The checking process at a glance
We choose the reasoning skill first, build a valid answer, test the wording, and explain the key step.
What the process protects
- Fairness: the prompt gives the information needed for the intended task.
- Clarity: the words point to the same rules for careful readers.
- Usefulness: the explanation teaches a method that works again.
Start with the reasoning goal
Each challenge begins with a target operation such as constraint propagation, information gain, reframing, bottleneck diagnosis, or sensitivity testing. The story is added after the reasoning goal is clear. The puzzle design and construction learning path connects that choice to clue writing, playtesting, and difficulty calibration.
Construct the answer and clue set
For closed logic problems, we construct a valid solution state and derive clues from it. The fair puzzle design guide explains how clue contribution and solution uniqueness are checked. For open real-world problems, we define what a strong answer must contain rather than pretending there is one universal response.
Test ambiguity
We list the assumptions a reader might make and search for alternate answers that fit the literal wording. If another answer works, we either narrow the prompt or explain why the problem is open-ended. Wordplay is signalled rather than used as a hidden escape hatch.
Write the explanation
The solution identifies the observation or deduction that changes the search space. It then connects that step to a reusable lesson. Explanations should help on the next problem, not merely close the current one.
Review after publication
Reader counterexamples are treated as useful tests. A challenge can be corrected, reclassified, or retired under the editorial and puzzle quality policy. Use the contact page to report a reproducible ambiguity or error.