Before I wrote a single line of the real game engine, I printed a grid on a sheet of A4, sharpened a pencil, and tried to break the whole idea in an afternoon.
The premise of Western Approaches is that the computer knows the true ocean and shows you none of it. You rebuild the battle by hand, in pencil, on a chart beside the keyboard — dead reckoning the convoy forward, laying direction-finding bearings from shore stations, crossing them into fixes. If that plotting loop isn’t satisfying with nothing but paper and a protractor, no amount of teleprinter atmosphere on the screen will save it. So the honest thing was to test the paper first, alone, before it could hide behind a nice interface.
The rig
One printed chart. A pencil, a ruler, a protractor. And a sealed answer key: a short scripted crossing where I’d pre-computed, with a script, exactly where the convoy and the U-boats truly were each watch — the numbers I’d be graded against but couldn’t see while playing. The rule I set myself was simple and unforgiving: play the whole crossing on paper, commit to my fixes in ink of belief, and only at the end unseal the truth and measure the gap. No peeking, no adjusting. If the gap was maddening rather than motivating, the design was wrong at the foundation.
What the paper taught me
Three things, none of which I could have learned from a screen.
First: a single bearing is nearly useless, and that’s the point. One line from one station tells you a U-boat is somewhere along a ray hundreds of miles long. It’s only when a second station hears the same transmission that the two rays cross and a position appears out of nothing. Drawing that first cut by hand — watching a fact assemble itself from two directions — is the exact hit of satisfaction the whole game had to be built around. On paper, it landed.
Second: the drift has to be invisible, and it has to hurt. My pencilled convoy and the true convoy pulled apart a little every watch — current, weather, the imperfect station-keeping of forty merchant ships. I couldn’t see it happening; I could only manage it, by re-anchoring to the commodore’s noon position when it came. The moment I understood that my beautiful plot was quietly lying to me, and that keeping it honest was my job and no one else’s, the game had its spine.
Third: being confidently wrong is a different, worse thing than being wrong. Some of my fixes were off. Fine — fog of war. But the fixes I was sure about and got wrong were the ones that cost me, because I’d acted on them. That asymmetry — that certainty should be expensive when it’s misplaced — became a core scoring rule later: the game now asks you not just where the enemy is, but how sure you are, and punishes confident error far more than an honest “I don’t know.” That entire mechanic was born on a piece of paper, from the specific sting of trusting my own bad line.
The verdict
At the end I unsealed the key and laid the truth beside my plot in a second colour. The gap was real — I’d drifted, I’d mistimed one contact, I’d trusted a bearing I shouldn’t have. And it felt entirely earned. Not arbitrary, not punishing: a legible record of every judgement I’d made across the crossing, right there on one dirty sheet. I sat looking at it longer than I expected to. That was the moment I knew the game was real — not because it was fun in spite of being on paper, but because it was fun precisely because it was.
Everything since — the hidden simulation, the teleprinter, the confidence stakes, the second convoy that pulls away from the first — has been in service of protecting that one afternoon’s feeling. The screen exists to ration what you know. The paper is where the game actually happens.