Learn
How to score options so the numbers mean something
Decision scoring is the step at which judgment becomes a number: this vendor earns a 7 on support, that one a 4. Scores make a comparison possible and visible. They also carry a risk, because a number looks more certain than the judgment inside it.
Anchor the scale in words
Before scoring, write what each point on the scale means for each criterion. On support, a 10 might mean a named contact who replies within four hours, a 5 next-day email replies, and a 1 a public forum only. With anchors written down, two people scoring the same option land close together, and a later reader can see why a 7 is a 7.
Converting measured figures
When a criterion is already a number, such as price, convert it with a fixed rule rather than by feel. One simple rule gives the best value a 10, the worst a 0, and places everything else in proportion. For costs, where lower is better, the score is 10 × (highest cost less this cost) ÷ (highest cost less lowest cost).
The rule keeps a small difference from looking large. Two prices 2 percent apart should never score 9 and 4.
Score the evidence, not the impression
Each score is a claim, and it should name what supports it: a reference call, a trial, a contract term. A score resting on a sales demonstration deserves a note saying so. Where the evidence is thin, score cautiously and mark the cell, so the total does not conceal a guess.
The common mistake
Scoring after deciding. Once a team has a favorite, its scores drift upward and its rivals drift down, each by a point or so that nobody would notice alone. Across six criteria the drift can settle the result. Scoring one criterion across all the options at once, rather than one option at a time, makes the drift easier to see.
Scores in a Decision Room
In Argumont, weights total 100 before any option is scored, and the engine works the sums exactly from your own figures. The Champion for each option argues for its scores, the Challenger presses on the generous ones, and the Inspector marks the claims behind them. The Judge's Verdict states which option leads and which score, if it moved, would change the result. The Human Decision is made with the reasoning behind every score open to inspection.
Three prices on one scale
Three vendors quote $84,000, $91,000, and $120,000 a year. Under the rule, the lowest quote scores 10 and the highest scores 0. The middle quote scores 10 × 29,000 ÷ 36,000, or about 8. Scored by feel, that middle quote might easily have received a 6, because it looked expensive beside the cheapest. With cost weighted at 30, the two-point difference is worth 60 points in the total, enough to reverse a close decision.
Questions
- What scale should I use for decision scoring?
- A scale of 0 to 10 or 1 to 10, with written anchors, works for most decisions. Finer scales add apparent precision without adding real information.
- Should I score with whole numbers?
- Usually. Whole numbers on an anchored scale are honest about how precise a human judgment can be.
Bring Argumont a Decision worth challenging.
Start with a paid 60 day pilot on your own real Decisions.