Return of the Obra Dinn is probably already familiar to you if you follow indie narrative games at all: winner of both the Grand Prize and Excellence in Narrative at the most recent IGF, the creation of Papers, Please designer Lucas Pope. I’m only getting to it now because I’m seriously behind.
The game puts you in the role an an insurance inspector, trying to work out what happened to everyone who used to be aboard a ghost ship. Thanks to a supernatural pocket watch, you’re able to revisit the time-frozen moment of death for every body you find, allowing you to explore one disturbing tableau after another to figure out who died, in what sequence, and why.
Unsurprisingly, it’s a very good game, one that achieves spectacle, surprise, and even some comedy within its confines. It’s worth playing with as little spoiler information as possible, so I’ve put less than usual above the fold. More design thoughts follow below, but they assume you’ve either played the game or read enough of a synopsis to be able to follow references.
Return of the Obra Dinn is one of those games that commands respect even when it’s making choices I wouldn’t have made. There are a bunch of nitpicks and style disagreements I could list: finding the final identities becomes laborious and fiddly; the system sometimes demands precision about modes of death when it was hard for me to guess what was intended from the scene; I wanted the pacing to pick up at a couple of points.
Finally, in terms of plot revelation, the mid-game offers you a sequence of de-escalating monsters — showing you the vast Kraken first and the mermaids last. The crab-riders sequence is the slackest portion here, narratively. The Kraken has the merit of terror and surprise, and I was appropriately gobsmacked when I first came up on deck and saw the tentacles sticking out of the water. And the mermaid sequence, meanwhile, is much more interpersonally intense, with a complicated set of allegiances and motives to understand, and the implication of some underlying fantasy lore.
But the crab-riders section is the place where I most pulled away from the game world, feeling restive about how long the story was taking, and losing confidence in its internal logic. When viewed in forward chronological order, the story does benefit from the escalation — the mermaids come in, then something else moderately bad happens, then the Kraken arrives — but the crab bloodbath isn’t doing as much work as the other sections, narratively speaking.
Contrast, here, the Murder section. There is much less nasty spectacle than when the crab-beasts are roaming the ship, but there’s still a lot to see. The execution by firing squad is perhaps my favorite moment in the game. We’ve spent so long at this point poring over the picture of the execution that it is breathtaking to have the scene come to life; to find the artist and see him from the outside for the first time; to discover that reality is both more and less legible than the artwork that showed it. And then the rest of the section is meaty, full of information about who has done what, and what kinds of people they are.
None of those reflections significantly diminished how much I respect this game, though. Return of the Obra Dinn commits to a few systemic design choices and then adheres to them with true discipline.
One: core mechanic. You will wander around the ship; you will find physical remains, you will use your pocket-watch to view the scene in which that being died, and you will then record the identity of the deceased, the method of death, and the identity of the killer.
That’s it. Anything that does not belong to this mechanic has been abstracted away so naturally that you might not really notice its removal. There are locked parts of the ship that become available over time, but this is metered by how many discoveries you’ve already made. Pope doesn’t clutter up the game with unnecessary locks and keys, for instance, as a more literal-minded designer might have done.
Two: the ship is both setting and main character. In the course of unraveling the mystery, you will develop a highly robust mental model of the Obra Dinn: its physical dimensions, its strengths and vulnerabilities, its resources, the damage it took over time; the people who lived aboard it, how they related to each other, how they spent their time, what their duties were. Some reviewers have complained (justly) that you don’t get to know individual characters very well; it’s possible to guess at the motives of a few of them, but the reasoning often appears in quite broad strokes.
The ship’s complement as a totality, however, we can understand fairly well. At the beginning, our crew is curious, varied, diverse. The early deaths are due to accident and illness, things that might happen to anyone at sea. But then factions begin to act on less noble motives — greed, superstition, cowardice, vengeance, an insatiable desire for pretty shells. The ship turns on or destroys the people who are most other, in the process losing access to their knowledge and insight. The captain makes a series of bad choices about whom to trust and what to do.
It is a collective rather than an individual tragedy, the narrative of a compromised and damaged group.
See also:
- Andrew Plotkin’s review
- Hints and nudges guide if you’re stuck
- Return of the Obra Dinn’s TV Tropes page
- Reading and Hypothesis, an older essay of mine about games (and non-game literature) that invites us to come up with and then test our hypotheses about what has happened here
Emily Short's Interactive Storytelling 
I had other reservations. I could not shake the feeling that Obra Dinn was the second of Lucas Pope’s “historically” set games to prominently feature national and ethnic stereotype as an effective and practical strategy for recognizing people, very literally. Which crewmembers remained as you transitioned into the laborious and fiddly back third of the ledger? Even as I scoured the ship for details and angles I might have missed, I found myself making guesses based on notions of what name sounded like it might belong to whose face — and I was dismayed when that approach got results.
In general I have come to chafe at the premise, which is multiply contrived to figure as a puzzle of a certain length. This is (of course!) the draw, but I wonder whether the puzzle could have dissolved away comfortably — perhaps the ending should not been fixed behind completing the ledger? It certainly did not suit the game well to return in its conclusion to the confined pacing of its opening moments. And the implications of the details reserved to that scene in the jail are mute and dull, especially compared to the vibrant revelation of the Execution scene, or of any of the action beats.
I’m not soured to it, precisely — the labor of this thing is incredible — discipline is a good word for the choices you describe here. And I believe that its story is, yes, that of a doomed group, rather than of individual characters — that is, besides the individual represented by the player, with whom we spend more time, and of whom we know much more. I wish more like the execution scene could have been done, to highlight the difference in perspectives — memory, knowldege, intimacy, what is and isn’t communicable across time.
Even though I highly respect the game and value its original design, I didn’t enjoy it much as a player. I found it to be very linear, and most of the time I felt that I wasn’t really doing anything, and didn’t have much agency. The game took me by the hand and showed me what it wanted me to see, then I had to make something of it to unlock the next scene. I didn’t really feel that I was investigating the ship since I didn’t have the freedom to perform the actual investigation.
Didn’t you feel this way at some point?
I wouldn’t describe it as linear. While there were a few sequences that did have to be experienced in order (especially in sequences where finding one body led directly to the next), the game allowed for some aspects to be explored out of order; and it very much allowed for the player’s *understanding* of the story to develop in a range of ways, for a range of different reasons. A lot of what I was doing from moment to moment was choosing to revisit particular passages of the book or look at different evidence to try to work out who was who. It sounds as though perhaps you actually found it easier than I did, if you felt like all the necessary evidence was presented to you without much effort on your part.
Agency? Well, yes and no. You have the necessary agency to move through the ship and manipulate a few items, though the range of what’s simulated is fairly light. In breakdowns of agency, that’s what Sam Ashwell would call grasp ( https://heterogenoustasks.wordpress.com/2014/09/22/a-bestiary-of-player-agency/ ), or what Stacey Mason might call affect (https://link.springer.com/chapter/10.1007/978-3-319-02756-2_3 and discussion at https://emshort.blog/2015/05/24/framed-invisible-parties-and-the-world-plot-interface/ ).
You have no diegetic agency (Mason) or Big Decision agency (Ashwell), which is to say, agency to change the course of events, over the main storyline. Nothing you do will change what occurred in the past. (Or, if you’re doing narratology terms, you can’t alter the fabula.)
The endgame is a slightly different matter. Though I didn’t try this myself, according to websites (e.g. https://obradinn.fandom.com/wiki/Endings ), it is possible to send off the book incomplete, and to get different endings as a result — or to blame the Captain for absolutely everything that happened to anyone, though that sounds like more of an Easter egg. Does this count as narrative agency? In theory, you’re changing exactly what happens at the very end, but I think most players will discount any but the “full solution” ending as types of failure rather than types of choice, especially since there is no obvious respect in which a lesser solution is “better” or reveals different information. Still, the outcome does vary.
Then there is the ergodic sense of agency (see for instance Karth, http://manuel.boutet.free.fr/EngagingWithVideogames.pdf#page=220 ). The ending of Return of the Obra Dinn would not happen without the player as a motive force driving the story forward. In that respect, the level and type of agency involved is actually quite similar to what you find in classic old-style adventure games, in the sense that your puzzle-solving involvement is required in order to chug through to the ending; what’s more, the work you’re doing in the game is also the work required to understand the story.
This may seem like an annoyingly pedantic answer, but I think there’s some value in sorting out what we mean when we talk about agency, since people tend to use the term to mean loads of different things.
All that said, “I didn’t enjoy it” is the kind of statement I can’t argue with. And you say you felt that this had to do with linearity, a lack of freedom and a lack of agency, which can still be a meaningful observation about your experience even if I would not agree with you in applying those descriptions to the game.
My personal enjoyment slackened twice, for different reasons. Once it was because I felt like the crab story wasn’t going anywhere and my sense of surprise and curiosity was diminished; my exploration wasn’t being rewarded with the kind of new discovery that I hoped. Once it was because I wasn’t sure where to find the next body and was wandering the ship looking for clues; that was because I’d been a bit dense about where to look for one of the corpses listed in the book. So that one was arguably my own fault, though it represents a type of failure state that the game allows players to fall into.
For the most part, though, I did have fun, and the fun I had was very much associated with learning the system of life on the ship.