Decision matrix
A matrix is a table: options as rows, criteria as columns, weights on criteria, scores in cells. The math is simple. The honesty is not.
learn
Freeze the criteria before you fall in love with an option
Criteria should be named before options are scored. Weights should reflect the actual trade, not the slide that needs to look balanced.
What it means
Too few criteria hide tradeoffs. Too many turn the matrix into mush. Five to eight is a working range for most operating decisions.
Why it matters
Hidden criteria still operate: politics, fear, a board member's hobby. Writing them is how you keep them from pretending to be analysis.
How to apply it
Draft criteria, freeze them, score, then document if a later change was justified.
A common trap
Using criteria so vague that every option can claim a ten.
How Argumont uses this
You set criteria and weights before the run. Scores are whole numbers with reasons. Changing weights after the fact should be visible.
Related pages
Decision scoring
A score says this option is stronger on this criterion, given this evidence. If you cannot show the evidence, the score is decoration.
Decision scoring
Argumont scores options against criteria you can inspect. Weights are explicit. Reasons are stored. A model agreeing with itself is not a score.
Compare options
Comparison work fails when each option is described in a different language. Argumont holds options to the same criteria, then lets you open every score.
Learn
These articles explain methods Argumont uses in the product. They are education, not a pile of keyword stubs, and not fake news with invented dates.
Set criteria in Argumont
Argumont challenges assumptions, researches evidence, compares competing approaches, and identifies the strongest decision.
Set criteria in Argumont