Dr. Drang’s fundamental rule is “Don’t,” and he spends the post explaining why and when he makes exceptions. The insight came from Allison Sheridan’s Macstock presentation and Dr. Drang’s own work with Numbers to build difference tables and tidy data. He begins by noting what makes spreadsheets attractive: a grid of cells that act as data containers without needing names or initialization, easy click‑and‑drag functions that auto‑fill references, automatic recalculation of dependencies, and flexible styling for reports. In short, a spreadsheet is a data store, a logic machine, and a presentation tool.

Yet the very ease that draws users also breeds problems. Over a decade of reviewing Excel files from engineering firms exposed Dr. Drang to tangled IF statements, inconsistent row formulas, distant cell references, missing cell names, and occasional mistakes. Headers were often multiline, data split across multiple sheets, and each file was essentially sui generis, forcing manual cleaning. Size limits also bite: Excel caps at 1,048,576 rows, Numbers at 1,000,000, while datasets like the US baby‑name collection contain 2,181,032 entries, forcing “cheating” to drop less‑popular names.

The conclusion is that the click‑and‑drag simplicity encourages poor organization and errors as spreadsheets grow. Dr. Drang argues that separating logic from data—using a program instead of a template—handles varying sizes and avoids the “Siren song” of spreadsheet convenience. He cites the Reinhart/Rogoff controversy as a textbook case of elementary spreadsheet mistakes that had real‑world policy impact.