the drive for precision in estimates (EAC) > Earned Value Management (EVM) User Group Forums

Unanswered Active Topics

Forums Search

Forums > EVMS Forums > EVMS Theory

Subject: the drive for precision in estimates (EAC)

Prev Next

You are not authorized to post a reply.

Author

Messages

waldengr

Posts:11

01/08/2010 2:30 PM
Over a long career I have learned that numbers carry unwarranted authority simply because an idea is represented by a designator (number) that implies all subjectivity has been removed from a proposition. Likewise, trend analysis tends to gravitate toward four decimal point precision, as if numeric precision will cause an otherwise "how goes it" trend indicator to become a safe, simple, binary decision-maker. Otherwise stated, it is all too common to mark EV performance boundaries with a crayon, cut with an axe and measure the result with a micrometer. Estimates and number theory seem extremely compatible on the surface (just get enough data points and you can remove all uncertainty). An experienced mathematician pointed out to me that the more data you have, the more confirmation of the direction and perhaps velocity of a pattern. HOWEVER, more datapoints do not mean that one can predict the very next datapoint with confidence approaching certainty (pure math aside). Example? The DIA collects and reports more datapoints about a stock than one can stand. And all you can learn from all that data is the general direction of stockmarket activity. No one has produced a formula that will guarantee knowledge of the precise opening and closing values of any single stock or mutual fund. And so it is with EV predictions and estimates. Over the last two or three years there seems to be a large movement toward using and improving EV analysis as a financial control tool. That is, EV data is being sliced and diced in a number of experiments and research pagers that demonstrates EV is becoming an attractive means to allow money movement in and out of a project on a daily basis. One example is attempting to predict daily use of money such that corporate treasury can reduce the amound of funds available to a project in order to leave the money in the treasury account to earn interest. The demand for precision in this instance is acute as it is necessary to be wholly confidently that a precise amount certain can be manipulated in and out of allocation to a project on dates certain, remain in treasury for a pre-determined number of days acruing a precisely predictable amount of interest. In short, there are numerous instances trying to use an axe for microscopic surgery. Companion to this drive to make trends accurate four digits to the right of the decimal is the insistance that EV can and will, by its very presence, ensure zero cost overruns or schedule slips. The result is that EV is manipulated so as to point to a predetermined result, rather that as a tool for the project manager to assess the overall direction of the project. EV has become a result in itself, rather than a pointer to significant trends in a project cost/schedule line. While it is fun and probably career-enhancing to play with numbers and formulae, we seem to be in danger of EVM losing its credibility because EV principles are being pushed and shoved into arenas for which EV is not designed. Eventually, disappointment in the number of malformed EV programs that produce poor results will likely result in EV being set aside for "the next killer app". Are EV practicioners too often guilty of promoting the mis-use of the tool? Are we spending to much time and writing about the numbers, while simultaneously downplaying an immutable law that EV can only be properly applied to and depended upon in a business environment where the truth about project execution is valued over the attractiveness of the chart?

You are not authorized to post a reply.

Forums > EVMS Forums > EVMS Theory > the drive for precision in estimates (EAC)