"Clash Games for a Blast"—Jonathan Blow of “Braid” and “The Witness” on Four-World Puzzle “Order of the Sinking Star” [Interview]

What kind of surprises might emerge when you combine the mechanics of different games? “Order of the Sinking Star,” developed by Thekla, is a puzzle-adventure game built on a warehouse-style puzzle foundation, where four worlds intersect.

Jonathan Blow, creator of *Braid* and *The Witness*, is the founder of Thekla and serves as the director of this title. Game*Spark interviewed Blow approximately two weeks after the game’s release. While players are free to solve puzzles in any order, unexpected solutions are addressed.What is the goal behind balancing free-form progression with precise puzzle design? We spoke with him about everything from why he chose the “warehouse keeper” concept to the roles of the characters and the story, the team’s development process, and his current impressions of the game.

Combining Different “Warehouse Keeper” Elements into One Grand Game

――You’ve previously worked on the 2D platformer *Braid* and the 3D first-person game *The Witness*. Why did you choose a “stacker” puzzle as the foundation for this project? What aspects of the “stacker” concept—its appeal and potential—caught your attention?

Jonathan Blow (hereinafter Mr. Blow): There were at least two appealing aspects to the “warehouse keeper” concept. First, it’s very simple. Just by having objects that fit within a single square interact with one another, a wide variety of puzzles emerge. As a designer, I feel it’s a very well-crafted mechanism.

When designing games, it’s easy to end up making the mechanics too complex. With Tic-Tac-Toe, on the other hand, you can understand the situation at a glance. There aren’t a lot of hidden mechanics; you can easily grasp where the pieces are and how they move.

Another reason was that I wanted to carry forward the fun of combining different mechanics, which I had explored in *The Witness*. Symbols were placed on panels where you drew lines, and each symbol had its own set of rules. Even symbols that didn’t make sense at first would create surprises when combined with other symbols after you’d learned their rules.

This time, we decided to expand our experiment with combining mechanics on a much larger scale. Rather than combining small symbols, we’re combining existing games themselves. It’s like a giant collision accelerator for game design. The image is of different games colliding and causing a massive explosion.

However, the mechanics of the games being combined need to mesh well. In warehouse-style games, for example, there are commonalities in how objects move—they all move along a grid. That makes it easy to imagine how things would work if you combined different games.

For example, combining *Monster Hunter* with a racing game might produce something quite unusual. While it could be an interesting combination, it’s not what we were aiming for this time. The “warehouse management” format helped us bring the game we envisioned to life.

――This game is on a massive scale—did you envision it being this large from the start? Also, why did you choose to complete it as a single game rather than splitting it into multiple titles?

Blow: At first, I thought it would be interesting to combine different worlds. In the original plan, there were six worlds, not four. If we’d proceeded according to that plan, we probably wouldn’t have even been able to finish it.

What we’re aiming for with this game is to connect the worlds and show players the fun of those combinations. For example, what happens if you combine an item from one world with a hero’s ability from another? We test each one individually. Even if we eliminate combinations that aren’t fun, we can’t achieve our goal unless we test and show players every single interesting combination.

As we explore these combinations, the scale naturally expands. If one world has seven possibilities and another has nine, that makes 63 combinations. When a third or fourth world is added, the number of combinations grows even larger. Simply following the concept of combining worlds results in a massive game.

It was probably inevitable that it would become this large. Perhaps I was fooling myself at first. If I’d known it would become this big and take this much time, I probably wouldn’t have been able to finish it.

I didn’t split it up because I felt that doing so would compromise the quality of the work as a whole. Everything is integrated into a single entity, allowing players to wander as they please and see things for themselves. A format that allows for free exploration suits this game best. Even if there had been ways to split it up, I didn’t want to do so at the expense of the game’s quality.

Also, even if I had split it up, it’s not certain that I would have been able to shorten the development period as much as you might think. No matter how I chose to build it, the necessary technical work took a lot of time. Art production took time as well, but if I had split it up, I might have finished sooner. It’s possible I could have released the first game in about two-thirds of the time. However, it wouldn’t have reached the same level of polish as the current version.I always want to create the best possible product.

Even with Free Progression, Unexpected Solutions Require Adjustments

――I think games with a linear progression make it easier in some ways to adjust difficulty and tell a story. Even so, why did you choose a structure for this title that allows players to tackle puzzles in any order they like?

Blow: When I was younger, the games I played were mostly linear. As puzzle games became more popular, problems began to arise with linear designs. In an action game, even if a section is too difficult to progress through, you at least know what you need to do. If you keep trying and improving, you’ll eventually get past it.

But puzzles don’t work that way. You might never solve them if you can’t come up with the answer. The main solutions the gaming industry adopted were to make the puzzles easier or to simply not include them at all.

At least in the West, there are games—such as the *God of War* series—where, the moment you enter a room and face a puzzle, the character starts giving you the answer within about 30 seconds. That’s far too short to think it through on your own. While this approach of immediately providing the answer had become widespread, I didn’t like it.What I wanted to create were authentic puzzles. While some players might not be able to solve them, my goal was to make puzzles that could be solved with enough thinking, enthusiasm, and persistence.

However, it’s not good design if a player gets stuck after completing about 10% of a game they’ve purchased and can’t play the rest at all. There are ways to provide hints or allow players to skip puzzles, but I don’t like either of those approaches. I want players to be able to move on to other areas.

The ability to take a different path when you get stuck. And the fact that you don’t have to solve every single puzzle to reach the end. Those are the two key points. Even if you get stuck on one puzzle, you can still clear the game if you make good progress on the others. It’s not just about puzzles—I simply enjoy games designed this way.

――When a player solves a puzzle in an unexpected way, how do you decide whether to accept it as a valid solution or treat it as a loophole that needs to be fixed?

Mr. Blow: In this game, we treat all of them as loopholes that need to be fixed. However, this isn’t because we don’t appreciate the players’ ingenuity. There are simply many different kinds of puzzle games.

The puzzles in this game aren’t just meant to test the player’s wits. Every puzzle contains an idea that hasn’t yet been revealed in the game. We want to convey the fun of objects behaving in unexpected ways.In the early stages, we first let players get to know the properties of objects, and once they understand how things work, we introduce new ideas one after another. What happens when the world intersects and objects interact with one another? We design the stages to evoke a sense of surprise and fun.

Of course, no matter how carefully we design the game, with a title of this scale, there’s always a chance we might overlook some unintended solutions. However, if a puzzle is solved in a way that deviates from our intent, players may finish without ever experiencing the ideas we wanted to showcase. Moreover, in many cases, the sense of satisfaction upon solving the puzzle is diminished. It’s like reaching the goal by moving objects haphazardly, but feeling no real sense of accomplishment.We want to avoid that kind of experience in this game.

That said, this doesn’t apply to all puzzle games. Some games embrace the idea of multiple solutions. In this title, however, we’ve chosen not to allow multiple solutions.

Worlds Intersect Not Only in Puzzles but Also in the Story

—Compared to your previous works, the story plays a much larger role in this game. Why did you decide to focus so heavily on the story this time?

Blow: Our approach to storytelling varies from game to game. For example, *Braid* placed a greater emphasis on story than *The Witness*. We’re not trying to add more story with each new title; rather, we choose the approach that best suits each individual game.

At the heart of this game is the idea of worlds intersecting. I was always thinking about how to better convey the fun of worlds intersecting. It might sound simple, but in the early stages of development, we weren’t able to express it as well as we are now, and I think it was hard to grasp even while playing.

That’s where the story comes in. By depicting what kind of place the setting is, the nature of the world becomes clearer. While the game mechanics convey this to a certain extent, adding a story makes it even easier to understand.

The heroes are a good example. When you understand not just their abilities but also their personalities through the story, they leave a lasting impression. They’re not just a red character—you come to understand who they really are.

Furthermore, as you progress through the game, the role of the story changes significantly. The story also becomes a means of directly conveying the game’s design philosophy—what this title is aiming to achieve.

Not every player will be interested in the philosophy behind game design. However, there are players who are interested in those aspects. This title is structured so that not only does your understanding of the story’s world deepen as you play, but the reasons behind why this game was created also become clear through the story.

――Did the large number of characters affect the puzzle mechanics or the game’s design? If there were any specific factors that influenced them, please tell us.

Blow: First of all, the game we based this on also had many characters, each with one or more abilities. From a purely mechanical standpoint, the functions of objects and the abilities of characters can be viewed in the same way. In practice, however, we often found that presenting them as character abilities made them feel more distinctive and was more enjoyable. That said, there were also cases where the opposite was true.

Most of the characters are based on Jonah Ostroff’s *Heroes of Sokoban*. They’re fantasy heroes in the vein of *Dungeons & Dragons*. The balance in the original game was slightly different; it felt geared more toward puzzle enthusiasts.

As I reviewed what worked well and what didn’t in this game, I realized that playing as the Bard was incredibly stressful. The Bard has the ability to charm other characters and objects and move them alongside himself. Despite being a powerful ability, I kept getting stuck in puzzles, which gradually became frustrating.

Since the ability is tied to the character, I want players to grow to like the character themselves. I don’t want them to end up disliking the character, and I certainly don’t want them to feel stressed while playing.

So, I put a lot of effort into seeing if I could flip the puzzle concept on its head. Instead of thinking about what to charm and how to do it to avoid getting stuck, I transformed the puzzles into ones where, in more open spaces, you consider how to use the ability to move objects from one place to another. The Bard changed significantly during development, so it’s a particularly clear example.

――I think *Braid* and *The Witness* also blend elements of fiction and non-fiction, designed so that the player’s perspective and way of thinking change as the game progresses. Is there a particular reason you like to connect fiction and reality?

Blow: It’s true that all three games connect fiction and reality. The approach differs each time, but I seem to enjoy depicting that connection to reality. I’m not entirely sure why myself. It’s hard to put into words, but perhaps it’s because I value a certain “authenticity” in the stories I tell.

If I connect the story to something in reality that I find interesting or valuable, it creates a sense of authenticity precisely because it’s based on something that actually exists. Of course, fiction also has its own unique sense of realism. However, depending on how it’s portrayed, it doesn’t always work out.

Sometimes, when playing a game, I feel that turning a story into a narrative can make a topic that should be taken seriously seem trivial. It’s as if people think, “It’s just a game, the characters are fictional, and it’s set in a distant world, so it has nothing to do with reality.” I have no problem with games that are purely meant to be entertaining, and I’m not saying they’re bad.

However, when a game is trying to address a serious topic but then backs off by saying, “No, just kidding. You might have thought I was getting too serious, but it’s just fiction,” it can actually make the story feel even more trivial. Perhaps I just don’t like using fiction as an escape route. Maybe I don’t want to leave that kind of escape route open for myself, either.

If I add just a little bit of reality, players can think about the game’s ideas in direct connection to the real world. However, I try to leave it up to the players to figure out the connection to reality and what’s actually happening on their own. I don’t want to dictate, “You should think of it this way.” In a sense, that’s just my style, and I can’t really explain it any better than that.

Relationship with the Original Creators and Puzzle Development Within the Team

――How would you best describe your relationship with the four original works and game series that served as the foundation for this title—as “being influenced by them” or “collaborating on their creation”? Please also tell us how the original developers were involved.

Blow: It’s not just a matter of being influenced. We obtained licenses for each game and then made adjustments to the level design and some of the mechanics.

One of the game’s appeals is the sense of surprise that arises when these worlds intersect. To create that surprise, we’ve incorporated elements from the original games while giving them a slightly different twist. However, the original games’ distinctive characteristics remain firmly intact. If you compare them side by side, you’ll see that even though the graphics are vastly different, there are commonalities in the stage layout and other aspects.

Early in development, we spoke briefly with the designers of the original games. We discussed what they were aiming for in each aspect, what they valued most, and why the designs turned out the way they did. We wanted to make sure our adaptations remained faithful to the originals.

That said, the original designers were not part of this game’s development team. The actual puzzle design was handled by different members.

――What discoveries or challenges did you encounter while collaborating with other designers to create these puzzles?

Blow: It was completely different from anything I’d done before. I handled all the design for *Braid* and almost all of it for *The Witness*. But this time, the scale was so large that I couldn’t have created it on my own.

Since this game is based on a previous title, other designers were involved from the conceptual stage onward. Furthermore, I needed help from others to bring this game to life. Fortunately, a group of truly outstanding designers came together for this project.

In the past, I used to come up with ideas on my own, but this time we held weekly discussions and shared what each of us had created. It was fun to create while discussing ideas together.If you’re making a different game every time, you need to adapt your approach to suit each project. A method that worked well in a previous game might not be suitable this time. If you keep using methods that don’t fit the project, it could negatively impact the game’s quality. With this project, it became clear that designing as a team was the right approach.

Another interesting aspect was the live streams during development. While we were building stages, viewers would sometimes offer help via chat. The credits list not only professional designers but also quite a few viewers who contributed ideas and helped create puzzles. It was also interesting to see how developers and players were connected through the creation process.

――I also watched your Twitch streams. Could you tell me what inspired you to livestream the design and other production processes, and how you went about making it happen?

Blow: I gradually started streaming the production process as well. Even before live streaming became widespread, I had been giving talks at conferences and the like, so I was used to speaking in front of people. Once live streaming became possible, I wanted to show people how I was actually programming. At first, the content was mostly technical, and I even streamed talks focused on technology. Eventually, I started showing the design process as well.

I think I was also influenced by my friend Casey Muratori’s stream series, “Handmade Hero.” It was a very popular series where he built games from scratch without using game engines or any pre-made assets. I thought I could do something similar.

However, I didn’t intend to show everything like “Handmade Hero” did; I just planned to give occasional glimpses into the development process. Some days I’d program, and other days I’d work on design.

I don’t think there are many programming streams on Twitch, especially ones by professional developers. If you work for a company, all sorts of issues can arise, so it’s probably difficult to get permission. Since I run my own company, I’m able to stream. While I occasionally see people from other companies streaming their programming work, I—at least—had rarely seen anyone share their design process.

I’ll be building the stage while talking about the challenges I’m currently facing. I’ll even show the moments when I’m stuck and unsure of how to improve things. I found it interesting to share this process of trial and error. I’d be happy if the viewers enjoyed it too.

――I think one of the things that makes Mr. Blow so appealing is that he creates a different game every time. For this project, you’ve even developed your own programming language, “Jai,” and a game engine. How has this custom development environment influenced your game design?

Mr. Blow: Programming in Jai is fun, and it speeds up the work, so it’s indirectly helpful. However, at least in my case, I tend to keep the technical aspects and game design ideas quite separate.

In the past, designers who couldn’t program had to convey their ideas and have programmers bring them to life. If programmers struggled to implement them, it could hinder the game’s design. But now, even designers who can’t program can create prototypes using AI. The problem of design being dictated by the difficulty of implementation may no longer exist.

I plan to use the same engine and language for my next game as well, but the underlying concept will be different. It will likely be a completely different game. In terms of creating something different every time, that won’t change going forward.

Post-Launch Reception and the Future

――It’s been about two weeks since release. How has the reception been so far? Have there been any unexpected reactions from players, or anything that’s particularly stuck with you?

Blow: Things are going well. Before the release, I couldn’t predict what kind of reaction we’d get.It’s always hard to predict, but this time it was especially unclear. *The Witness* is an unconventional game—even in the trailer, you see a series of line-drawing puzzles. Some people might think, “This looks just like a newspaper puzzle. Is there really any point in playing this?” I wasn’t sure if people would accept it, but in reality, many people have played it.

This time, the format is familiar to anyone who plays games. However, “warehouse-style” puzzles aren’t exactly a hugely popular genre. I’m not entirely sure why that is. Personally, I find them very appealing, but they can be too difficult at times; while experienced players might appreciate the fun, it might not resonate as well with first-timers.

So, I made a special effort to convey the appeal of “warehouse-style” puzzles. I feel that message came across. Whether on Steam or consoles, the review scores were higher than I expected—a pleasant surprise.

I’d be happy if players continue playing and come to enjoy it even more. In particular, most of the reviews for this game on digital distribution platforms come from people who’ve played enough to evaluate it but have only experienced a small portion of the game as a whole. That’s just how massive this game is. I feel like players are enjoying it more the further they progress, but since it’s still early days since release, the real journey is just beginning.

Right now, I’m truly relieved. With indie games, even good titles can sometimes go unnoticed. I’ve seen examples of genuinely good games getting buried, so there was always a lingering anxiety in the back of my mind. However, this time, many people have played it.

I think one reason so many people played it was our partnership with the publisher. One of the reasons I signed with Arc Games was that there are far more games out there now than before, making it harder for people to discover new titles. I believe my game has a good reputation, but it’s been 10 years since *The Witness*.Someone who’s 19 now would have been 8 or 9 back then—perhaps too young to play it. I’m truly grateful for the support from PLAYISM in Japan and publishers in other regions.

If a game isn’t good, players will know it no matter how much you promote it. On the other hand, even a good game might go unnoticed because so much is happening every day. Still, I believe that if we make the effort to get the word out, it will reach people. If they give it a try, they’ll surely recognize the game’s value.

It’s been two weeks since release, and I’m truly relieved now. A weight has been lifted off my shoulders, but the work continues. We’re currently fixing the minor issues within the stages I mentioned earlier and making further improvements.

――Finally, while I imagine you’ll continue polishing this game, if there’s anything you can share at this point about plans or concepts for your next project, please let us know.

Blow: It’s still a bit too early to talk in detail about the next project. Right now, I’m focused on this game, and if I release information too early, players might think, “This isn’t what was promised,” if the content changes during development. I want to avoid any mismatch between expectations and reality.

I’ll share more once we’re at a stage where we can show it to you. It’s not quite fully formed yet. We won’t begin full-scale development on our next title until 2027.

*Order of the Sinking Star* is currently available for PC (Steam), PS5, and Nintendo Switch 2.