In all practical problems associated with the selection and assessment, almost always criteria have different importance. The priorities of criteria are expressed by weighted coefficients. Every decision maker has to establish criteria priority scale. This can be done directly - degrees of criteria importance defined by judgment of experts, or indirectly - degrees of criteria importance calculated by alternatives themselves. Many multi-criteria decision methods allow some kind of compensatory between criteria. The low performance on an important criterion can be redeemed in overall aggregation by good performance on few other less important criteria. In this paper, we present the method which provides that such an important piece of information must be preserved: If an alternative does not satisfy a dominant criterion, then its overall aggregation value is zero.