Share
Embed
Estimation Process
The background and theory behind the estimation practises presented in the workshop.
Jacob Chencha
April 28, 2016
Transcript
22 slides
-
What is an estimate An approximate calculation or judgment of
the value, number, quantity, or extent of something. -
Why do we estimate 1. Plan budget 2. Schedule project
periods 3. Make reasonable commitments -
Steps in Estimation 1. Estimate Project Size (Function points, Story
points) 2. Estimate the Effort (Developer hours, Man Month) 3. Estimate the Schedule (Calendar dates) -
Project size Total scope of a project. Includes breadth and
depth of the feature set. As well as the program difficulty and complexity. -
Project Effort One person's working time for a month, or
the equivalent, used as a measure of how much work or labor is required or consumed to perform some task. -
Our Assumptions 1. Non trivial program 2. Efficient use of
programming tools 3. Use of modern programming practices 4. Active risk management 5. Excellent physical environment 6. Integrated use of communication tools 7. Use of agile techniques -
Precision vs Accuracy The only good kind of estimate is
an accurate one. Failure to acknowledge imprecision is a sign of a bad estimate. -
Examples Bad Things To Estimate 1. Lines of Code 2.
Key Strokes/Hour Vary wildly between developers and languages -
Hofstader’s law Hofstadter's Law: It always takes longer than you
expect, even when you take into account Hofstadter's Law. -
Velocity Velocity is quite a simple concept. It is the
measure of the number of units of work a team completed in a given time interval. This is commonly measured in story points completed per sprint. Velocity is the sum of the estimates of delivered (i.e., accepted) features per iteration. -
Team size Optimal team size = schedule in months /
number of months in schedule Bigger projects take longer, but they also have bigger teams; and the inefficiencies associated with larger team sizes mean that effort increases disproportionately faster than schedule. -
Schedule Compression 2 Your estimate: 5 months Man Month estimate:
15 Months What client wants: 4 Months Schedule compression factor = ⅘ = 0.8 -
No shortcuts “There is no single development, in either technology
or management technique, which by itself promises even one order of magnitude [tenfold] improvement within a decade in productivity, in reliability, in simplicity.” ~ Fred Brooks -
Effects of compression compressed schedule effort = initial effort ÷
schedule compression factor Compressed schedule effort = 15 / 0.8 = 18.75 Man Months -
Maximum compression Communications and management overhead ensure a maximum ceiling
on compression. Maximum compression 0.75 (Boem 1981) -
Brooks Law Adding manpower to a late software project makes
it later Why: 1. Ramp up time 2. Communication overhead 3. Limited divisibility of tasks -
Estimation tips 1. Don’t be cornered into giving a quick
estimate 2. Keep metrics on projects so that you have a benchmark to refer to 3. Use estimates given by the developers 4. Estimate at the smallest pieces possible (Law of large numbers) 5. Include common and easy tasks on your estimates 6. Give a range rather than a point estimate 7. Don’t be an optimist (Refer Hofstader’s)