# The most useful line in the log just says nothing > A five-day sourdough log tracks temperature against rise, and the entry that reads NOTHING is the one that finally shows the pattern. August 19, 2026 · https://justog.club/stories/sourdough-feed-log 03.09, 21C, SLOW. 04.09, 22C, BETTER. 05.09, 19C, NOTHING. 06.09, 24C, DOUBLED IN 5 HRS, ringed in orange. 07.09, 23C, SAME AGAIN. A faint flour smudge sits across one corner, from a hand that had been kneading a minute earlier. Four of the five lines are progress. One of them is a dead end, and it's the line that made sense of the other four. ## What is this notebook page actually tracking? A starter's rise, measured against the room it's kept in. A sourdough starter is alive — flour, water and wild yeast fermenting together — and how fast it rises depends on temperature as much as anything the baker does to it. Writing the date, the room temperature and a one-word result down each day turns a living, variable thing into a record somebody can actually read back and compare. ## Why does the same starter behave so differently from one day to the next? Because yeast activity speeds up and slows down with temperature, and a kitchen's temperature moves around more than most people notice. A few degrees either way changes fermentation enough to turn a lively rise into a flat one, which is why 19C and 24C produce such different results from what's otherwise the same flour, the same water and the same starter. The keeper isn't logging inconsistency for its own sake — they're logging it because the room, not the recipe, is the variable actually moving. ## Why does the entry that says NOTHING matter more than the ones that show progress? Because it's the point where the pattern became visible. SLOW and BETTER read like a starter warming up over two ordinary days. NOTHING, at the coldest reading of the five, is the outlier that turns two data points into a trend — it's the day cold enough to stop the rise altogether, which tells the keeper more about the room than any of the days it worked. Two entries later, DOUBLED IN 5 HRS at the warmest reading confirms it, and gets ringed because it's the proof, not the discovery. ## What makes a feed log worth keeping at all? A number written down on the day that felt like nothing happened. A log with just the good days in it is a highlight reel, not a record — it can't tell a keeper why a loaf that worked in July went flat in October. Writing the failed rise down alongside the successful ones is what turns five ordinary mornings into a working theory about a kitchen's own temperature, one a keeper can actually act on next time the room feels the same. ## Does this apply to anything besides sourdough? Anywhere a slow, temperature-sensitive process gets kept honest by writing down the days that didn't work. [A cheesemaker's make sheet](/stories/cheesemaker-make-sheet) tracks the same kind of live process — pH and temperature against a decision — but there a wrong reading costs real money on cheese somebody's selling. A feed log on a kitchen counter has nothing riding on it beyond one loaf, which is exactly why nobody but the baker would ever have thought to keep it. [A pickling crock's date label](/stories/pickling-crock-date-label) is the closer match: another domestic record corrected once the food itself proved the first guess wrong, kept for nobody's benefit but the household's own. ## Isn't a day that produced nothing the least interesting one to write down? It's the reading that makes sense of every other one on the page. A log with four good days and no bad ones can't distinguish luck from a working method — every entry looks like confirmation of the same thing. The day that stalled is the control the keeper didn't set up on purpose, and it's the reason DOUBLED IN 5 HRS two days later gets ringed instead of just noted: by then the keeper already suspected why. [A freezer label that carries just a question mark](/stories/freezer-leftovers-labels) is the same shape of surprise from the other direction — the entry that looks blankest is the one somebody ends up staring at longest. ## Does it print well? Yes — a ruled page of short, regular lines sits naturally on a chest print. Charcoal ink on bone keeps it reading as a notebook rather than a graphic, and the orange ring should stay a single loose hand-drawn oval rather than a tidy printed circle — it's the one mark on the page added after the fact, once the pattern was already clear. The flour smudge wants keeping soft-edged and faint; a hard-edged smudge reads as a printing error rather than flour. ## How do I get one made? Describe the log at [JustOG](/) — the dates, the temperatures, and which entry turned out to matter. Pick a direction, drag the crop frame, see it composited on the real garment, and it's made to order and shipped. Designs other people have published are in the [shop](/shop). Five days, one ring in orange, and a smudge nobody bothered to wipe off before writing the next line.