The variable nature of the trade-offs in this example does not mean the weighted sum of scores method will not work!  If we change the measurements (in a nonlinear fashion), for example by using gallons per mile instead of miles per gallon for the efficiency attribute, these lines will become considerably straighter.  Often it is possible to find simultaneous transformations of both measurement systems that turn the preference curves into a family of parallel straight lines.  When this occurs, provided that the new measurements are used in the formula, a weighted sum of scores method will capture this set of preferences.
	Thus, the first thing we try to do is determine whether a mere change in how some attributes are measured will allow us to formulate a simple score function.