Software project planning is a difficult task in which even small errors can lead to large financial consequences.  As a result, there has been research conducted to assess new methods to perform software project planning.  In this research, there is not a standard error measure or evaluation criteria.  This paper seeks to test multiple algorithms from the literature using a combination of algorithms (COMBA) approach.  Different error measures are tested on the combined algorithms to see if single error measures are reliable.  The results indicate single error measures tend to be reliable, but that the dataset the error was measured on matters greatly, coinciding with a similar study done previously.  Finally, future works using a combination of algorithms approach are suggested.