A Post-Apocalyptic Chess Thought Experiment

Lately I’ve been (re-)enjoying shows and movies like Fallout, Pluribus, Arrival, and The Gorge.
They all explore communication, inference, and meaning when shared context is missing or fragile. Along with a new membership subscription, I also started exporting some chess puzzles and discovered how Forsyth–Edwards Notation (FEN) works, which got me thinking about how much shared structure chess normally assumes.

That sent my mind in a strange direction.

Thought experiment:
Imagine a post-apocalyptic world. This is not Armageddon chess. You can still grow food, boil water, and live reasonably well. However, all cellular networks, landlines, the internet, computers, and even electricity are gone. You do have a hand-cranked induction flashlight and a small mirror.

Somewhere in the distance, you see what might be another person, but only faintly. You have no binoculars. You cannot walk toward each other, nor do you want to leave your precious encampment. The only way you can communicate is by signaling with light, most likely using short and long flashes.

Assume the only shared knowledge you can reasonably rely on is recognition of SOS, that neither of you knows Morse code beyond that, and that you both have very basic beginner-level chess movement knowledge.

From there, how could the discovery process be optimized so that you both eventually realize you are describing a chess board, and then a chess move, without assuming piece placement, notation, or even explicit agreement to play chess?

I am not asking for a finished protocol yet (no acknowledgements or error codes), but rather:

• What would you signal first?

• What patterns would you most naturally recognize and write down?

• How would you minimize assumptions while still converging toward chess?

I have my own ideas and will share them later, but I am curious how far shared cultural knowledge alone can take us.

I’d for sure just bash you over the head the first night and steal your supplies after watching you fiddle with a flashlight and some wooden figurines all day. “Idk honey he’s flashing the light again, hopping around 3 jumps while holding that small wood horse…the radiation levels here musta been through the roof to deteriorate a man’s mental facilities so quick”.

Just kidding. I liked your thought experiment… but I doubt all be the guy to solve this. How long do you think you’ve spent with this thought so far?

Ok I thought of something.

I hold up a chess piece to signal my intent to move. Then simply click out coordinates with individual light presses.

To start it all off at first interaction I’d hold up the board and the light. I’d put finger on a1 and do a click. a2 is two clicks… Then big pause and repeat clicks for numbers. So Qxc7 is I hold up queen and flash 3 times for C ,pause, and then 7 flashes. Actually pretty simple .


It was funny thinking how slow it could be by me just grabbing a piece and you flashing yes or no to each and every variant it can move before I grab thenext piece and repeat.

never mind since you can’t entirely see them. The only answer I see is you have to invent your own version of a morse code type language. Which is crazy if first thing you say after this magnificent feat is “you want to play chess?” ….it’s funny because in this scenario they are very basic beginner chess players.

Thanks for the laugh, that mental image made my day.

To be fair, someone might try to bash me over the head, but think more along the lines of The Gorge. I am very safe, very isolated, and mostly just bored out of my skull.

I have been thinking about this on and off for a couple of weeks. The part that keeps pulling me back is not the encoding itself, but how you get from a universally recognized SOS to the realization that a chess board is being described at all.

Fair point, and you are absolutely right to catch that. No line of sight, no details, no wooden horse spotting involved. At best, all you know is that something intentional is happening on the other side.

And yes, inventing a crude signaling language from essentially nothing is the whole madness of it. Which is exactly why the idea of eventually landing on “want to play chess?” from an SOS signal still makes me smile. Two beginners, no shared language, no shared symbols, just persistence and boredom doing the heavy lifting.

The real question for me is whether meaning itself can be negotiated at all under those constraints, without the two people ever seeing each other.

Alright I’m declaring it impossible, especially factoring in a human life span. I googled two things that you mentioned as parameters. The wind up induction flashlight I think kills all hope. It’s to weak. I was thinking maybe if had strong flashlights you can start by pointing your beams at constellations and use those to establish more words/signals possibly.

At this point I’m finding it so impossible that it feels like this was reverse engineered to be literally 100% impossible.… this problem is seemingly equivalent of talking about conversing with a possible alien species xxxx amount of light years away. All the same problems and limitations

Maybe…. You could use the fact their are 2 letters in SOS to create a system . Like S is 19th in the alphabet, so 19 flashes. Then 15 for for O. Hoping they catch on and can apply the system to the entire alphabet. Be slow as heck but could work.