Everyone has made one. The pizza that came out better than anything before it, that you photographed, described to people for weeks, and have never been able to make again.
The reason is not skill and it is not luck. A pizza is the output of maybe a dozen variables, several of which you did not choose deliberately, and memory keeps about three. Ask someone what their finished dough temperature was on the night of the good pizza and they will tell you what the weather was like.
A log fixes this, and it does not have to be elaborate. The one below is short enough to fill in while the oven preheats, and every field earns its place by enabling a specific diagnosis.
Why memory is not a record
Memory reconstructs rather than retrieves. It keeps the outcome vividly, the conditions barely at all, and fills the gaps with what usually happens. It is also biased toward whatever you were paying attention to. If you were focused on hydration that night, you will remember the hydration and not that your kitchen was unusually warm, that you were near the end of a bag of flour, or that the dressed pizza sat on the peel for a minute while you found the anchovies.
The uncontrolled variables are exactly the ones you do not remember, which is why they stay uncontrolled. That is the argument in why the same recipe gives different results every week, and the log is its practical answer.
The minimum fields, and what each one lets you diagnose
Each field below is followed by the question it answers. If a field on your log does not answer a question, take it off.
Dough
1. Date and time of mixing. Enables: calculating actual elapsed fermentation, rather than intended. This is the anchor for every other time on the sheet.
2. Flour: brand, type, and lot or batch code if printed. Enables: diagnosing the run of pizzas that mysteriously changed. Flour is an agricultural product and specifications vary between lots. When a dough that had been consistent suddenly behaves differently at the same hydration, the first question is whether you opened a new bag. Without the lot recorded you can never answer it.
3. Full formula in baker's percentages. Enables: reproducing at any batch size, and comparing across bakes. Grams alone tell you what you did once; percentages tell you what you did in a way that transfers. See baker's percentages for pizza.
4. Water temperature at mixing. Enables: hitting a dough temperature target next time. It is the lever you actually control.
5. Finished dough temperature, measured at the end of mixing. Enables: the single most valuable diagnosis on the sheet. Fermentation runs at the temperature of the dough, not of the room, and mixing adds an amount of heat that varies with method and equipment. If your schedule keeps producing different results, this is usually the reason. Detail in desired dough temperature.
6. Mixing method and duration. Enables: separating gluten-development problems from fermentation problems. A base that tears is often under-developed rather than under-fermented, and you cannot tell the two apart retrospectively without this.
Fermentation
7. Every stage, with a duration AND a temperature. Bulk, ball, retard, and final proof each get their own line with both numbers. Enables: everything. A time without a temperature is not a record any more than it is an instruction. This is the field people most often half-fill, and a half-filled fermentation record is why schedules do not transfer between kitchens or seasons. See fermentation is time and temperature.
8. Ball weight and how many balls. Enables: comparing bases across bakes. Ball weight changes proof rate, the thickness you can open to, and the finished diameter, so a comparison across different ball weights is not a comparison.
9. Ambient kitchen temperature. Enables: explaining the seasonal drift that makes a summer schedule wrong in winter. Also the field that lets you predict, rather than discover, that tonight's dough will move faster than last time's.
10. Condition of the ball at the bench, in your own words. Slack, tight, domed, spread, skinned, sticky. Enables: linking a stretching or launch problem back to fermentation. This is the observation that connects the dough half of the sheet to the bake half.
Handling and bake
11. Dwell time on the peel. Enables: diagnosing launch failures, which is the biggest cause of a lost pizza and the one nobody records. An honest "about ten seconds" or "about a minute" is enough. See launch failures.
12. Dusting medium used. Enables: separating a sticking problem from a burnt-base problem. Different media release differently and scorch differently. See what to dust your peel with.
13. Baking surface, its measured temperature, and preheat duration. Enables: diagnosing pale or soggy bases. The dial setting is not the temperature of the surface the pizza sits on, and preheat duration matters as much as the setting. Without an infrared thermometer, record the preheat and the setting and mark the surface temperature unknown. See deck temp versus dome temp.
14. Bake time, and position in the sequence. First pizza of the night, or fourth? Enables: diagnosing the common pattern where pizza one is excellent and pizza three is pale. Baking surfaces lose heat to each pizza and need time to recover. Without the position recorded, this looks like random inconsistency.
15. Topping load, described honestly. Enables: separating a wet centre caused by the dough from one caused by half a jar of passata. See wet vegetables on pizza.
Outcome
16. What you were trying to change from last time. One line. Often "nothing." Enables: the entire point of keeping records. A log without an intent field is a diary.
17. The result, in specific terms. Base colour, rim height and colour, crumb, underside, any gum line, any raw or wet band. Enables: symptom-based diagnosis later. "Good" and "bad" are not diagnosable. "Pale underside, well-coloured rim, thin gum line under the sauce" is. See diagnose your bake by symptom.
18. One line on what you would change next time. Enables: the next entry. Write it while the pizza is in front of you, because tomorrow you will not remember.
That is eighteen fields, most of them a word or a number. It takes a couple of minutes.
Fields that seem useful but aren't
Brand names of every ingredient except flour. Flour lot variation genuinely moves results. The brand of your olive oil across forty bakes is data you will never use.
Precise times to the minute. Fermentation does not resolve at that scale. Nearest quarter-hour is plenty and increases the chance you fill it in at all.
Humidity. Sounds scientific, and is very hard to link to a specific outcome without controlling far more than a home baker controls.
A satisfaction score out of ten. It compresses everything you need for a diagnosis into one number that tells you nothing about which variable to move.
Anything you would have to guess at. A guessed number recorded as fact is worse than a blank. Write "unknown."
Recording the outcome usefully
Photographs are the highest-value, lowest-effort part of the record, provided you take the right ones. Three shots:
The underside. The most informative photograph in pizza and the one almost nobody takes. It shows conduction: how much colour the base picked up, whether it is even, whether the dusting burnt, whether there are pale rings where the surface had not recovered.
A cross-section through the rim. Cut through and photograph the cut face. The only honest evidence of what fermentation and handling actually produced.
The whole pizza from above, for shape, blistering, colour distribution, and topping coverage.
Shoot in consistent light. Comparing a pizza photographed under a warm bulb with one shot by a window will mislead you about colour, which is one of the things you are trying to judge.
How to run a one-variable test across two bakes
The log makes this possible. The discipline makes it useful.
- Pick one variable. Hydration, finished dough temperature, preheat duration, or dwell time. One.
- Write the intended change down before you mix. Field 16. Committing in advance stops you retrofitting a story onto whatever happened.
- Hold everything else steady. Same flour, ideally the same bag. Same ball weight, same schedule at the same temperatures, same oven setup, same topping load.
- Bake in the same position in the sequence. First pizza against first pizza, not first against fourth.
- Record both bakes fully, photographs included.
- Compare specific outcomes, not overall impressions. Underside against underside, rim against rim.
Change three things and an improvement teaches you nothing you can act on, while a regression leaves you unsure what to reverse. It is the same discipline we apply to the formulas we publish, set out in how we test.
Using the log to build your own house formula
After a run of logged bakes you stop having a recipe and start having a formula with conditions attached: this flour, at this hydration, mixed to this dough temperature, on this schedule at these temperatures, at this ball weight, on this surface at this preheat.
That is a house formula, specific to your kitchen and your equipment, which is exactly what makes it work when someone else's recipe does not. Once you have one, the log changes job. It stops being an investigation and becomes a control: run the formula, note anything that deviated, and only run a deliberate one-variable test when you want to move somewhere new.
Copy the eighteen fields above into a notebook, a spreadsheet, or a phone note, in the order they are listed here: dough, fermentation, handling and bake, outcome. Nothing about the format matters. The sheet is not the point, the record is — though paper beats a phone when your hands have flour on them.
Mistakes to avoid
Building a log so detailed you stop filling it in. A short log completed every time beats an exhaustive one abandoned after three bakes.
Recording times without temperatures. The most common and most damaging omission on any dough record.
Skipping the log on a night that goes badly. Those entries are the valuable ones. Nobody ever diagnosed anything from a run of successes.
Guessing at a number you did not measure. Write "unknown." An invented figure will mislead you months later when you have forgotten you invented it.
Filling it in the next morning. You will remember the outcome and not the conditions, which is the exact failure the log exists to prevent.
Changing several things at once and then consulting the log for an explanation. It cannot give you one.
In Plain English
A bake log is a short record of the conditions that produced a pizza: the formula, the flour lot, the finished dough temperature, every fermentation stage with both a time and a temperature, the handling, the bake surface and its preheat, and a specific description of the result. Eighteen fields, a couple of minutes.
This is for anyone who has made one excellent pizza and cannot reproduce it, and for anyone whose results swing week to week for reasons they cannot name.
Fill it in as you go rather than afterwards, because the conditions are what memory loses first. Use it hardest on the nights that go badly.
Next step: before your next bake, write down the one thing you intend to change and nothing else. That single line converts a run of unrelated attempts into a comparison you can actually learn from.
Where to go next
- Back to the pillar: the full bench-to-oven sequence
- Why the record is necessary: why the same recipe gives different results every week
- Reading the outcome: diagnose your bake by symptom
- The four variables the log is built around: the four numbers that decide your crust
- The field most people never fill in: desired dough temperature
Copy the eighteen fields into whatever you will actually fill in at the bench, and how we test sets out the same discipline as we apply it to our own published formulas.