’s happening on screen.

3. What methods do you use to identify flaws in the game balance?

Our main tools when adjusting the balance are the use of stats, mathematical models and special tests – both automated ones and live ones with players (e.g. the Public Test or Super Test). More specifically, we look at the “plane” and the “player audience”.

Probably our most important tool, however, is always the analysis of gathered stats from the game. It takes into account all the actions that players perform both in-battle as well as in the hangar. The stats tell us about player behavior, their view on efficiency and overall gameplay experience of planes and many other game-related features.

First of all, we look at the efficiency metrics of all planes. As such, we assess the “difficulty level of planes”: How successful are both newbies and veterans when using a particular warbird? We also analyze plane efficiency in correlation to the time of day because, for example, during the day there are mostly players of every skill level playing while in the evening and at night, generally speaking, the hardcore pros come out to play. When looking at the stats, we also take into account the module setup of a plane, its mounted equipment and ammunition type as well as the tier of the battle they get matched into. We also distinguish between regions: Players on different servers prefer planes of different nations, which leads to a differently-looking setup of teams in battle, which, in turn, also influences the way one and the same plane fares in combat.

When we speak about „efficiency“ we mean how much damage, on average, a plane can cause on priority targets for its class, its hit ratio, its longevity in battle (including the frequency of suffered critical damage or fire on-board), a plane’s profitability in terms of credits and XP and many more factors.

If when analyzing data, we discover that the shown stats are outside of our admissible or expected range, we begin an in-depth investigation in which we look for the reason of the differing and how to correct the issue. In many cases, it’s not so easy to come up with a solution. For example: You can improve the efficiency of an Attack Aircraft in many ways, by increasing its damage-dealing weapons output, its accuracy, its durability or by reducing its time to turn-metric, which would allow the aircraft to keep firing on its targets for longer. Of course, another method we employ is to look at the feedback of our community. We regularly check the players’ opinion on things on the forums. Besides that, every plane goes through a regular check-up that’s being performed by members of the Super Test. Certain plane statistics are being assessed based on the results that we see from our regularly-held tournaments (this mostly applies to duel-oriented fighters). Why do we do this? Easy! Firstly, because there’s a higher concentration of top players in tournaments, capable of maxing out every aircraft to its top performance. Secondly, all used setups during such tournament battles are chosen in a way to achieve maximum dogfighting success. All of this allows us to easily assess the balance of aircraft used in tournaments, based on the conditions that players encounter in competition. Lastly, of course, each of our balancing staff members has their own opinion about which plane is “imba” and which one “s#%cks”. However, we don’t take any of these opinions as unshakable axioms and merely regard each one of them as a reason to, once more, take a look at the statistics.

It has taken us a while to arrive at the current balancing methods. We have learned a lot and found out how to avoid past mistakes. As the game grew, developed and became more complex, so did we and our methods in the balancing department. Let me give you some examples from our field of work. Not long before release, we’ve given all of our testing participants a couple of warbirds, among which the Bf 110 C-6. This specific machine is big and thusly not very agile, but possesses a huge gun, which, in the hands of a capable pilot, seemed like it could cut through the competition like a knife through butter. It was important to note though, that not everybody managed to master the aircraft, which is why its overall win ratio settled at a normal 48%. Other key metrics were within the usual range as well. It seemed like the C-6 was an aircraft like any other. Then we released the game, took away the C-6 from all the testers and made it a Premium aircraft. The plane soon became one of the most popular Premium vehicles, and many of those, who mastered it previously, bought it after release as well. That meant that now, not everybody, but mostly only skilled pilots were playing with the C-6, which skewed its statistics and turned it into an undeniable powerhouse that newcomers had a tough time dealing with. After all, nobody is happy when they start playing a new game, seem to be getting along fine, and then hit a brick wall when their own plane is torn to pieces after being hit just once from an overpowered-seeming monster plane. That’s only funny for those shooting but not those being shot down. In such a case, the balance is clearly off. And the reason for that was the inappropriate distribution process (when we tested the plane as an omnipresent warbird in the beginning).

Another example from an even earlier stage of the game is the F2A, whose stats during Closed Beta seemed far from impressive. Back then, we didn’t yet take into account different configurations when assessing a plane’s efficiency. Once we did add this metric into the consideration, we found the following: In its top configuration with its top gun the F2A simply destroyed everything in its path and was thusly extremely successful. At the same time, on its stock configuration, the same F2A was horrible, didn’t win, didn’t do damage and didn’t earn its pilots any XP. Because of this, many players didn’t bother researching it, skipping the plane for its successor without installing any module upgrades. The win ratio of this plane drifted apart: For every one pilot who played on a top F2A configuration and simply obliterated everything, there were 10 other stock pilots, who couldn’t even compete for scraps and pulled the win ratio downward.

4. To achieve all this, do you play the game yourselves or do you mostly look at feedback?

All members of the balancing team necessarily all play themselves, both on the live server and during regular sessions on the Super Test. It’s our Community Managers who follow the general mood of our player audience in the various places where they interact with them, in order to be able to then send us weekly digests that inform us about the goings-on in the forums. In those digests we find all the questions that players currently inquire about. That’s why you can be sure that if the players have something to say about balancing (be it a question, identifying a problem or maybe even a solution) then we get informed about it in a very timely manner. If you have specific questions, please continue voicing them in the various channels. We’ll do our best to get back to you with an answer as soon as possible.

5. What is the specific key moment or metric that makes you decide to boost or nerf a plane?

As I said before, it’s impossible to identify just one single reason to do anything with a plane. A good plane balance is dependent on a positive equilibrium between many factors.

6. What other things can you do to balance a plane besides adjusting its flight characteristics?

As you point out correctly, in most cases, we go to the modification of the flight characteristics of a particular plane. However, there are other ways that can have an equally strong, if not stronger impact: For example we can adjust the global underlying combat mechanics (like the flight model or the damage model). We can add or subtract modules from the setup of the plane. We can also make economic changes (how profitable a plane is in terms of its XP and credits earnings potential). Lastly, we can also make changes to the weight that a particular warbird receives from the matchmaker, which is going to influence the team composition of battles that a player ends up in. I want to point out, that in terms of the matchmaker, we don’t have any plane in the game that’s being placed in the upper or lower echelon of tiers more often than others. When composing a team, the matchmaker looks at the number of flights, plane classes and tiers – however, not specific plane models, which are irrelevant for our matchmaking.