More than a weighted average
The matrix first converts every criterion to the same 0–1 scale. A criterion marked Higher is better gives the highest observed value 1 and the lowest 0; a cost or delay criterion marked Lower is better reverses that scale. The normalized values are then multiplied by the normalized weights and added.
This keeps euros, days, percentages and subjective 1–10 scores comparable without pretending they use the same units.
Sensitivity analysis
A ranking can look precise while depending on one fragile assumption. For every criterion, the tool varies that criterion's weight from 0% to 100% while keeping the other weights in their existing proportions. The result shows the weight range over which the current winner remains first and which alternatives can take over outside that range.
If a tiny weight change flips the winner, treat the ranking as sensitive rather than definitive. If one option stays first over a wide range of plausible weights, the result is more robust.
Dominated options
An option is marked as dominated when another option is at least as good on every criterion and strictly better on at least one. A dominated option can never become the best weighted choice while all criterion weights remain non-negative, so spotting one is often more useful than another decimal in the final score.
What the score does not decide for you
The calculation is only as good as the criteria, measurements and weights you enter. It does not turn subjective judgments into objective facts. Use the sensitivity view to expose where the conclusion depends on assumptions, and keep non-quantifiable constraints outside the score if forcing them into a number would be misleading.
Is my data uploaded?
No. The matrix is scored in your browser, so nothing you add is sent to gratistools.be. See the privacy page for details.