Does the level of technical documentation in a recipe actually affect whether you make it again?

Documentation quality is a real variable but I think it operates indirectly. The specs themselves don't drive repeatability, but they shape how you interpret a failure, and that interpretation is what actually determines whether you try again. A well-documented method that underdelivers still gives you a theory about what went wrong. That theory keeps the door open. For five-ingredient work specifically, I'd argue the documentation threshold is lower than people assume, because the structure of the recipe itself forces transparency. With only a handful of components, there's limited complexity to hide behind. A loosely written five-ingredient recipe is often still recoverable because the relationships between components are visible and learnable quickly. You can reconstruct the logic after one run in a way you genuinely cannot with a more layered dish. On the retroactive rating issue, I think most people conflate these without realizing it, and that's fine as a practical matter. If your second attempt is better because you corrected something, the recipe still gave you a workable foundation to build on, and that's worth something. The more interesting case is when your second attempt is worse precisely because you tightened it up. That outcome suggests the original method had productive looseness built in, and that's actually useful information about the recipe's character rather than a flaw in your evaluation.