(If you’re super spoiler-sensitive, probably don’t read this, just play the game? Though if you’re only mildly spoiler sensitive, it should be fine, I’m just going to talk a bit about the approach to problem solving in a couple of places.)


Liesl and I played Return of the Obra Dinn together. It’s a good game to play with somebody, I think: 90% of the time, we had the same ideas, but there were several situations when she noticed something that I didn’t, and I can easily imagine that that would have led me to bang my head against the game for hours longer if I’d been playing it alone.

Which is, of course, not unique to this game: it’s something that happens with logic games or puzzle games or point-and-click adventure games. Maybe the point there is that, in games like that, the distinction between single-player and multi-player games isn’t so clear as it is in action games: if you’re trying to understand an environment or a context, then there isn’t so clear a link between the number of players that the control affordances suggest and the number of players that can best enjoy playing it together? (I had a similarly positive experience playing Her Story with Miranda and Liesl, talking about what the scenes meant and coming up with ideas for words to try.)


Obra Dinn certainly has good puzzle design, and I’ve gotten more impressed with that design as I’ve thought back on the game. When we went through the scenes in the game the first time, we basically ended up identifying one trio of characters per chapter; I assume that’s by not an accident, and it meant that we always were making progress. Though sometimes that progress felt substantial (the introductory chapter, where you can identify almost every death), while sometimes it felt much more minimal. (If a chapter has a dozen or more fates, then only solving three doesn’t feel like a success!) So at that point we’d seen everything, it felt like we’d only solved a few of the deaths (and we certainly hadn’t solved half of them, maybe a third?), and it wasn’t at all clear how we’d solve the rest.

At that point, my assumption was: we’d spend one more evening trying to figure things out, and then we’d stop. And that would have been totally fine: Obra Dinn was already a neat game, and it’s a relatively open-ended one, so I respect a game that leaves it up to you how much you want to engage with it, that feels satisfying to leave at different spots in it.


And, for most of that next evening, that was still my working theory! We took harder looks at scenes, we filled in information more thoroughly than we had before, and we managed to chip away; but I still didn’t see how we were going to make it past even identifying half of the people on the ship.

But, while doing that, we were getting a better feel for the game; and clicking on the question mark next to pictures pointed out some more useful routes for investigation. (As a side note: I’m really impressed by how the game didn’t foreground that information in a tutorial, it just had the question mark sitting there for us to click on when we happened to notice it, and that happened at the right time!)

We ended that session with two realizations. One was that we were actually fairly close to identifying all of the officers / one-off people, so we could switch to a process of elimination style approach there. The other was that Liesl noticed something that gave us a clue to the fate of the four people who apparently left on a boat. So, with those, we had an active reason to keep on playing: the game continued to be doling out discoveries to us.

And we did indeed manage to succeed in identifying all of the officers in our next session: a bit of a back and forth, enough to make us feel like we were figuring things out, which ended up getting us over a hump, to where it felt like we really would be able to identify everybody.


We had one more session after that, where we did a further divide and conquer and identified all of the Topmen and then all of the Seamen. And, honestly: for some of those, we just guessed! There were a couple of instances where we could only identify people up to their role / ethnicity but hadn’t yet figured out their name; rather than going over the scenes yet again to see what we’d missed, we tried things a couple of different ways to see what unlocked.

Which is another impressive aspect to the game: it wouldn’t have been as good if you could brute-force single characters at a time, but some level of brute-forcing is going to be necessary. For example, Liesl and I had multiple discussions about whether certain deaths counted as being spiked or as being speared – ultimately, we ended up just trying some people both ways.

And the way the game gave you credit for identifying sets of three people was a great way to balance that tension: no single-death brute forcing, but if you’re in a situation where you’ve accumulated a few people that you’re 80% confident of, then you’ll be able to learn which ones are right, or if you have two guesses for somebody’s name or death, then you’ll be able to figure that out.


We actually didn’t figure out all of the deaths: we got to 56 of the 58 non-epilogue deaths, and while we knew who those two characters were, we just couldn’t figure out how to describe their deaths. So we looked those two up; and that was also the right choice, we would have been banging our head against those two for quite a while otherwise.

But that was okay, too: the game didn’t feel any less satisfying because of that. Which is another data point on my claim above: Obra Dinn feels to me like it’s designed in a way to allow you to leave at a wide range of times while still feeling satisfied.


I could go on: pleasantly different art style, interesting contrast of the static nature of the death scenes versus the dynamic nature of the regular navigation of the ship versus the audio of the death scenes. Really neat game: I’ve never played something quite like it, and it’s put together very well.

Post Revisions:

This post has not been revised since publication.