Welcome to Systematic Sudoku!
The weekly posts are suspended for the summer, to be resumed September 6. I’m leaving this overview post for those who may be encountering Sysudoku.com for the first time. It’s to let you know what this blog is about, and what you might find among its 250 posts and 20 pages.
It’s about methods by which humans can solve difficult Sudoku puzzles, with the office software that comes with every laptop, but without programs known as solvers. The blog is a narrative following the discoveries made week by week over five years. Thanks to experience before starting, and some good luck, the principles I began with have held up to produce a system of solving that is good for puzzles of every difficulty, from newspaper single – stars to monster “unsolvables”.
Over a five year period, my outlook and writing tools have naturally evolved, though the principles have not. At this point, I’m taking a posting holiday to update existing posts, add links between posts, and improve navigation pages.
Early posts deal with the basic task of finding all candidates, and marking them. The sysudokie approach incorporates the solving of easier puzzles, and avoids flooding the grid with hundreds of candidates that could have been more easily eliminated on a less cluttered grid. The marking of all candidates necessary for advanced methods is done in a manner unique among Sudoku writers. Once comfortable with this basic technique, solvers are able to do part of it mentally, adding another pleasurable challenge to Sudoku solving.
Most of the 2012 posts, are about advanced solving, which exploits logical relationships between candidates. Then the blog turns to evaluating collections of puzzles, and the writings of other authors on human solving techniques. This process continues today, as new collections and writings emerge.
Also continuing is the gradual introduction of “extreme” methods for exceptionally tough puzzles. Most of these puzzles swamp the solver with too many candidates, concealing the relationships accessible to human vision. But even these puzzles can be solved “by hand” by extreme methods based on trials. A trial is conducted by assembling a set of candidates that are true or false together in the puzzle solution. The puzzle is much more solvable, once the trial set is found to be true or false.
More experienced new readers may agree with me that much of the available Sudoku solving advice is more suited to computer solving than human solving and will discover human engineered methods here that are found nowhere else. Among them, a trial version of Sue de Coq called Single Alternate SdC, a scratchpad algorithm for locked and almost locked sets and fish, finned or not. Also new forms of wings and wing eliminations by forcing chains, a railway graphics tool for finding all XY-chain eliminations, bent naked sets, nice loop coloring, and new forms of pattern analysis. They may want to join me in a currently ongoing discussion of exocets. Also you may be pleased to know that the work of Denis Berthier is on the Fall 2016 Sysdudoku agenda.
Sysudoku Template Tools
The puzzle grids you see on this blog are made by dragging numbers of the “given” font into a ©PowerPoint presentation slide, then dragging in clues and pencil marks of a different font as the puzzle is solved. Slides are saved at significant stages to make a presentation file record of the solving. Graphic lines, curves and icons are added.
This template is available to readers, for installing puzzles in the same way. It enables you to follow the solving process step by step from a blog trace. The trace tells exactly what is done, but the reader, looking at her own detailed grid at that point, supplies the reason why it is done. There is no better way to learn Sudoku at any level.
Other templates are available to represent the logical relationships between candidates in various ways, to support advanced and extreme techniques. Fish, chains and wings are more easily spotted on these templates. Any template you see in a blog post can be obtained free by email.
The blog is accessible in several ways, via an extensive set of linked pages. Following the blog chronologically, you can “drop in” by simply scrolling down the screen, or by using the month by month link roll along the right side. Each link brings a page with the first lines of the post description of the month’s posts. Click for the full post, with comments section.
A bar menu across the top accesses the following pages:
For a Sudoku beginner, or a beginning Sysudoku reader wanting a quick overview of Sysudoku basic principles.
A glossary explaining terms used in the blog, and often citing terms used elsewhere for the same concept.
Order of Battle
A set of flowcharts recommending an order in which to use Sysudoku techniques.
Explanations of how to read, and write the two types of Sysudoku marking traces. Trace rules prescribe which new clue or locked set to follow up next. The regular trace is depth first, following up all effects of one cause, before taking up the next one. The trial trace is breadth first, following up each cause on a new list for one step, then repeating the list for one step. Besides being a tool for learning, traces enable you to analyze exactly where your human solver circuits went wrong. A special trace for trials allows trial contradictions to be documented graphically.
A breakdown of topics into pages linking to posts and more detailed pages, guiding you to posts for explanations and examples.
Illustrations of the solving templates and links to posts introducing them.
Background on the blog, and Sudent, its author. That’s me.
You’re welcome to send in a comment. I appreciate them all, but only publish those that would interest other readers.