centerwhether

Centerwhether is a decision-making principle in decision theory and operations research for selecting actions under uncertainty. It focuses on the central tendency of the outcome distribution for each possible action and prefers the action whose central outcome most closely matches a predefined target. The approach prioritizes stability and predictability alongside traditional measures like expected value or worst-case minimization. Centerwhether has been discussed in economics, engineering, and AI planning.

The term blends center, a central-tendency concept such as the mean or median, with whether, reflecting the binary choice of action based on fit to a goal.

In practice, for each candidate action, the distribution of outcomes is modeled. A center value is computed (mean or median), a target is set, and a loss is assigned from the distance between center and target. The action with the smallest loss is chosen. Some formulations also include dispersion or downside risk.

Applications include pricing under demand uncertainty, project selection, and routing under uncertain travel times. Example: if two pricing options yield different average profits, the option with an average closer to the target profit is preferred, especially when risk is controlled by an added term.

Critics argue that the method depends heavily on the choice of center, target, and loss function, and can overlook tail risk or volatility. In practice, centerwhether often overlaps with existing criteria such as mean-variance analysis or robust optimization.

See also: robust optimization, decision theory, target-based decision making.