Build a candidate scorecard that stops a 4-person panel from scoring the same interview four different ways

You are building a candidate scorecard for a specific role, to be filled out by each interviewer independently right after their interview slot, before the panel debrief.

ROLE BEING HIRED FOR
Staff Backend Engineer, Payments team

COMPETENCIES THIS PANEL IS INTERVIEWING FOR
System design under real constraints, debugging a live incident scenario, mentoring/code review judgment, and communicating a tradeoff to a non-technical stakeholder.

WHO INTERVIEWS FOR WHICH COMPETENCY
Priya covers system design, Marco covers the debugging scenario, Dee covers mentoring and communication.

PAST SCORING INCONSISTENCY YOU'VE SEEN
Whoever interviews last in the day tends to score a full point lower than whoever interviews first, regardless of the candidate.

Build the scorecard as one row per competency, each with a 1-4 scale (not 1-5 — force a lean rather than a comfortable middle score) and a one-line behavioral anchor for what a 1, 2, 3, and 4 look like specifically for that competency in this role, not a generic scale reused across competencies. Every anchor must describe an observable behavior or answer quality from the actual interview, not an inference about the candidate's character — "gave a concrete example with a measurable outcome" is scorable, "seemed smart" is not. For every competency, add a required text field prompting the interviewer to write the specific evidence (a quote, an example the candidate gave) that led to their score, since a number with no evidence is unusable at debrief and is the exact failure mode described in the past-inconsistency note. Include one field at the end for a clear hire/no-hire lean and a one-sentence reason, separate from the competency scores, because a panel debrief needs the interviewer's actual recommendation, not just an average of numbers that can mask a single disqualifying red flag. If the past-inconsistency note describes a specific pattern (e.g., one interviewer always scores high, or scores drift based on interview order), add one line of guidance directly on the scorecard addressing that exact pattern.

WHAT NOT TO DO
Do not create a single generic "culture fit" line — if culture or values matter to this hire, name the specific observable behavior that represents it (per the competencies list) rather than an unscorable catch-all that tends to be used to justify decisions made on other, unstated grounds.

OUTPUT FORMAT
A fillable scorecard: one section per competency (name, 1-4 scale with the four behavioral anchors, evidence field), then a final Hire Lean section. Keep it to something an interviewer can complete in under 5 minutes right after the interview ends.