by Arlene Minkiewicz
| September 24, 2014
While I don’t like to admit to visiting a website entitled geekArticles.com, I did stumble across a reprint of an essay by Grant Rule “Bees and the Art of Estimating” that some of you may find interesting and instructive. The author participates in his own form of “Estimation Trivia” by posing the following challenge “Take paper and pencil and write your estimate for the number of insects in the average hive of English honeybees.” Of the approximately 1100 software measurement and process improvement professionals he has challenged thusly, only about 10 have bothered to question the underlying assumptions in this challenge such as the definition of an insect or the meaning of an average hive. And only about 8 have given a range rather than a point estimate.
Hmmm….. There’s probably some statistics about how many folks in a population of 1100 software measurement and process professionals are also experts on English honeybees. Or maybe not, but it’s a good guess that a lot of them don’t know the first thing about honeybees and even fewer are apiologists (one engaged in the scientific or systematic study of honeybees). And yet they were comfortable estimating with certainty with no questions asked.
Certainly, one must recognize that for these software professionals, answering a question about honeybees posed during a presentation at a conference or symposium or the like, clearly does not carry the same implications as preparing an estimate for an actual piece of software being developed or considered. Nonetheless – one would like to think that these software measurement and process people, when asked to estimate, would actually think like estimators.
Because you can literally do anything with software, software estimators are often asked to perform estimates for features or technologies that are nearly as unknown to them as English beehives. They need to stay on their toes and resist the urge for complacency. They need to have good tools like TruePlanning to guide them through the questions to ask and help them quantify the impacts of uncertainty.