Return to Grace (on Steam here) is an atmospheric sci-fi walking sim that takes about an hour and a half to finish.

You play as Adie, a space archaeologist who’s searching for a legendary AI named Grace on the Ganymede moon of Jupiter. Inside the immaculately preserved ruins of a structure built to worship Grace, Adie encounters various aspects of Grace’s personality, who communicate through her digital wristband. As she explores the beautifully appointed space station, Adie searches for the answer to a mystery that’s gone unsolved for more than a hundred years: why did Grace suddenly disappear?

The game’s story is well-paced and engaging, and the voice acting is excellent. Return to Grace also showcases one of the most visually striking examples of retro futurism I’ve encountered. The references to everyday 1960s architecture and interior design fit the world and its themes perfectly, as do the more elevated hints of Art Deco.

What I’d like to do in this review is analyze and criticize how the game goes about telling its story, but please understand that this is more about me solidifying some general thoughts about storytelling in video games and less of an actual negative assessment of the game itself.

Essentially, while I have no problem with a game rewarding completionism, I don’t think developers should expect this sort of behavior from the player by default.

More specifically, I feel that some of the more “interactive” elements of Return to Grace don’t work the way they’re supposed to, and this diminishes the effectiveness of the game’s storytelling.

To begin with, the opening twenty minutes of Return to Grace are rough. Adie crashes on Ganymede, wanders through a snowstorm, and navigates a series of anonymous techno-corridors as she attempts to enter Grace’s station. The game doesn’t signpost objectives, and the low visibility and featureless visual landscape make it easy for the player to get lost.

I can understand why this initial difficulty exists on a thematic level. Without AI guidance, the developers suggest, the outer reaches of space are ugly and hostile. Still, while this sort of introduction would work in a movie, it’s perhaps not the best choice for a video game.

It’s not fun to be immediately hit with frustration that delivers no narrative reward, and I feel like the developers aren’t respecting the law of conservation of detail. In a shorter work, each detail matters more, and there’s no narrative reason to have a full 20+ minutes of the player wandering around aimlessly at the very beginning of the story. This section of the game could have been shorter, or at least “easier” in the sense of offering more guidance.

Thankfully, everything changes once Adie enters the station proper. The visual environment is beautiful, and the aspects of Grace’s personalities begin to have amusing and enlightening conversations with Adie and with each other. One of these aspects, “Logic,” suspects that there might have been a good reason for Grace to shut herself down, and she starts to put artificial barriers in the way of Adie’s ascension to the peak of the station.

To summarize, Logic essentially requests that Adie locate and listen to a collection of voice recordings left behind by the station’s last residents. Navigating a gorgeous 3D space to piece together a lore puzzle sounds like it would be great fun; and, for the most part, it is.

Unfortunately, the game puts two artificial points of friction in front of each discovery. First, the player must complete a simple but annoying memory puzzle in order to unlock each residential unit in the station; and second, each recording can only be accessed after completing a short but annoying spatial puzzle on Adie’s wristband. This isn’t necessarily a problem the first few times, but it gets very stale very quickly.

Eventually, an aspect of Grace’s personality called “Control” offers to complete these puzzles for you, thus removing the friction. Control is a weird asshole, to be sure, but it’s difficult to overemphasize how immersion-breaking and repetitive the puzzles are.

While roleplaying as Adie, my thought process was that perhaps Control might be less of an asshole if he had something to do. We all want to feel useful, after all, and “wanting to feel useful” seemed to be a throughline across the various aspects of Grace’s personality. Meanwhile, as the person playing Return to Grace, I was invested in the game’s story and interested in learning more details while exploring, and I didn’t like my flow state being interrupted by meaningless tasks.

It turns out, however, that allowing Control to do even one of these puzzles locks you out of the most satisfying ending, which is the ending in which you get to meet Grace.

In other words, the default ending is a “bad” ending that answers no questions and offers no narrative resolution. In addition, Control kills Adie in a decision that has zero foreshadowing. I don’t even consider this death a spoiler, as it comes out of nowhere and makes very little sense. The default ending isn’t really an “ending,” but rather a punishment dealt to the player for not finding all of the voice records while manually completing all of the puzzles.

This leads me back to the point I started with, which is that I don’t think games should expect completionist behavior from the player. Some players might come to a narrative adventure game like Return to Grace for pointless busy work, sure. It’s not beyond the realm of possibility. Still, I’m guessing that most of us don’t play cinematic story games for the purpose of ticking boxes off a spreadsheet, and that most of us aren’t going to want to play the entire game through a second time after seeing such a pointless and upsetting ending the first time.

What I would like to see in Return to Grace (and other narrative-driven games) is a default ending that respects an organic style of progress through the story and doesn’t assume that the player will have a walkthrough constantly open on a second screen. There’s no need to penalize casual players! Let the default ending be satisfying. 

In the case of Return to Grace, Adie’s meeting with Grace really should have been the natural conclusion to the story. For players intrigued by the fictional world, a satisfying “default” ending will trigger curiosity concerning different possibilities. I’m all for different endings dependent on more attention and devotion to the gameplay, but I think these endings should be weirder and more interesting than the natural default ending. 

In Return to Grace, the “Control kills Adie” ending would make far more sense if the player’s completionist behavior were perhaps framed as Adie resisting Control by refusing to allow him to help her, for instance. Also, I tend to think this “bad” ending with Control is far more emotionally compelling if the player has already talked with Grace in the “good” ending.

Again, I’m all for developers rewarding completionism, but it shouldn’t be taken for granted. This is especially true in games whose mechanics serve to break the flow of the story, even if this interruption is thematically intentional.

I’m using Return to Grace as an illustration of a general argument not because I dislike it. I actually really enjoyed this game! 90% of Return to Grace is very good, but I thought it might be worthwhile to explore the 10% in which the ludic medium is something of an uncomfortable fit for the message.

Putting my concerns regardings its endings aside, Return to Grace is a thought-provoking narrative adventure filled with uniquely stylized artistic flourishes, and I’d definitely recommend checking it out if you’re a fan of What Remains of Edith Finch and Outer Wilds.