Culture change

Most Models Are Wrong

“The most that can be expected from any model is that it can supply a useful approximation to reality: All models are wrong; some models are useful”.

~ George E. P. Box

George E. P. Box

George Edward Pelham Box FRS (18 October 1919 – 28 March 2013) was a British statistician, who worked in the areas of quality control, time-series analysis, design of experiments, and Bayesian inference. He has been called “one of the great statistical minds of the 20th century”. He repeated his aphorism concerning the wrongness of models in many of his papers.

The first appearance (1976) reads:

“Since all models are wrong the scientist cannot obtain a “correct” one by excessive elaboration. On the contrary following William of Occam he should seek an economical description of natural phenomena. Just as the ability to devise simple but evocative models is the signature of the great scientist so overelaboration and overparameterization is often the mark of mediocrity.”

 ~ George E. P. Box

He wrote later (1978):

Now it would be very remarkable if any system existing in the real world could be exactly represented by any simple model. However, cunningly chosen parsimonious models often do provide remarkably useful approximations. For example, the law PV = RT relating pressure P, volume V and temperature T of an “ideal” gas via a constant R is not exactly true for any real gas, but it frequently provides a useful approximation and furthermore its structure is informative since it springs from a physical view of the behavior of gas molecules.

For such a model there is no need to ask the question “Is the model true?”. If “truth” is to be the “whole truth” the answer must be “No”. The only question of interest is “Is the model illuminating and useful?”.

~ George E. P. Box

What’s A Model?

The Marshall model belongs to the group of models collectively referred-to as Scientific Models.

“A scientific model seeks to represent empirical objects, phenomena, and physical processes in a logical and objective way. All models are in simulacra, that is, simplified reflections of reality that, despite being approximations, can be extremely useful. Building and disputing models is fundamental to the scientific enterprise. Complete and true representation may be impossible, but scientific debate often concerns which is the better model for a given task.

Attempts to formalize the principles of the empirical sciences use an interpretation to model reality, in the same way logicians axiomatize the principles of logic. The aim of these attempts is to construct a formal system that will not produce theoretical consequences that are contrary to what is found in reality. Predictions or other statements drawn from such a formal system mirror or map the real world only insofar as these scientific models are true.

For the scientist, a model is also a way in which the human thought processes can be amplified.”

“Models are typically used when it is either impossible or impractical to create experimental conditions in which we can directly measure outcomes. Direct measurement of outcomes under controlled conditions (see Scientific Method) will always be more reliable than modelled estimates of outcomes.”

The Marshall Model

I’ve written a number of blogs posts (plus a White Paper) on the Marshall Model, and its relationship with Rightshifting, so I’ll not repeat that material here.

How is the Marshall Model Useful?

  • Explains the fundamental source of productivity – or lack of it – in organisations generally.
  • Predicts the likely path of attempts to “go Agile or “be Agile”, embark on Digital Transformations, adopt Lean or Theory of Constraints, etc..
  • Situates a range of approaches to business productivity along a spectrum (the Rightshifting spectrum), in order of effectiveness.
  • Defines the challenge facing organisations that wish to significantly improve their productivity and effectiveness.
  • Illustrates the role of the collective psyche (within social systems).
  • Offers a way forward to higher productivity, joy, engagement and seeing folks’ needs better met.
  • Provides interventionists with insights in how to intervene in organisations seeking to improve, similar to the way the Dreyfus Model provides interventionists with insights in how to intervene in situations where individuals seek to improve their skills. (How to best adapt and adopt styles of intervention to suit where the organisation is at, in its journey towards maximum effectiveness).
  • Offers a seed for building a shared mental model of the factors governing an organisation’s relative effectiveness, as well as a means to understand the mental models typically in play within organisations.

Some time ago I wrote a post on how folks might use the Marshall Model.

Aside: Please let me know if you would value an elaboration of any of the above points.


“Truth … is much too complicated to allow anything but approximations.”

~ John von Neumann

The Marshall Model is not Truth. It is truthy, in that it has some utility as described above. It is a hypothesis, one I’d be delighted for folks to debate, dispute and discuss. Do you have, for example, your own go-to model for explaining organisational productivity? Where does the Marshall Model sit, for you, on the spectrum of “highly useful” through to “not very useful at all”? Would you be willing to share your viewpoint or hypothesis on organisational effectiveness and productivity?

– Bob

Further Reading

All Models are Wrong ~ Wikipedia Entry
Scientific Modelling ~ Wikipedia Entry
George E.P. Box ~ Wikipedia Entry
Mental Models ~ Wikipedia Entry
Models Are The Building Blocks of Science

The Marshall Plan

I guess most people, when they start a new job or client engagement, have in mind the things they want to do and see happen. Most likely, things they’ve seen or made happen in previous jobs or engagements. Along with, maybe, some things they’ve read or heard about and are minded to try out, given the opportunity. (And what better opportunity than the honeymoon period of a new job or client?)

We might choose to call this an agenda.

My Agenda

I’m no different, excepting perhaps the items that feature on my agenda:

  • Invite participation in discussing “who matters?” (with respect to i.e. the work and the way it works)
  • Empathise with the emergent community of “folks that matter” (not exclusively, but as a priority)
  • Invite folks to listen to each other’s volunteered observations, hear each other’s feelings, and explore each other’s needs.
  • Invite folks to solicit and then begin attending to each other’s requests (explicit and implicit)
  • Offer and provide support to folks and communities in their journeys

Note: I’ve not included on my agenda anything about specific actions that I myself might want to do and see happen, beyond the items listed. Specifically, although I’ve written often about strategies such as Flowchain, Prod•gnosis, Rightshifting, the Marshall Model, self-organising/managing teams, the quality of interpersonal relationships and interactions, etc., I don’t bring these into my agenda. If folks discover these strategies for themselves, they’re much more likely to understand their fundamentals, and maybe come up with even more effective strategies.

The Antimatter Principle is the only strategy I’ve regularly written about that recognisably features on my prospective agenda, and then only by extending invitations to participate in that strategy. (Note: Attentive readers may just notice the tip of the Organisational Psychotherapy iceberg peeking out from the above agenda).

I’ve reached a point in my journey where, keen as my ego is to see all my ideas (strategies) made manifest, my experience tells me that’s not the way to go for the best outcomes for the community as a whole.

As for the Marshall Plan, I believe it’s best, in the longer run, to have the folks involved (in particular, the people that matter) do their own discovery and learning. Discovering for themselves, over time – through means they also discover for themselves – effective strategies for attending to folks’ needs (often including the principles underlying those strategies). I see my role in this Plan as supporting – in whichever ways folks request, or say they need – this collective endeavour. Such support quite possibly to include actively helping the discovery and learning, whenever there’s an explicit (albeit refusable) request for me to do so.

– Bob

Further Reading

The Benefits of Self-directed Learning

A Hiding To Nothing

Most large companies are on a hiding to nothing if and when they decide they’re “going Agile” for software development. “Going Agile” can only ever deliver the outcomes these companies seek if the whole organisation is prepared to change some of its fundamental beliefs about how organisations should be run.

Sought Outcomes

What outcomes do larger compares typically seek from “going Agile” in their software development teams? Here’s a partial list:

  • A more coherent, disciplined approach to software development
  • Improved governance and oversight
  • Improved estimates
  • Better due-date performance (reliable on-time delivery)
  • More visibility into project roadmaps
  • Common standards
  • Better project organisation
  • People working “in sync”
  • Senior management confidence (in e.g. the teams’ ability to deliver)
  • Higher staff motivation and engagement

Why These Outcomes Are Unrealisable

What’s not to like in the outcomes these companies seek by “going Agile”? Although maybe not comprehensive – they lack, for example, outcomes like “joy in work”, “folks getting their needs met”, “improved flow” and “customer delight” – there’s a bunch of stuff here I could get behind.

Setting aside the observation that some of the above “outcomes” – such as “common standards” and “people working in sync” – are more solutions than needs, “going Agile”, per se, is not the answer for delivering these outcomes. At least, not within the Analytic mindset world view.

Why the Analytic Mindset is the Blocker

With an implicit Theory-X, local optima (manage the parts separately) perspective, any and all solutions attempting to deiver these outcomes through “going Agile” are doomed to undermine the very outcomes sought.

It’s likely to start well, with much interest and hope expressed by the staff. After all, who wouldn’t want more autonomy, more mastery, more purpose in their work? But as things progress, existing company policies, rules, attitudes, etc. will begin to assert themselves. To the detriment of staff morale, motivation and engagement. Pretty soon, staff will begin to question the sincerity of the management in their support for “going Agile”. Pretty soon, it will start to become apparent to anyone who’s paying attention that existing policies, rules, etc., have to change fundamentally to see the outcomes sought begin to happen.

And with declining staff engagement in “going Agile”, and reducing enthusiasm for understanding the principles necessary for making Agile successful, progress will slow to a crawl. At this point, middle-management, who have to carry the burden of making “going Agile” happen will also begin, quietly, to question the wisdom of the senior management direction. This will lead to their more often reverting to orthodox, “tried and tested” (and less personally burdensome) ways of working. And more often baulking at the effort needed to push through adoption of more Agile practices.

What To Do?

So, what’s a company to do? Most companies will not realise, or want to hear, that the Analytic mindset is fundamentally incompatible with successfully “going Agile”. So, another Agile adoption failure is in the making.

Personally, having helped various companies face up to this challenge, I’d say:

“It’s extremely unlikely that you’ll want – or even be able – to give up your existing world view. At least in the short term. So something’s got to give. And it’s probably better that you give up on “going Agile”. But DON’T give up on wanting things to be better. Park your Agile aspirations, and try another path, another solution. After all, it’s the OUTCOMES you seek that matter, not some specific – and cargo-culted – solution.”

So what might that alternative solution look like? What can an unrepentantly Analytic-minded organisation do to improve its software development outcomes?

My recommendation would be to focus on the interpersonal relationships within and between departments. Help developers understand and relate to customers (both internal and external) better. Help other folks within the organisation better understand and relate to developers.

Leveraging these improving relationships, encourage multi-party, cross-function dialogue about the outcomes sought, and what folks of every stripe can do, every day, to begin to shift the organisation’s rules, policies, structures and assumptions.

In a nutshell, be less autocratic, directive and strategic, and more democratic, collegiate and opportunistic.

And remember, I’m here to help.

– Bob

Not Obviously Wrong

What’s obviously wrong in software and product development? The list is continually changing, but here’s some stuff which was not obviously wrong ten or twenty years ago, which has recently become obviously wrong, at least to many people in the world of software development:

Obviously Wrong

  • Big batches and queues of work (aka Waterfall)
  • Utilisation (i.e. keeping resources fully busy)
  • Ignoring stakeholders
  • Big Design Up Front (BDUF)
  • Violence in the workplace
  • The daily commute

And here’s a list of stuff which has not (yet) attained the status of “obviously wrong” – and so appears in the list labelled “Not Obviously Wrong”:

Not Obviously Wrong

  • Estimating
  • Management
  • Command and control
  • Telling (ordering) people what to do
  • Leadership
  • Specialisation
  • Cost accounting
  • Projects
  • Big developments in big chunks with big groups of people
  • Ignoring the costs of delays
  • Testing
  • Demanding compliance to defined ways of doing things
  • Separating ownership of the way the work works from the people that do the work
  • Agile
  • SAFe
  • Scrum
  • Kanban Method
  • Work
  • etc.

How do items get to move from the one list to the other? (note: everyone has their own two lists, and each item moves at different times for different folks). How do your two lists look, at the moment?


Looking at this another way, the obviously wrong list above has items that, although once not obviously wrong, now appear on many folks’ obviously wrong list, having made the transition through e.g. a process of reflection, evaluation, discussion and above all UNLEARNING.

No Hashtags

FWIW, it occurs to me that we might choose to regard the raft of #No… hashtags on Twitter as opportunities to consider in which of our own – and others’ – lists the related (hashtagged) topic appears.

– Bob

Testbash Dublin and Organisational Psychotherapy

As I mentioned in my previous post, I’m just back from presenting an interactive session on Organisational Psychotherapy at Testbash Dublin. Some folks seemed confused as to the relevance of Organisational Psychotherapy to testers and the world of testing, so I’m happy to explain the connection as I see it. (And please note that many of my previous posts on Organisational Psychotherapy may also help to illuminate this connection.)

I’ll start by riffing on something Rob Meaney said during his presentation:

“Significant quality improvements [aren’t] driven by testing. They [are] driven by building relationships and influencing the right people at the right time.” ~ @RobMeaney #TestBash

Quality (and other) improvements come from improved relationships. This has been a theme on this blog for some years now. For example see: The Power of Humane Relationships.

I asked a key (for me) question during my session (several times):

“If we accept that (as per the Marshall Model) it’s the collective mindset of the organisation that determines its relative effectiveness, how do we propose to support the organisation if and when it choses to do something about its mindset?”

Unsurprisingly perhaps, I heard no answers, excepting my own proposal for a means to that end: Organisational Psychotherapy.

I wonder how many folks involved with testing ask themselves and their peers the question “How can our organisation become more effective at testing?”. Or, using the #NoTesting frame, “How can our organisation become more effective at delivering quality products and services?”


Organisational Psychotherapy is not just about improving product quality, however. Through improved relationships, and a shift in how the organisation relates to its people (i.e. from Theory-X to Theory-Y), the quality of life at work also improves. Put another way, we all have more fun, more job satisfaction, and get to realise more of our potential at work. Further, for all the folks that matter, their several needs get better met. And, as a bonus for the organisation itself, it gets to see its people more productive and engaged. What’s not to like?


I’ve also written elsewhere about using the Antimatter Principle in practical ways during software development. For example, during development we eschew requirements gathering in favour of incrementally elaborating hypotheses about the needs of all the folks that matter, and then conducting experiments to explore those needs. I can envisage teams that still need testers adopting a needs-focused approach to driving testing. For example, putting into place various means by which to answer the question “how well does our product meet the needs of the people that matter to us?”.

Practical Applications

On a related note, some folks asked me about practical applications of Organisational Psychotherapy in their day-to-day work as testers. Here’s just a few applications which immediately come to mind:

  • Improving communication with the people that matter (i.e. developers, fellow testers, management, stakeholders, customers, etc.). I find NVC (nonviolent communication) skills and practice particularly useful in this context.
  • Clarifying what works and thus what to do more of (Cf Solutions Focus). This can improve team retrospectives.
  • Helping the people that matter (including ourselves) feel better about what we’re doing (Cf. Positive Psychology).
  • Understanding each other’s strengths, with a view to having the right people in the right seats on the bus (Cf. StrengthsFinder).
  • Eliciting requirements (if you still do that) (Cf Clean Language).
  • Building a community (such as a Testing CoP or a multi-skilled self-organising product team) (Cf Satir Family Therapy).
  • Improved cooperation with higher-ups (empathy, Transactional Analysis, etc.).
  • Dealing with blockers to changing/improving the way the work works.


I’d love to hear if this post has helped put my recent Testbash session in context.

– Bob

Do Nothing That Is Not Play

If we think about it calmly for but a moment, one logical outcome of nonviolence is folks not working for a living, but playing for a living.


Many companies seem desperate for “innovation”. Is innovation more like to come about through folks doing what they’re told (“working”) or through playing with things, as they see fit? I posit the latter is far more likely.

And no, this is not some flight of fancy. Do we really embrace wholehearted the set of assumptions labelled by Douglas McGregor as Theory-Y?

Theory-Y assumptions are: (1) physical and mental effort are natural and most people (depending on the work environment) find work to be a source of satisfaction, (2) they generally, on their own motivation, exercise self-control, self-direction, creativity, and ingenuity in pursuit of individual and collective (company) goals, (3) they either seek responsibility or learn to accept it willingly, and that (4) their full potential is not tapped in most organisations.

Could it be that folks’ “full potential is not tapped in most organisations” because they are obliged to “work” rather than play?

Could it be that engagement (and productivity) would take an amazing leap forwards if we invited folks to “play” rather than “work”?

Could you have the courage to experiment to find out for yourself?

Or are we all so in thrall to what Walter Wink calls The Domination System that play as an alternative to work is undiscussable?

– Bob

Further Reading

Serious Play ~ Michael Schrage
The Importance of Play (A Valentine for Marshall Rosenberg, part 2) ~ John Kinyon
What If #7 – No Work ~ Think Different
The Human Side of Enterprise ~ Douglas McGregor
Theories of Motivation ~ Think Different

Change is a Waste of Time

Change is a waste of time. And money. And human potential.

That’s not to say the fruits of change are not sweet, but the cost of those fruits, especially given the way most companies go about change, is egregiously excessive.

Put another way, if we could spend less time, effort, money, management attention etc. on change, the waste would be less. And the fruits more abundant and sweeter.

Despite many saying they want change, few need change. I mean, it’s not like anyone’s job, livelihood, career or self-image depended on it. At least, that’s the perception. And those few who DO need change, are those that seem to be stymied at every turn when trying to introduce it. 

What to do? 

How about we take a leaf out of the Book of Therapy, and simply support the few – the few individuals, the few organisations – that feel the need to change? Maybe their improved wellbeing might then encourage a few others?

– Bob

%d bloggers like this: