Personnaly, I think the variance is to large to use. As a 'project' grows in size, the complexity increases and therefore the time to complete each object also increases. Additional issues include the number and capability of the individuals working on the project. I would assume that wanting to do such forecasts arises because you have some larger applications in mind, so the trivial (address book, calculator, ... ) are not of interest here. Long ago, I did project management for software development and developed a process which served fairly well. I basically sat with the senior individuals (three to 5) who would lead the major segments of the effort and discussed the project -using a fairly comprehnsive 'technical specification' and developed a deliverables list for the project at a level which could have individual assignments for each item. This list was used (by the senior individuals) to assign eack of these items to a specific individual and THAT individual was given the task of providing the estimate of the time necessary to complete the item. The senior individuals were, of course, available to their team members to discuss any issues and help the less experienced give representative estimates. To get the overall schedual, the various tasks were then layed out in the traditional predecessor/successor requirements and adjustments were made to generate ths initial schedual and budget. I learned that different individuals had a reasonably consistient pattern of their estimnates, usually giving an initial estimate over or under the actual by more-or-less the same percentage, but that I could only reasonably establish this value for individuals where I had a track of at least three projects, so I could generally adjust the schedual and assignments to suit. Overall, I managed approximatly thirty projects over a span of about seven years using the process and only had one of them run more than 10% off track (this was the corporate "standard" for 'success).
With the project management tools available today, I would think that it would be posssible to manage projects somewhat better and achieve or better these results. To me, the important elements of the process included the individuals' self generated estimates of time/effort required and schedualing of milestones for review and evaluation of each item on a very short interval. I generally would not permit any current schedual activity to go without review with at least the individual and his senior to go any more than two weeks without at least an informal status review, and critiacl path items might be discussed on a daily basis. These 'reviews' were NOT an attempt to intimidate, but simple informational discussions to identify where and how to supoplement the effort, and particularly when the item was in jeporady.
I know this is not the handy-dandy formula you ased for, but it may be of some help.
MichaelRed
m.red@att.net
Searching for employment in all the wrong places
