Blog

#NoEstimates – A Strange Argument

I’ve been reawakened to trying to unravel the #NoEstimates argument.

I first came across it when Glen Alleman (@galleman) of http://herdingcats.typepad.com/ fame liked my Free Estimating Course. Shim Marom’s  (@shim_marom) article http://quantmleap.com/blog/2014/02/consolidating-the-estimatesnoestimates-debate/ is a great summary & cross reference to relevant posts from other people’s blogs.

Estimating Course Entry Page

Free Estimating Course

Seems the #NoEstimate community’s issue is that most peoples estimating ability is crap so an approach that avoids being misled is to be preferred – I thoroughly agreee with that chain of thought. They then get quoted as saying “so we won’t estimate”. I’m not sure from what I’ve read that that is a fair attribution but either way its a chain of thought that I strongly disagree with. Lets educate people in what good estimates are and what good use of estimates by the roles involved is.

1) What I read is the #NoEstimates community actually says “Errm well we do need targets like dates and budget…” so their thrust is a shift in the dependant & independent variables – something DOD 5000 did in “Cost as Independent Variable” for US Defence contracts probably 35 yrs ago. The DOD 5000 series is a massive sprawl of ideas like Product Capability Trade-Spaces, Integrated Product Teams, Through-Life-Cost over Acquisition Cost, Earned Value, Statements of Outcome (to replace S_of_Work), Gain/Pain share contract types and more. If you want to be smart about commissioning & executing projects and delivering benefits then estimating capability IS necessary but discussing just estimating in isolation cannot lead to useful conclusions – We are looking at Complex Adaptive Systems (teams) within CAS (projects) within CAS (industries, Economies, Political philosophies, Government and governance et al) – Seeing the whole is necessary before attempts to discuss the components.

Two overlappping prbability density functions2) There seems from the debate to be a move towards understanding of what an estimate actually is, but there is a huge body of evidence showing many think its an estimate is a number – An estimate is a machine that each time you crank the handle spits out a collection of scales and interdependent ranges whose audit trail includes an initially growing collection of factors and later a stable set. The ranges reduce as velocity (or SPI-CPI) stabilize across sufficient data to be reliable.

All estimates underpin decisions. There are only two reasons to want an estimate 1) to trade-off between options of opportunity cost – ultimately with the only resources we truly own – our intellect and our time and second to coordinate when results depend on collaboration. Inescapably this means estimates – the froth is just what to estimate, when, how and to what degree of precision to yield accuracy that is useable in decision making.

Conclusion: where ever you stand in the #NoEstimates debate you need to know the concepts, tools & techniques, process steps and procedures because whether you think we need cost estimates or cost constraints, duration estimates or delivery constraints you still need estimates!

tu’pence worth
Simon

 

UA-34759907-1