A worked comparison

Three designs for the same road

Raising a road might get people to the hospital during a flood. It might also change how the water sits around the houses beside it. Both can be true at once, so residents and planners can try more than one design.

An illustration. Any real comparison needs data and methods that suit the question.

1. Ask a question. Name a place and a change you want to try. An assistant works out which data and which methods your question needs.
2. Review the proposal. You see the place, the conditions, the assumptions, and what will be measured. Nothing runs until you approve it.
3. Compare alternatives. Put today’s conditions next to each proposed design, on a map, in playback, and in numbers.
4. Inspect the evidence. Every result keeps its inputs, its methods, its assumptions, and its limits. The owner can open the exact data the runs used; people it is shared with see the results and their basis.
5. Revise and contribute. Change an assumption, add another design, bring evidence of your own, or argue for different criteria for judging the tradeoffs.

Two people can agree on the arithmetic and still disagree about which tradeoff is acceptable. Perspectives record what each person cares about, including the things that should not be turned into a number.

A comparison does not make two proposals equal

Putting two proposals side by side does not make the evidence behind them equal. Each one stands on its own data, methods, assumptions, and constraints, and you have to look at those separately. If nobody has the hydraulic or ground-condition data, a road simulation cannot tell you the design is safe or buildable. That takes more data and an expert who can read it.

The illustration below uses a simpler example than the road above, two service options compared by journey time, to show the shape of a recorded question. It is not a live simulation, and it is not a prediction about any real city.