Rule 2 of 13 · Chapter I — Before You Point at a Number
Assume every number you publish will be optimised by the cheapest available route, not the route you had in mind.
Why this rule exists
People under pressure do not optimise the outcome, they optimise the measured thing, and they do it with the least effort that works. The route you imagined — better work, more skill, deeper fixes — is almost never the cheapest one available. Planning as if everyone will take the expensive virtuous path means your metric is designed for a population that doesn't exist.
In practice
Before publishing, spend ten minutes thinking like someone who needs the number to move by Friday and doesn't care how: reclassification, timing shifts, selection of who gets counted, definitional edges. Assume the cheapest of those routes is the one that will be taken, then ask whether you can live with that. If you can't, change the measure or close the route before it is published, not after someone uses it.
When it doesn't apply
Where the cheapest route to moving the number genuinely is doing the work well — a tightly defined measure in a well-instrumented process — this assumption costs you a short exercise and nothing more.