Show Me the Numbers Designing Tables and Graphs to Enlighten (Stephen Few)（Z-Library）

【One-Line Pitch】 A practical, example-driven guide to designing tables and graphs that tell the truth clearly, aimed at anyone who has to turn data into decisions—analysts, managers, report builders, and students of data communication. 【Book Arc】 - **Opening (~0%–10%)**: Frames the core problem—most quantitative presentations are poorly designed, sometimes to the point of misinformation—and previews the book's teaching method: learn concepts, then apply them through exercises and real-world scenarios. - **Early (~10%–30%)**: Builds the statistical foundation for display choices: averages, spread, standard deviation, correlation, and the relationships quantities can express (ranking, ratio, correlation), so you know what a number set is actually saying before you draw it. - **Early–Middle (~25%–40%)**: Establishes the table-versus-graph decision rule—tables for look-up and precise one-to-one comparison, graphs for seeing patterns across a set—and covers table structure (unidirectional vs. bidirectional layouts, summary plus detail). - **Middle (~38%–55%)**: Moves into visual perception: how sensation and perception work, preattentive attributes like color hue and intensity, and Gestalt principles (enclosure, closure, continuity) that explain why some designs group and read effortlessly while others fight the eye. - **Late (~55%–90%)**: Applies perception theory to concrete graph and table design—choosing the right graph for the message, encoding values so they can be compared accurately, and stripping away decoration that distorts or distracts. (Excerpts do not cover the specific chapter-by-chapter sequence here.) - **Ending (~90%–100%)**: Closes with two newer chapters—"Silly Graphs That Are Best Forsaken" (donut, circle, unit, and funnel charts) and "Telling Compelling Stories with Numbers"—plus Excel appendices for table lens displays and box plots. 【Key Takeaways】 - **Poor design is a communication failure, not a data failure** (Opening): even excellent information is useless if its story is badly told, and almost no one has been trained to design tables and graphs well—this book exists to fill that training gap. - **Match the display to the use** (Early): ask whether the information will be looked up as individual values (use a table) or examined as a whole for patterns (use a graph); this single question resolves most format debates. - **Tables win on precision, graphs win on pattern** (Early): tables encode values as text we read directly, support multiple units of measure, and combine summary with detail; graphs trade that precision for instant intelligibility of large data sets. - **Understand the statistic before you visualize it** (Early): sums and means can hide radically different distributions—the two-warehouse shipping example shows identical sums, means, and medians masking very different variation, which is why spread and standard deviation matter. - **Perception is not neutral** (Middle): preattentive attributes like color intensity and shape are processed instantly, but combining two attributes forces slow, attentive processing; distinct values must be far enough apart to actually read as distinct. - **Gestalt principles govern grouping** (Middle): enclosure, closure, and continuity mean borders, fills, and layout create groupings—sometimes illusory ones—so design deliberately rather than accidentally. - **Some chart types should simply be abandoned** (Ending): the second edition singles out donut charts, circle charts, unit charts, and funnel charts as current misuses best forsaken. - **Numbers need narrative** (Ending): before presenting quantitative information you must uncover its story, then tell it in ways that help others understand—well-told statistical stories have identifiable characteristics. 【Reading Tips】 - **Deep-read the perception chapters (~38%–55%)**: this is the conceptual engine of the book; skimming it will make later design rules feel arbitrary. - **Work the exercises rather than reading past them**: the book deliberately invites you to struggle with design problems before showing one possible solution, which is where the learning sticks. - **Skim the statistical refresher if you're already fluent** (~10%–30%), but pause on the warehouse variation example—it is the clearest demonstration of why summary statistics mislead. - **Treat the book as a shelf reference**: the author intends it to be pulled down occasionally, not read once and shelved. - **Use the Excel appendices hands-on** if you build reports in Excel; they turn table lens displays and box plots from concepts into artifacts. 【Coverage Limits】 This guide is based on stratified excerpts covering roughly the first half of the book in detail, with only table-of-contents-level visibility into the later graph design chapters and the two new second-edition chapters; specific design rules, figure examples, and chapter titles from the second half are not covered here.