The useful result
Reproduce a weighted score and keep tied options tied until a real criterion changes.
- Reproduce a weighted score and keep tied options tied until a real criterion changes.
A decision matrix becomes useful when its assumptions are visible. Give Astra criteria and scores you supplied, then ask it to show arithmetic rather than present taste as objective judgment. The trick is to include a deliberate tie: it reveals whether the assistant follows your rules when the table does not produce a convenient winner.
Select the named model only if it is available in your interface. API controls and consumer-app controls differ. You can practise the technique on the supplied fictional inputs with an available compatible assistant first. This lesson is documentation-based; examples are expected results, not a claim that we ran the named model.
Before you start
- A fictional comparison with no private supplier information.
- Weights: fit 50%, support 30%, portability 20%; each score uses a five-point scale.
Review evidence
Documentation-based review · 2026-10-18
Provider documentation checked on 18 October 2026. Examples and checks are original illustrative exercises, not recorded platform tests. Cover artwork is conceptual and does not imply provider sponsorship.
Supply the scoring contract
Define what a score of one and five mean for each criterion. For practice, use A=(4,2,5) and B=(3,5,3) in fit, support and portability order. State that scores are fictional and do not describe real vendors.
Example / expected result
Illustrative example; not a recorded model result.
A: 4×0.5 + 2×0.3 + 5×0.2 = 3.6.
Check your result
Weights total 100% and all scores use the same direction: larger is better.
Ask for visible arithmetic
Request a table with every weighted component, total and unresolved trade-off. Prevent the model from adding a hidden price or brand preference. It may suggest a new criterion, but it must not silently apply it.
Compare fictional options A=(4,2,5) and B=(3,5,3). Weights for fit/support/portability are 0.5/0.3/0.2. Show components and totals. Keep ties; do not introduce new criteria. List one question that would resolve the practical trade-off.Example / expected result
Illustrative example; not a recorded model result.
A=3.6; B=3.6. Tie. Ask whether stronger support or better portability matters more for this use.
Check your result
B also totals 3.6; any asserted sole winner is unsupported.
Review the decision rather than the prose
Recalculate the totals in a spreadsheet or by hand. If you change weights, version the table and explain why. Keep a missing score blank until evidence is available rather than filling it with an average.
Check your result
A reviewer can reproduce both totals and identify whose preference changed the decision.
When it goes wrong
A weighted table does not make subjective scores factual. Avoid real purchase recommendations without checking current specifications, terms and prices. Tiny score differences can be smaller than the uncertainty in the ratings.
Make it your own
Change portability to 30% and fit to 40%, keeping support at 30%. Recalculate: A becomes 3.7 and B stays 3.6. Explain that the result changed because of your preference.
Useful next resources
Sources and review
Last checked Oct 18, 2026. Features and account limits can change. Verify important results against your original material.
Did this help you complete your task?
Tell us what needs improving. No account or private output required.
Report incorrect information or a broken linkGenius Forges Editorial
Practical AI lessons with checked sources, labelled examples and human verification.
