Now that I’m not working in
There was a good turnout of just over 30 people to hear Seb talk. There was also some beer and nibbles laid on too which is always helps you deal with skipping dinner, especially when you know you’ll be frequenting the local pub afterwards.
The title for his talk comes from a misattributed quote of Disraeli, but it nicely sums up what many people feel about the whole issue of attempting to predict how long a software project, or even just a single task, will take. The initial part of the talk focused on how bad we are at the process of estimating and looked into some of the “science”, such as Planning Poker, that suggests we are better at estimating relative to some known level of complexity rather than trying to predict an absolute amount. However it seems this has now been debunked too. Hopefully the crowd-sourcing aspect still provides value, as does it being another conduit for conversation and the breaking down of stories.
We didn’t just get to listen to Seb but got to take part in a little experiment too. This involved us answering some general knowledge questions, such as when Bram Stoker was born, with a range estimate rather than a single value. The range however was supposed to be something we thought was 90% accurate, which should have lead to us getting around 9 out of 10 questions right. Of course most of us were way out even with that kind of leeway.
This nicely brought home the point of how bad we are at estimating and the quiz theme continued with us trying to answer the age old question of “how long is a piece of string”. Although again, rather than giving a measurement we just needed to compare it to the others and say which was longer. Naturally they weren’t all laid out end-to-end but looped around in different ways to make the process much harder; another simple but effective game.
There were various references to popular books that cover these topics, such as Waltzing with Bears and Impact Mapping, with plenty of quotes and snippets that intrigued me. I already own a couple but suspect a visit to my local online bookshop is imminent as there is plenty more to lap up here, particularly in the area of breaking down user stories. One particular slide with a flowchart titled “How to Split a User Story” looked most promising [1].
The final part of the talk looked at why we are often being asked to provide these estimates in the first place – what use are they to The Management that demands them? Seb delved into the area of Return on Investment (ROI), which is a fairly simple calculation, assuming you can of course quantify the “value” that a project will bring. This area of business analysis seems even murkier than the process of estimating that we are being asked to do. Presumably this was aimed at internal projects as he’d already covered estimating for external clients at the very beginning with a selection of quotes from builders.
The talk lasted well over an hour and
certainly provided plenty of leads for those that wanted to learn more. The
evening was rounded off nicely with a trip to a local pub (Bree Louise) which
had a number of real ales on tap. This was a great reminder that I should make
it into
[1] http://www.agileforall.com/wp-content/uploads/2012/01/Story-Splitting-Flowchart.pdf
16 October 2014
Bio
Chris is a freelance developer who started out as a bedroom coder in the 80’s writing assembler on 8-bit micros; these days it’s C++ and C#. He also commentates on the Godmanchester duck race and can be contacted via gort@cix.co.uk or @chrisoldwood.