While Cygames had never developed a fighting game before, the Japanese studio knew it wanted to create one based on its popular mobile RPG Granblue Fantasy. To turn that dream into reality, it enlisted the help of one of the best fighting game studios in existence: Arc System Works, the company behind the Guilty Gear franchise and Dragon Ball FighterZ. Together, the two companies co-developed Granblue Fantasy: Versus, which gaming review site GameCrate called “a fighting game for everyone.”
With a fresh perspective for the genre, Cygames took the opportunity to reinvent certain fighting game mechanics. These fundamental changes paid off as Granblue Fantasy: Versus has been heavily praised for offering a fighting system that is accessible to newcomers, but also inventively adds layers of depth for veterans in the process. To see how Cygames and Arc System Works pulled off this amazing feat and to gain insight into the collaborative workflow between the companies, we interviewed Cygames Creative Director Tetsuya Fukuhara, Arc System Works Lead Planner Kazutoshi Sekine, Arc Systems Works Development Manager Hideyuki Ambe, Arc Systems Works Art Director Hideaki Sawada, Arc Systems Lead Modeler Souichi Kawasaki, and Arc Systems Lead Programmer Masashige Kazuhito.
Compared to Arc System Works' past efforts like Guilty Gear and Dragon Ball FighterZ, Granblue Fantasy: Versus is a more methodical ground-based fighting game that values positioning over extensive combo memorization and juggling. Why was this slower pace a good fit for the game?
Cygames Creative Director Tetsuya Fukuhara: Our fighting game is based on Granblue Fantasy, which is a mobile game, and while Granblue Fantasy players are gamers, many active players either didn’t own a console or have not played a console game in an awfully long time. Because of this, we thought it would be best to simplify the button inputs and aim for a slower pace in order for these players to be able to enjoy a fighting game for the first time in a while, if ever. Even if we take Granblue Fantasy out of the picture, fighting games these days have become faster and more complex, which can create barriers to entry. I thought this would be a good place to reset things.
While many special attacks can be performed with traditional quarter-circle movements, they've also been distilled down to single button presses, at the cost of longer cooldown times. This makes these attacks more accessible to newcomers, yet also adds a layer of strategy for fighting game vets as it conceals the windup of charge attacks. How did the studio come up with such a novel fighting mechanic?
Fukuhara: In terms of fighting mechanics, we received various suggestions from Arc System Works. But skills, their usage, and cooldown specifications have been there since the game’s inception. It feels like the most representative expression of Granblue Fantasy, and if you look at it from another perspective, if the title was not Granblue Fantasy, the system would probably have an odd feeling to it. We debated over whether a fighting game could really work with the franchise, and I believe the end result turned out well.
The game has been compared to titles like Soul Calibur with its simple auto combos in conjunction with having a dedicated block button and Super Smash Bros. for its accessible special attacks. Did the team look at other fighting games for inspiration?
Fukuhara: Rather than getting inspiration from other games, I first looked inward and gathered feedback from our staff. Since we are not a company that specializes in making fighting games, we had many employees who were either unfamiliar with them or weren’t very skilled in that arena. With the feedback that I received, the two most common problems were that the commands were too difficult and that they couldn’t block. So, we decided that these would be the best aspects to tackle, and I would say that is where a large part of our inspiration came from.
Granblue Fantasy: Versus' combat system has been heavily praised for being accessible for beginners while offering depth to seasoned fighting game veterans. Can you elaborate on how you struck this delicate balance?
Fukuhara: I believe that the skill system, blocking, and rules for various other actions all mesh very well together. The skills and blocking mechanics, in particular, are simple, but they haven't been fully utilized in other 2D fighting games before. In this regard, there are some rules in this title that can be re-evaluated, or even redefined. Since it comes from a genre with a rich history and set style, we knew that one mistake could make the game off-putting to our audience. However, both companies really understood the concept of what we wanted to create, and together, we continued to move forward without much hesitation. Ultimately, I believe that this resulted in a game that almost anyone can enjoy.
Arc System Works Lead Planner Kazutoshi Sekine: When players start a new game, beginner or not, they tend to want something that is straightforward so that they know exactly what to do. More hardcore players tend to want a line-up of multiple options so that they can actively choose the best one. As a result, for this title, the first thing we did was to split the input methods into two different types and then created a situation where players can choose the best input method depending on the situation.
Fighting games, well, more so, competitive games in general, have constantly changing and reactionary gameplay, so having to think thoroughly about your choices creates a game that can be enjoyed for much longer; at least, that is what I believe. For example, when you react with an anti-air move, you had to choose to use that move in that situation. This decision is simple, but if improvement is the goal, even advanced players will still be satisfied making them. We created the system with this concept in mind.
On another note, if you use a technical command for a skill instead of the simple command, we tried to prevent the technical commands from being too advantageous. Since fighting games are so technical, adding even the slightest advantage can make a big difference. This time, we wanted some parts of the game to be simpler to digest, so with some practice, it should be possible to catch up, technically speaking. We prioritized having to read opposing players over creating a highly technical game.
The characters in the game are quite varied with close-ranged characters, distance-based fighters, grapplers, and more. How did the team approach making a diverse cast of characters that were both fun and balanced to play?
Fukuhara: At launch, fighting games have a clean roster with a diverse fighting style and gender ratio. I believe that there is a theory behind it all, even in the character layout. During the initial stage, we consulted with Arc System Works to shape and select the characters that would be included based on these theories. During this phase, we, of course, found many characters that would fit into different categories such as speedy, grappler, and so on. If we use martial arts characters, as an example, we ended up with two of those, with Ladiva at launch and then Soriz as DLC.
Arc Systems Works Development Manager Hideyuki Ambe: First, we had Cygames submit a list of characters that they wanted to use and had them create a quick summary on whether the character would be close or long-range, their power level, their speed, and how technical the characters would be. We then used that list to match several characters to different fighting styles so that the characters can be designed in a balanced manner.
At that time, there were still characters that didn’t make the final roster. We had a vast amount of characters to choose from, so we were able to pick and choose a cast that everyone would be satisfied with.
The character designs were easy to link with fighting styles, so once we had the roster decided, it didn’t take much to select a style to go with for each character. We started with Katalina and began to work on other characters. Each time we implemented a new character, we would repeat playtests and adjust the balance. In addition, we had a few set rules to follow in order to maintain that balance. For example, damage at the edge of the screen would be different, how long hit stun lasts, and so on. On top of these rules, if the character’s personality can shine through, then we’re satisfied.
Granblue Fantasy: Versus features a novel beat-'em-up RPG mode with fully voiced performances and a loot-upgrading system. Can you walk us through some of your design goals on that front?
Fukuhara: Nowadays, games where you can win or lose tend to be looked down upon. Although fighting games are not limited to competition, I believe that it is important to convey the competitive spirit in a fun way, while highlighting the appealing aspects of the characters in order to maximize the attractiveness of a game through different approaches. When we got feedback from people who are not good at fighting games but are interested in them or have never played a fighting game before, we came to the conclusion that the number one reason people feel averse to fighting games is because they don’t want to lose.
For me personally, this is about the fighting games that I used to play. If I did not feel like playing online against other players, I would default to arcade modes. Arc System Works’ Blazblue had a very in-depth single-player mode, and I would play it quite often. It is very enjoyable, and I thought that it would be great if we could add something like that to our title.
Because it is Granblue Fantasy, both companies unanimously decided to center it around the term RPG. The significance of the RPG mode is that you can proceed in the RPG-like story, while also learning the fighting game mechanics in the process. In addition to the basic game mechanics at the beginning of the game, boss fights act as a sort of tutorial for specific mechanics. The Ares battle focuses on high and low guarding, while the Colossus battle focuses on jumping and evasive tech. All the way until the end of RPG mode, players will pick up skills that are applicable in competitive play.
While the game looks like a beautiful 2D fighter, it's actually a stylized 3D fighting game with art assets that make it look 2D. Why was this design decision right for Granblue Fantasy: Versus?
Fukuhara: I believe that the best way to translate the artistic style and modeling of Granblue Fantasy was through the wonderful limited animation and art style created in Guilty Gear Xrd. For example, you can use 3D physics to create something like a moving cape, but with limited animation, you can add cool movements and silhouettes one frame at a time, though this does take a little bit more time to create. This is something that is not possible with 3D physics. With this method, I believe that it is possible to express the beauty of the original illustration without compromising performance.
Arc Systems Works Art Director Hideaki Sawada: This may be largely due to the fact that the original art for this title had a certain warmth to it that fits very well with the popular anime-like 2.5D style used in Guilty Gear Xrd. The models used for the background art have very few linear parts and create a distorted look as if it were drawn by hand. The character models were also created in a way that facial expressions and body movements are expressed differently depending on motion and angle. In general, I was very conscious of these things and was careful to have it resemble a hand-drawn style.
The attacks feel very satisfying and weighty. How did the team execute here?
Arc System Works Lead Planner Kazutoshi Sekine: This is not a game that is centered around continuous combos, so we set the hit stop to be on the longer side overall. However, tempo is especially important in fighting games, so we set the hit stop to be at the maximum [threshold] without losing the sense of speed.
Considering the game was developed by both Cygames and Arc System Works, can you delve into how that relationship worked and detail what the workflow was like for the development team?
Fukuhara: For general task allocation, Cygames oversaw production for concept art, 2D illustrations, battle dialogue, and music. Technical aspects, the creation of 3D models, and scene production (scripting, effects, sound effects, and such.) were headed by Arc System Works, but each element and the overall direction was managed by Cygames.
Since the Arc System Works staff had a great love and understanding of Granblue Fantasy, they often offered suggestions for battle specifications and character motions. As someone from the Cygames side, and as the person who is directing the entire Granblue Fantasy IP, I personally acted as the direct contact and gave as much specific feedback as possible, so that the development could move forward quickly.
Ambe: Our workflow allowed us to interact with each other frequently to increase the quality of the game every time. We were always amazed at the surprisingly fast response from Cygames and the information, consultation, and confirmation that we needed in each scene during development. The responses were detailed and easy to understand, so the development team was able to proceed stress-free. We understood that the feedback from Cygames was for the betterment of the project, so even if they were sometimes difficult, our development team aimed to always meet and exceed Cygames’s requests. Even if Cygames proposed an idea that might be a heavy burden for us, they were such a pleasure to work with that our staff was able to continue development in a very motivated manner. As a result of this repeating process, I feel that we have developed a high-quality game.
Why was Unreal Engine a good fit for Granblue Fantasy: Versus?
Arc Systems Lead Modeler Souichi Kawasaki:
Unreal Engine was a good fit because:
The threshold for editing is low, making it easy to create materials as needed.
It is easy to import materials and implement them into games without needing a programmer to be the middleman.
It is easy to update model data, such as adding bones, and the existing data will largely not be affected by these changes.
Arc Systems Lead Programmer Masashige Kazuhito: I would like to add:
At the start of the game’s development, we already had some development experience using UE, and it has the advantage of being simple to learn.
Because it is a free-to-use program, there are many users, which means that in many cases, there is information readily available on the internet when there is a specific function that we want to implement or when a problem occurs.
After implementing the graphic data, and having it reflected on the screen, it’s easy to make design changes without needing a programmer.
Do you have any game design tips for other developers?
Sekine: These are some things that I always keep in mind:
Have a consistent concept for gameplay (what kind of interactions you want to create, what kind of experience you want players to have).
Determine in-game rules (what’s playable, things that can’t be done).
Increase the things that can be done within those rules to create variations in playstyles.
Having said that, do not let the rules consume you and let the ship steer itself (if you become too obsessed with the rules, your playable range will become narrower).
Even if you're playtesting the game yourself, always play with a critical eye.
What you thought was correct yesterday may not be correct today, so re-examine.
Regardless of whether you're working on an original IP, have respect for the creation.
It is okay to be kind to yourself once in a while.
Thanks for your time. Where can people learn more about Granblue Fantasy: Versus?