| novelty and solution bias | The proposed change is named as a Method version before identity is established. | Begin from the admitted-Method or candidate-account branch and preserve it through the return. |
| success-story bias | One completed or favorable case becomes proof of effectiveness, fit, or worth. | Type the direct observations and state the limits material to the proposed use. |
| plan-completion bias | A scheduled trial or configured tool is reported as performed Work. | Keep WorkPlan, actual Work, supporting Systems, and direct results separate. |
| survivorship bias | Held, stopped, failed, or missing-evidence cases disappear from the trial account. | Account for each eligible case, its result or disposition, and the reason where the intended result was not obtained. |
| measurement availability bias | A visible cycle-time or usage measure becomes the improvement objective. | Start from the operating decision and qualify the measure’s subject and use. |
| rollout bias | Publication, training, access, or local use becomes organization-wide adoption. | Limit the return to the named operating use; use OPS.20 only for a separate population-level cultural question. |
| tool-change bias | A prompt, board field, model, or support change is called a new Method. | Apply the reusable-semantics identity test and retain the actual changed object otherwise. |