The three methods outlined in are the scattergraph method, the high-low method, and the method of least squares (regression).

For this discussion, I'd like you to tell me which method you think is most practical for use in "the real world" for purposes of estimating and budgeting for future costs. Is there a method that might be more objective but might end up being less reliable? Why might this be? Alternatively, is there or should there be room for subjectivity in estimating future mixed costs?