Coming late towards the #NoEstimates discussion, I’m surprised about a few of the presumptions which have gone unchallenged, and I’m also surprised about the lack of some rudimentary points that nobody appears to possess made to date. The purpose of this information is to condition unambiguously things i see because the arguments in support of estimation in software development and set #NoEstimates in context.
1. Estimation is frequently done badly and ineffectively as well as in an excessively time-consuming way.
My opportunity and that i have trained up to 10,000 software professionals better estimation practices, and trust me, we view every imaginable horror story of estimation done poorly. There’s no doubt that “estimation is frequently done badly” is really a true observation from the condition from the practice.
2. The main reason for poor estimation is generally insufficient estimation abilities.
Estimation done poorly is most frequently because of insufficient estimation abilities. Wise people using good sense isn’t sufficient to estimate software development projects. Studying two page blog entries on the web is not likely to educate anybody how you can estimate perfectly. Good estimation isn’t that hard, once you’ve developed the skill, however it isn’t intuitive or apparent, also it requires focused self-education or training.
3. Many comments for #NoEstimates demonstrate deficiencies in fundamental software estimation understanding.
I do not expect most #NoEstimates advocates to accept this thesis, but as somebody who does know a great deal about estimation I believe it’s obvious on its face. Here are a few good examples
(a) Are estimation and forecasting exactly the same factor? So far as software estimation is worried, yes they’re. (Just perform a Google or Bing search of “definition of forecast”.) Estimation, forecasting, conjecture–it’s the same fundamental activity, so far as software estimation is worried.(b) Is showing someone several images of kitchen remodels which have been completed for $30,000 and implying the next kitchen redesign could be completed for $30,000 estimation? Yes, it’s. That’s an implementation of the technique known as Reference Class Forecasting.