Turn complex choices
into clear decisions.
Compare strategic options with transparent weighted scoring, editable trade-offs, sensitivity analysis and reproducible uncertainty simulations.
Decision command center
A transparent ranking based on your current options, anchors and normalized criterion weights.
Strategic option ranking
Higher weighted scores are better · 0 to 100
Criteria contribution
Point composition of each weighted score
Performance matrix
Edit each option's raw values. Higher/lower preference and fixed reference ranges determine its normalized score.
Scenario laboratory
Move strategic priorities and see immediately how the preferred option changes.
Strategic weighting
Weights are normalized to 100% for every calculation
One-way sensitivity
Sweep one criterion from 0% to 100%, rebalancing all others proportionally
Uncertainty intelligence
Stress-test uncertain performance inputs through a seeded, reproducible simulation, not a predictive AI model.
Monte Carlo experiment
1,200 scenario draws · identical inputs yield identical results
Each raw input varies independently using a centered triangular perturbation of ± the selected fraction of its anchor range. This is an assumption stress test, not a probability calibrated to observed errors.
Score distributions
Simulated 10th to 90th percentile utility intervals
| Option | Base | P10 | Median | P90 | Win % |
|---|
Methods and operating boundaries
Every calculation is local, inspectable, and explicitly assumption-dependent.
Each criterion has a low and high reference. For higher-is-better, utility is (value − low) ÷ (high − low) × 100. Lower-is-better reverses that result; values outside anchors are capped at 0 or 100.
Raw weights are normalized to total 100%, then utility is the weighted average of normalized criterion scores. Equal scores are ordered alphabetically for display. Weights and scales encode judgment, not objective truth.
The selected criterion varies from 0% to 100% in increments of 5%; remaining criteria divide the balance in proportion to their raw weights, or evenly if all other weights are zero.
A deterministic seeded generator draws two uniforms whose average yields a centered triangular value. Each input is perturbed independently by ± the chosen fraction of the anchor range and clipped to its anchors.
The model reveals all factors and score contributions. Check independence of criteria, double-counting, anchoring bias, and whether evidence really supports each input. No evidence verification is automated.
No tracking, data uploads, paid services, model downloads, or external code dependencies. Changes persist only in your local browser when storage is enabled. Export JSON to preserve or transfer your work.