Skip to content
Methodology

Why the Lab is different

Traditional recipe websites answer “What should I make?” CucinaLane also answers:

What changed?
Why did the bread change?
What should I test next?

The loop

Most recipes are recipe → result. Here, every bake follows:

1 Hypothesis2 Recipe3 Controlled variables4 Bake5 Measurements6 Observations7 Comparison8 Conclusion9 Next experiment

Think of it as a lab notebook and a debugger for bread. Recipes have versions; experiments start from a control; a comparison shows exactly what changed; the diagnosis suggests the next test.

Core principles

  1. 01
    Weight, not volume

    Every quantity is in grams. Cups hide a ±20% error that swamps the effects we are trying to measure.

  2. 02
    Baker's percentages

    Formulas are stored as percentages of flour, so a recipe means the same thing at 2 rolls or 20, and two recipes can be compared directly.

  3. 03
    Record the environment

    Room, flour, water and refrigerator temperatures change fermentation speed as much as the recipe does.

  4. 04
    Measure dough temperature

    Final dough temperature is the most controllable fermentation input. We calculate water temperature to hit it.

  5. 05
    Fermentation is biology, not a timer

    Times are guides. Volume increase and the poke test decide when dough is ready.

  6. 06
    Change variables intentionally

    One lever per experiment, compared against a named control, so a result can be attributed to a cause.

  7. 07
    Record failures

    A rejected experiment is data. It tells you what not to try again and narrows down what matters.

  8. 08
    Compare controlled experiments

    Every experiment shows exactly what changed from its control and what stayed the same.

  9. 09
    Separate observation from hypothesis

    What you measured, what you think caused it, how confident you are, and what to test next are kept distinct.

  10. 10
    One bake never proves causation

    A ±1 rating difference is within tasting noise. Results are provisional until they are replicated.

  11. 11
    Optimize for the eating experience

    The scores that matter: the crunch of the first bite, jaw fatigue by the last bite, and whether the roll holds a full cheesesteak.

How results are reported

Hypothesis
What we expect, and why, before baking.
Expected effect
Predicted direction for crumb, crunch, crispness, tenderness, chew, flavor and structure.
Actual result
Scores on a 29-attribute profile. Crispness and hardness are separate on purpose.
Conclusion
Supported, not supported, or confounded, judged against the control.
Confidence
Low until replicated, and lower whenever more than one lever changed.
Next experiment
The smallest test that would confirm or reject the leading explanation.

A note on the benchmark

The flagship project uses the house-baked seeded roll at Angelo's Pizzeria in South Philadelphia as its sensory benchmark. No recipe here claims to beat it until bake results say so; each is described as designed to challenge the Angelo's benchmark. CucinaLane is independent: it isn't affiliated with, sponsored or endorsed by Angelo's Pizzeria, and none of these recipes are theirs.