Scorecard — System Design Mock
Copy into sessions/<date>-<id>-<slug>/scorecard.md and fill. Mirrors
guides/interviewer-guide.md §7.
| Candidate | |
| Date | |
| Question | (bank ID + title) |
| Calibrated to | (company · level — senior · staff · principal · em · senior-em · director) |
| Kind / domain | (Build · Review · Strategy / domain tag) |
| Attempt | (N on this question · N overall) |
| Mode / pace / difficulty | |
| Elapsed | (actual, from the shell — not the budget) |
Scale: 1 below bar · 2 approaching · 3 at bar (hire) · 4 above bar
Every score carries evidence — a verbatim quote or a
§10timestamp. A dimension that was never probed is unscored (—), never a default 2. A wall of 4s means the interviewer under-probed and the sheet should say so.
Scores are against the target level’s bar. Weights (H 1.5 · N 1.0 · L 0.5) come from
levels.md §2 for that level.
| # | Dimension | Weight | Score | Evidence (quote / timestamp) |
|---|---|---|---|---|
| 1 | Requirements & scoping | |||
| 2 | Estimation (drives decisions) | |||
| 3 | High-level design | |||
| 4 | Technical depth (deep dive) | |||
| 5 | Trade-off reasoning | |||
| 6 | Operations / reliability / cost — manager bar | |||
| 7 | Communication & collaboration | |||
| 8 | Handling ambiguity & pushback | |||
| 9 | Leadership signal — technical (IC) or org/delivery (manager) |
Weighted score = Σ(weight × score) ÷ Σ(weight), scored dimensions only: ____ / 4 (show the arithmetic)
Outcome at the target level (levels.md §4: ≥3.25 strong hire · 2.75–3.24 hire · 2.25–2.74 lean
no · <2.25 no hire; a 1 on any H-weight dimension caps it at lean no):
Help needed
| Rung | Count | Where |
|---|---|---|
| (a) re-ask the requirement | ||
| (b) narrow the area | ||
| (c) name the tension | ||
| (d) offer a binary choice | ||
| floor hit — could not proceed without the answer |
Only (a)-level nudges → scoring well. Needing (d) to move at all → not at bar.
Overall help: ☐ none ☐ light nudges ☐ frequent steering ☐ interviewer led throughout
Time management
☐ finished cleanly ☐ rushed the end ☐ ran out of time ☐ finished early with room to spare
Where the time actually went (from §10 timestamps, vs budget):
| Phase | Budget | Actual | Note |
|---|---|---|---|
| 1 Requirements | |||
| 2 Estimation | |||
| 3 High-level design | |||
| 4 Deep dives | |||
| 5 Wrap |
The curveball
What was thrown, when:
Reaction — calm/frozen/defensive · revisited only the affected part or rebuilt everything · did they say what it cost:
The delta vs the cue sheet
In the checklist, not raised by the candidate — and, honestly, which of these actually mattered versus which are checklist noise:
| Item | Mattered? | Why |
|---|---|---|
Raised by the candidate, not in the checklist (these are often the 4s):
-
Verdict
Top 2 strengths, each with a quote:
1. 2.
Top 2 things to improve, each with the moment it showed up:
1. 2.
One drill for next time (exactly one, small enough to do this week):
Overall: ☐ strong hire ☐ hire ☐ mixed / lean no ☐ no hire
Apparent level: ☐ senior ☐ staff ☐ principal ☐ em ☐ senior-em ☐ director — and in one line, where it differed by dimension (e.g. “staff depth, senior framing”)
Self-assessment accuracy (what they said at debrief step 1 vs what the sheet says — a strong signal in its own right): ☐ accurate ☐ too harsh ☐ too generous
Interviewer self-audit
Honesty here protects the value of every score above.
- Leaks (components named before they named them, praise given mid-round, leading questions):
- Probes I should have asked and didn’t:
- Did I under-probe or over-help?
- Dimensions this session genuinely could not measure: