An optimisation problem chooses x to maximise an objective f(x) subject to constraints g(x) ≤ 0. With several objectives there is usually no single best x; the set of non-dominated options forms a Pareto frontier. To choose a single point you must add information: weights that combine objectives into one score, or constraints that set minimum acceptable levels. Changing the weights moves the optimum along the frontier. The mathematics is exact once objectives, weights and constraints are given.
Weighing in debate is the comparison of impacts that matter for different reasons: more housing versus community belonging, scientific integrity versus innovation, financial security versus personal autonomy. Optimisation thinking helps in two ways. First, it exposes dominated options: if a proposed alternative policy beats the status quo on every dimension both sides care about, the debate is over the alternative, not the trade-off. Second, it shows that among frontier options the disagreement is about weights, which are value judgments that need moral arguments (rights, fairness, vulnerability), not more calculation.
In the redevelopment motion, compare three options: (A) no redevelopment, (B) redevelopment with a guaranteed right to return, and (C) redevelopment without guarantees. Score each on housing supply, community preservation and cost to the city. If B is close to C on supply and far better on community preservation, C may be nearly dominated and the Proposition wins easily. If guarantees greatly raise cost and reduce supply, the outcome depends on how much weight community preservation receives. In the interactive, change the weights and watch the ranking change, while the dominated option never wins.
Dominated: another option is at least as good on every criterion. No weights can make it the unique best.
Weights imply that values are commensurable: that some amount of one can always compensate for a loss of another. Rights-based views reject this for some values (a person's basic rights may not be traded for aggregate gains), which is better modelled as a constraint than as a weight. Any numerical weighting on this site is an illustrative model, not a measurement of moral importance.