Reconstruction method
This is a bibliographic and methodological reconstruction, not a reproduction of Knuth’s books, papers, programs, or archival materials. It deepens the uploaded 300-case page by reorganizing the material into a Conway/Ahlfors-style research instrument: strategy cards, formulas, diagnostic questions, failure modes, case traces, source-control layers, and worked demonstrations. Strategy tags overlap; percentages do not sum to 100%.
Core thesis
Knuth’s method is not merely “write algorithms clearly.” It is a full-stack discipline: choose a representation, make the machine model explicit, prove invariants, compute exact costs, preserve examples, publish code and prose together, and keep the artifact alive through corrections and editions.
Deep reading unit
Each case is read as a miniature research pipeline: object → representation → algorithm → trace → invariant → cost model → exercises → bibliography → artifact maintenance.
Why this style
The Conway/Ahlfors format makes method visible: cards expose strategy; formulas expose compression; ranking exposes prevalence; corpus rows expose evidence; demonstrations expose how to apply the method.
The Knuth strategy engine
Overlapping prevalence ranking
Bars show count divided by 300 cases. Since a case carries multiple strategy tags, the displayed prevalence is a method-frequency map rather than a probability distribution.
Decision tree for reading Knuth as method
300-case corpus
| # | Case | Family | Deep reading move | Strategies |
|---|