#NoEstimatesBook and #KnowEstimating

An open comment

Vasco, Hi

Thanks for sharing your book 🙂 I’ve read Chapter 1, I’m ready for chapter 2 – your condition of supply was to provide comment – here is some.

  1. I’m unconvinced that you have sound insight.
  2. It isn’t just software that needs to learn how to estimate

Take a look at this picture:

How you choose to express your insight suggests to me that you are sure the world sees a circle and you know if they had your perspective they would realise the truth is a square.


The truth is we are all better off when we all embrace Cockburn’s oath of non allegiance ( . ‘they’ aren’t wrong, just incomplete and you are not the truth; just part of it. Wider perspective reveals more reality.

Phrases like “utopian faith*” that load emotion onto argument are shabby distractions from reasoned argument. We can all do it, it’s unhelpful; if we can’t make the point without “utopian/shabby” we should not prejudice the point with them.

(*ch 1 pg 14 last line 3rd para)

Software is only different from the rest of the human endeavor in that it’s protagonists are all young enough to believe they see a new world never seen before (btw 30 years ago I wrote pl/1 on ibm mainframes, 20 yrs ago it was K+R C on Unix.)

Software is NOT hard done by, not a special case, not subject to unique problems. There are complete equivalents in many disciplines. Software actually has some unique advantages. Heavy engineering has to design then physically construct and people can get hurt and the weather doesn’t cooperate. Software is easy compared – ironic that it’s protagonists whine more.

There are clearly components of the #NoEstimates debate that well up from challenges of current ‘best practice’. Truth is incompetence is rife. Truth is estimating is hard and expensive to do well. Truth is what is needed for competence is close to completely mis-understood and or absent. Competent estimating starts in business vision- mostly a sandy (or even just frothy) rather than rock-solid foundation.

The first and often missing insight is the difference between accuracy and precision. Since elsewhere you talk of 4% accuracy I’ll guess that either you don’t correctly distinguish a foundational perspective or you recognize and just ‘go with’ the lack of precision in popular use of language. We need to define our vocabulary if we are to communicate reliably; precision has variation within it, accuracy is binary. Not 4% accuracy but “error of under 5%”. Another concept – some errors are immaterial – not all estimates need to be of uniform precision. Precision costs. Accuracy can always be achieved instantly and at low cost if precision is unneeded.

Chapter one is all I’ve seen so far. The only conclusion I can draw from it is you are seeking to ‘prove’ a bias decided on before the analysis is done through the use of a selected set of references. The answer is NOT to ‘dis-‘ half the world. Ultimately you’ve set-off down a path to a position that is unwinnable.

The world runs on money and the people who own it want estimates. It isn’t a case of ‘just go with’. It is the sea to Canute’s decree; he showed even kings cannot command the tide. Durations and dates ARE NOT ENGINEERING they are politics. Effort is engineering. The translation from effort to duration is project – actually portfolio – management (IE Politics still and add sociology -groups- and psychology – individuals – as agents in a complex adaptive system – or actors in Latour’s network).

I do COMPLETELY agree that setting the delivery date and fitting in scope ‘a la’ DSDM’s MoSCoW is hugely more sensible in many context than the games played by deadlines. SO we must recognize that deadlines are set in the market place by irresistable forces due to cashflows and capital market processes (etc). I also completly agree with Goldratt – He pointed out that 5 + 5 = 13, everyone hides padding. So we must recognise people’s needs for self-protection in a world largly outside their own control is an irresistable force. not #NoEstimats but #KnowEstimating.

I can add 100 other ills. I can also add the solutions. Starting with “estimates are guesses” will never solve the issue. Estimates are NOT guesses; that really is a profound mis-statement and inexcusable if an attempt to mislead, inexcusable if made through ignorance or even if made as just hyperbole. #NoEstimates is a misleading label that ultimately weakens the cause as does a rallying cry to spark xenophobic flame wars.

‘Now’ is an isochron that separates greater certainty (the past) from greater uncertainty (the future). Estimation, the future, risk, uncertainty, pre-diction, intention, plans, aspiration, pre-scription, endeavour, hope, fore-casting are all one. Put another way the need to estimate is inescapable. It is the way the world works. The world works on capital. Capital markets demand prediction and proportionally reward the appearance (real, imagined, faked or simulated) of accuracy and preferably at improved precision and improved speed.

It is what we estimate that matters. It is how we estimate that matters, it is when we estimate that matters, it is what we understand an estimate to be that matters, it is how we use estimates that matters, it is our understanding of the principals of estimating that matter, it is what is communicated when estimates are provided and received that matters, it is maturity of estimating capability that matters, it is the role and duties of an estimate’s giver and recieve that matter most. #KnowEstimating

I guess I could reinforce each ‘nicely selected quote’, BUT equally I can counter each. I’d like to see chapter 2 and I’ll be happy to share some insight.

I also have a draft book. I suggest its chapter 28 deflates the #NoEstimates hash tag in under a couple thousand words. #NoEstimates is a neat short-hand attention grabber but what we all really mean should be #KnowEstimating for what it is. The draft manuscript is on the link – The title is my publisher’s choice and NOT mine! Indeed I railed against it!!

Once again thanks for sharing

With debate we can drive the state of the art forward – estimating is one broken part of a bigger broken and misnamed topic.