It is one basic assumption of the FACTORFINDER-Model that all factors are non-compensatory related. This is implied in the concept of the deviation in conjunction with the threshold value of the target criterion. Ideal values can not compensate for goal-adverse values in other factors. The influence of non-ideal values remains hidden as long as the threshold of the target criterion is surpassed and the variant is acceptable.
From alternatives with non-ideal values in some factors the impression may be derived that favourable values in some factors can compensate for adverse values in other factors. That is a consequence of the threshold value, which lets even imperfect alternatives look acceptable. Each success looks as successful as if it were perfectly ideal. That gives the impression of compensation between factors.
Deviations from the ideal - such is the assumption of the model - add up. That holds true even for minor deviations that itself would not let the variant fail. Ideal values can not make good for non-ideal values in other factors. Deviations in subcritical factors may add up to a more than critical deviation from the ideal. Their impact acts nonetheless and turns those alternatives inacceptable. Deviations remain hidden as long all non-ideal values are not important enough to let the variant as a whole fail. Only in the end we can judge whether all non-ideal factors were relevant enough to produce failure.