Make the trade-off visible.

Give every direction the same question.

A short note, a visual card and a conversation can all communicate an idea, but they ask for different effort and attention. Compare them against the same intended result. A board containing unrelated projects can produce numbers without producing a meaningful choice.

Name two or three plausible directions. Keep the level of detail similar: a rough sketch should not compete with a fully finished production plan unless that difference is part of the decision. Describe what each direction actually includes before rating it.

Write criteria that mean one thing.

Choose a few criteria that distinguish the directions. “Useful to the reader”, “easy to finish”, “interesting to make” and “easy to learn from” are the starting example. Replace those labels when your project needs a different question. Avoid counting the same concern twice under slightly different names.

Every rating on this board points in the same direction: zero is weak, five is strong. If a criterion concerns effort or cost, give a higher rating to the easier or less costly option. Do not enter raw hours or prices into a five-point rating field. Write your rating rule in the downloaded note when someone else will review it.

Weights describe your priorities.

A priority of zero removes a criterion from the calculation. Five gives it the largest available weight. The score is the weighted average: sum of weight times rating, divided by the total weight. If all priorities are zero, the board pauses rather than dividing by zero or inventing a winner.

The calculation treats rating steps as comparable and combines them linearly. Those are modeling assumptions. The result is a description of the judgments you entered, not an objective measurement, a probability or a guarantee that the highest-scoring direction is the right one.

• Check that a higher rating is always better.
• Use the same interpretation for every direction.
• Set priorities before reacting to the ranking.
• Keep a written reason for a surprising rating.

Try to move the conclusion.

Change one important priority and inspect the result. If a small change reverses the leading direction, the decision depends heavily on that judgment. That is useful information: the next move may be a small test of the disputed assumption, rather than more decimal places in the score.

An exact tie remains a tie. The board does not silently select the first option. Two close scores can also deserve a comparison through making. Keep the criteria and ratings with the final note so the reasoning remains inspectable after the session. The board does not save itself automatically.