🎯 Select · All Roles · Core Skill
The Weighted Decision Matrix
The Select step’s core tool: criteria in rows, concepts in columns, weights times scores — and a written conclusion that shows judges the team, not the spreadsheet, made the call.
🎯 What a Decision Matrix Is
A weighted decision matrix is how your team turns “we argued and picked one” into visible engineering reasoning. It sits at the heart of the EDP Select step: list the criteria that matter, weight them by importance, score every concept against each criterion, and let the arithmetic show which concept best fits the goals you defined before you fell in love with a design.
Judges read matrices closely because a good one proves the decision came from evidence. The matrix doesn’t make the decision for you — the matrix informs, the team decides — but it forces the reasoning into the open where a judge (or a teammate) can follow it.
⭐
The traceability rule. Every row in your matrix must trace back to a criterion or constraint you already wrote in your
Criteria & Constraints entry. If “looks cool” isn’t in your criteria, it doesn’t get a row. Write criteria first, brainstorm second, score third.
1 · Criteria
Copy the criteria rows straight from your Criteria & Constraints entry. Each one should be measurable — cycle time, weight, build hours, reliability — not vibes.
2 · Weights
Give each criterion a weight 1–5 by how much it matters to your strategy. Set weights before any scoring — weighting after you’ve scored invites bias toward a favorite.
3 · Anchor the scale
Define what a 1 and a 5 mean for each criterion. “Cycle time: 5 = under 3s, 1 = over 8s.” Anchors make two people score the same concept the same way.
4 · Score each concept
Score every concept 1–5 per criterion, using measurements, prototype tests, or research evidence wherever you have them. Note the evidence source next to contested scores.
5 · Multiply & total
Weight × Score per cell; sum each concept’s column. The totals rank the concepts.
6 · Write the conclusion
One short paragraph: what won, by how much, whether the team agrees, what the matrix might have missed, and the final call. Then log what you rejected in Discarded Ideas — rejected concepts with a clear lesson are rubric positives.
🧮 Worked Example — Intake, Three Concepts
Scores are anchored 1–5; weights were set before scoring. Evidence came from two prototype sessions and match-video research.
| Criterion | Wt | Claw
S | W×S | Roller
S | W×S | Flex-wheel
S | W×S |
|---|
| Cycle time | 5 | 2 | 10 | 4 | 20 | 4 | 20 |
| Reliability (jam rate) | 4 | 4 | 16 | 4 | 16 | 3 | 12 |
| Driver skill required | 3 | 2 | 6 | 5 | 15 | 4 | 12 |
| Weight | 2 | 4 | 8 | 3 | 6 | 3 | 6 |
| Build time | 2 | 5 | 10 | 3 | 6 | 3 | 6 |
| TOTAL | — | | 50 | | 63 | | 56 |
✅ The written conclusion that belongs under this table
The roller intake wins 63–56–50, driven by cycle time and driver skill. The team agrees with the result. One factor the matrix under-weights: the flex-wheel handles a tipped-over game piece better, which matters if loaders jam. Decision: build the roller intake for V1; revisit if tipped-piece pickup becomes a bottleneck at our first event. — Written by: S.K. · Witnessed by: Coach T.
❌ What loses rubric points
Roller wins 63. (Nothing else — no agreement check, no missed-factor reflection, no explicit decision, no signature. The math is treated as the decision instead of evidence for it.)
⚠
The retro-fitting trap. If the team picked a favorite first and the scores mysteriously make it win, judges can usually tell — anchored scales and evidence citations are your defense. Score from measurements and tests, and if the matrix result surprises you, say so in the conclusion. A documented disagreement with the matrix is stronger notebook evidence than a suspiciously unanimous one.
📖 Where It Lives in Your Notebook
The matrix is the bridge in a three-slide chain. Your Subsystem Comparison tables (Brainstorm step) gather the numeric evidence — one table per subsystem, three concepts each. Each comparison feeds a Decision Matrix (Select step), one per subsystem: drivetrain, intake, scoring, defensive. The winners combine into your Preliminary Design Selection. Do one matrix per subsystem — a single whole-robot matrix hides the real trade-offs.
Untraceable criteria
A row that appears from nowhere. Every criterion must exist in your Criteria & Constraints entry first.
No anchored scale
Unanchored 1–5 scoring turns the matrix into opinions with extra steps. Define the endpoints per criterion.
Everything scores 3
If all concepts tie on most rows, your criteria aren’t discriminating — replace them with ones that actually separate the concepts.
No written conclusion
Totals without a paragraph of interpretation read as “the spreadsheet decided.” Judges want to see the team own the call.
Scoring before weighting
Set weights first. Weighting after scoring is how a favorite quietly wins.
What you call a decision matrix, industry calls a trade study. NASA runs formal trade studies before committing to spacecraft architectures; automotive teams use Pugh matrices (a decision matrix scored against a baseline concept) to pick between design directions. The skill is identical: define criteria before concepts, weight before scoring, and document the reasoning so a reviewer can audit the decision years later.
🎤 Interview line: “We selected each subsystem with a weighted decision matrix — every criterion traces to our constraints entry, the scales are anchored to measurements, and we wrote a conclusion documenting where we agreed or disagreed with the totals.”
Your intake matrix totals 47 for Design B and 44 for Design A. A teammate says “so we pick B, done.” What should go in the notebook before you move on?
⬛ Nothing — the math speaks for itself
⬛ A written conclusion: the result, whether the team agrees, any factor the matrix missed, and the explicit final decision with signatures
⬛ Re-run the matrix until the gap is bigger