Calibration
Under, just right, over.
| Situation | Under | Just Right | Over |
|---|---|---|---|
| An unfamiliar term in a meeting | Lets the term pass so the meeting can move. Stays lost for the rest of the discussion. | Asks a brief question or notes it and looks it up the same day. Comes back only if it changes the work. | Stops the meeting for a full explainer. The agenda does not finish, and people stay late to recover the decisions. |
| A new library that might help this week | Ignores it and stays on the familiar path even when that path is slow. | Gives the library a short, time boxed try. Keeps the old path if the try does not clearly help this ticket. | Rebuilds the ticket around the library. The original work slips while they compare options. |
| A neighboring team's design review | Skips it because the service is not theirs. Misses a change that will land on them. | Reads the doc, attends if it touches their work, and asks a question that tests one assumption. Leaves when the rest is not their problem. | Sits in every review and debates each design. Their own tickets wait. |
| The first approach stops working | Keeps forcing the same method. Does not read, ask, or try another path. | Pauses, looks up how others solved it, and tries one different approach with a time box. Returns to the original if the detour is worse. | Drops the ticket to survey every method they can find. Days pass with nothing working in the branch. |
| A ticket outside your stack | Leaves it for the specialist. Does not open the docs or attempt a first cut. | Takes the ticket, learns the slice it needs, and checks with someone who knows the area before merging. | Uses the ticket as a reason to study the whole stack. The ticket sits while they take a course. |
| Data that does not match the plan | Talks past the numbers so the plan can stay. Does not open the raw source. | Pulls the data, names what they do not understand, and changes the plan only where the evidence is clear. | Rewrites the whole plan from one surprising chart. The team chases a new theory each afternoon. |
| A question you cannot answer in your own review | Bluffs or shuts the question down so they still look like the expert. | Says they do not know, writes the question down, and comes back after they have checked. Keeps the review moving. | Turns the review into a live research session. Reviewers sit while they read docs on the call. |
| Someone asks why we do it this way | Answers with that is how we do it and moves on. Does not check whether the reason still holds. | Explains the history, admits the parts they are unsure about, and looks it up if the answer may be stale. Changes the practice only if that check warrants it. | Turns the question into a rewrite of the team's process. Shipping pauses for a long debate. |