How an isometric RPG was created in Russia — the development history of Pathfinder: Kingmaker

Alexander Semenov, Senior Editor App2Top.ru Q: How did the development start in your case? Who came first with a brief: “Let’s make a CRPG”?

Oleg Shpilchevsky

Oleg Shpilchevsky, head of Owlcat Games studio: Now it already seems that it happened by itself. The core of our team is me, Alexander Mishulin and Viktor Surkov, we have been working together for more than 13 years, and Sasha and Vitya have been working together for even longer, about 18 years [All come from Nival. Oleg led the development of Allods Online. Alexander was the designer of the “Cursed Lands” and “Demiurges”. Victor was responsible for the art in Silent Storm. They worked together on the iconic Heroes of Might & Magic V project for Nival. — Editor’s note]. We have long had a desire to make a story RPG. Therefore, when the opportunity arose to implement this idea and get funding, none of us doubted.

How long did the pre-production last?

Oleg: Depends on what you mean by that. Vertical Slice lasted about eight months, after which we launched the first Friends & Family alpha test.

Was there any concern that the market is not ready for a large number of such projects (for example, Tyranny sales are noticeably less than Pillars of Eternity sales, and Dreadfire sales are lower than Tyranny sales)?

Oleg: And sales of Divinity: Original Sin 2 are noticeably higher than Divinity: Original Sin 1 and all these games. No, there were no concerns about the audience. There are about three million people who like classic RPGs. There are not very many games in this genre per year. Therefore, a person has enough time to play everything.  If we talk about concerns, I would say that when developing, there are always fears “is your game good enough to be wanted to play it”.

You probably initially calculated what the size of the team should be. How accurate was the forecast? How many people were attracted in total?

Oleg: The team, which is in the state, has remained calculated – 31 people. At the same time, we use outsourcing for some tasks, and now it turned out to be more than we expected. For example, at the peak of the project, approximately 30 external specialists worked with us.

What were the initial deadlines for development?

Oleg: We really wanted to do everything in two years.

How comfortable was the stated deadline? Or did you have to face a crunch by the end of development?

Oleg: In general, during the entire development, we worked quite hard, but without strong overwork. There were small crunches before the release of important versions (alpha and beta tests). And the last couple of months have been quite difficult.

Looking back, what would you have changed at the planning stage?

Oleg: I would cut off some content: there was a lot of it in the game. Much more than we had originally planned. For example, if you initially laid about 800 thousand words of text, then in the end you got more than 1.2 million. On the one hand, “more is not less”, and, as it seems to us, the game turned out to be interesting. On the other hand, the additional content “ate up” the time that we laid down for the final polishing.

Can you share the budget or at least tell me how much minimum should be put into development if the team wants to make an isometric CRPG?

Oleg: Unfortunately, this is confidential information. But making games is expensive. Let’s put it this way: the money we raised on Kickstarter [with the help of crowdfunding, the game raised $909 thousand. — Ed.], helped expand the game. By ourselves, without initial funding from the parent company, we would not have been able to make a full game with the collected funds.

What stages did you divide the development into and how much time did you put into each?

Oleg: Everything is very conditional and depends on many factors: platform, team, tools and genre. In our niche (classic RPG on PC), the time intervals for different stages of development are approximately the following. The concept will take 2-4 weeks. Pre-production strongly depends on the availability of a working engine and pipeline, but for a new project it is worth laying at least 3-4 months. Vertical Slice takes 6-8 months. In general, it takes about a year from the beginning of development to the First Playable product, which is not a shame for beta testers to show. Next comes the production, it scales from the planned amount of content (8-12 months). This stage should be divided into milestones with testing of the finished content in beta tests. In addition, you need to leave at least 3-4 months to finalize the project.

In one of your previous interviews, you noted that the game is designed for a period of 40-80 hours. How are such things usually calculated in general?

Oleg: There are speculative estimates based on the experience of similar games and planned content. And then they are clarified when the first content is collected (in the Vertical Slice process). It was then that we realized: to see the whole story embedded in the original Pathfinder: Kingmaker module, the player will take at least 80 hours (as a result, while statistics say even more time).

You have chosen Unity for development. How much did it have to be redone for the project?

Oleg: This is a difficult question. The engine, as well as the main components: render, resources, initial pipeline, are not radically altered at all. If such a need arises, it means that you made a mistake in choosing the engine. But project-specific elements (editors, pipeline fitting) are added always and everywhere. We have also fully invested in these areas. But in general, this is endless work – there is always something to improve.

What was the main difficulty when working with Unity?

Oleg: The engine’s proprietarity is a double–edged sword. Redoing the “internal” is always evil, since then any engine update turns into a headache for you. We immediately decided that we wouldn’t do that. In general, I think this is the right decision – it made it possible to switch to new versions fairly painlessly throughout the development. But some bugs, of course, sometimes interfered. At such moments, hands reached out to buy a code Unity and “do it right”.

Obsidian, inXile and Harebrained Schemes also make their isometric games on it. But they usually work with pre-rendered backgrounds. Why did you decide to do everything in 3D?

Oleg: Companies like Obsidian or inXile use, in fact, one “over-engine” that sets them a pipeline of work with art. They know the technology, are used to working with it. In general, the principle of “work where you can” is one of the main criteria for choosing an engine, as it saves time. At that time, we had no experience working with a rendered environment. Moreover, we have always worked with honest 3D, and the 2D environment seemed to us to be an excessively limiting factor. Therefore, the choice was made in favor of a 3D environment.

Which is cheaper to do — 2D or 3D?

Oleg: The 2D from the above example is the final result, which is what the player sees. To make it, 3D models are also created, from which the 2D environment is then rendered. Therefore, here, rather, it is necessary to ask what level of tech is used when working with art assets, how many “fashionable” chips and technologies developers want to use in the project. This can increase the amount of work on art several times, although the result together will still be formally called 3D. Therefore, it is very important to immediately set the “minimum technological bar” and assemble a “picture close to the final” on it as soon as possible in order to understand whether players will like it, or whether it is necessary to “wind up” the art stack of technologies.

You have invited Chris Avellon to consult. It still seems to me that it was primarily a marketing ploy. How much did his name help in the promotion?

Oleg: Chris is a strong designer and a “strong name”. Both of these components were important. I would even say that a strong narrator with a Western mentality was fundamentally necessary for us. It was important to have a person in the team who could assess the consistency of content for Western markets from the point of view of his mentality. The fact that this “name”, of course, was only a plus. It is difficult to assess how it helped in the promotion. It’s not buying traffic, but offhand we can say that yes, it helped, and it helped a lot. They always say about Chris: he is a figure of world importance in narrative igrodev.

Can you share any important thing that he taught or opened his eyes to?

Oleg: From my point of view, Chris was good at coming up with ideas for “spectacular content presentation”. When at some point “it is necessary to surprise”. For example, he offered an excellent idea of what is called meaning, that is, the philosophical background of the unfolding plot.

Returning to marketing. Do I understand correctly that Kickstarter has become for you first of all a way to declare yourself, and whether or not the game would have gained money in the end was not so important?

Oleg: Yes and no. Everything about marketing – yes. Kickstarter is a great tool to make your project recognizable at the initial stage. But it was really important for us to make sure that our interpretation of the classic RPG interested the audience. If this attention, expressed in the funds raised, would not have been there, most likely, there would not have been a project, since making a favorite game is good, but making a game that is not needed does not make much sense. So Kickstarter gave us, as they say, a start in life.

There were rumors that part of the amount on Kickstarter came from the management itself Mail.Ru Group. What do you think about it?

Oleg: That’s not so. For the game direction Mail.Ru Group [Owlcat Games is one of the internal studios of the Russian Internet giant. — Editor’s note.] The Pathfinder: Kingmaker project was an experiment. And Kickstarter, I repeat, helped us understand that the players have an interest in the project. If there was no positive reaction from the audience, there would be no game.

Of the marketing activities that you undertook at the stage of preparing the release, which, in your opinion, played the best?

Oleg: Participation in PAX was unexpectedly effective. This allowed us to communicate fairly closely and honestly with the community.

A very important question is: why did they not go to early access? The audience would be more loyal.

Oleg: Good question. Unfortunately, without A/B testing, it is difficult to say how it would be better. By the time of the release, many activities had already been launched, and they all became preparations for the release of the game.

And, I’m sure, a sore question for the team: how did it happen that at the start of the project was still, judging by numerous reviews, raw?

Oleg: I would highlight a couple of points in answering this question. First, as I said, underestimating the content. We should have finished production a little earlier and had more time to polish. Secondly, to be honest, although theoretically we understood the complexity of testing such games, but the combinatorics of the verification options turned out to be very large. To properly test one of the branches of the game, it takes at least a week, and there are hundreds of such branches. Simple functional testing is difficult to guarantee stability.

We are now faced with cases when the content passed dozens of times with a certain combination of previous solutions turns out to be an error. In any case, the team is now very energized to quickly fix the main points that the players complain about. Moreover, despite the mistakes, we like the game. Even people who give us not very positive feedback usually write in the spirit: “I like everything, fix it just here!” And we are working on it.

By release. You have chosen a window between Marvel’s Spider-Man and Assassin’s Creed Odyssey. Why is this time, why couldn’t it wait until January?

Oleg: From the point of view of the audience, our projects are not particularly competitors to each other. I repeat: we focused on very specific players. It was important for us to create a game for them and then reach out to these same people with the story “look what we have done for you.” It looks like we’ve got it.

What were the estimates based on the initial circulation? How do you evaluate the sales of the first weeks in general? Below/above expectations?

Oleg: Again, the details are confidential. After some time, perhaps we will be able to tell them [According to SteamSpy, the number of owners of the game as of October 30, 2018 is from 100 to 200 thousand. At the same time, the service estimates the sales of Pillars of Eternity II: Deadfire from Obsidian, which was released in May. — Ed.].

What’s next?

Oleg: First — policing. Then a few planned DLC and an exit to the console.