01
Distinguish a score from an endorsement
A calculated score summarises the inputs supplied to a rule. It does not establish that the rule captures every feature worth considering. In a Simerit example, label the result as a weighted total under the displayed assumptions. Keep any final project decision separate, especially when a qualitative concern has not been included in the table. This makes room for a thoughtful discussion about the criteria rather than encouraging readers to treat the largest number as an automatic endorsement. The arithmetic can be correct while the model remains incomplete.
02
State the direction and scale of each criterion
Before combining scores, explain whether larger values mean more of the desired property and whether the criteria share a scale. Mixing a time in minutes with a preference rating would require an explicit transformation before addition could be interpreted. A simple teaching example can avoid that issue by supplying all criteria on the same zero-to-four scale, with higher values preferred. Even then, the spacing between scores is an assumption of the exercise. The identity should support that explanation through clear tables and labels rather than imply precision that the inputs do not possess.
03
Make revisions part of the presentation
Use the same Simerit spelling across the worksheet, scenario list and exported notes. Give each comparison a short identifier and record when its inputs change. A revised weighting can sit beside the original weighting so readers see the consequence of the change. Avoid styling a provisional result as a permanent badge. A useful visual hierarchy puts the alternatives and criteria first, the calculation second and the interpretation after that. This order lets a reader understand the comparison even if they disagree with the preference choices behind it.