A Review of “The Undoing Project: A Friendship that Changed the World”, by Michael Lewis, 06 December, 2016

If only it had changed the world!

If you have ever:

  • Gambled – not just casino-style, but on a hunch, a business idea, a theory, a project etc.
  • Interviewed for staff
  • Negotiated anything from some money off at a store to a £multi-billion contract
  • Tried to weigh the odds of one event happening versus another
  • Are involved in any legal “encounters"
  • And many other life situations requiring a decision to be made too numerous to mention …

you need to read this book and then “Thinking Fast and Slow”. If you get through them both (and really internalise what they say!), it will totally change your understanding of how people make decisions. Important wouldn’t you say?

This is the equivalent pop-management book of “Thinking Fast and Slow”, by the Nobel Prize-winning author, Daniel Kahneman – which is definitely not pop-management, but is ground-breaking – or would be if people could i) understand it, ii) break through the dry presentation approach he uses. A review of this book will come later, so back to The Undoing Project.

It is in the typical style of Lewis (of “Moneyball” and “The Big Short” fame), which makes it an easy read. It is really a story of two brilliant outsiders working in the field of psychology and their institution-busting ideas, but starts with a theme from Moneyball to set the context. He introduces an outsider named Daryl Morey who worked up the numbers (on a statistical basis) on how to select would-be star players for the Houston Rockets basketball team – using the model he had built, rather than what most managers used such as “experience”, “the look of the player”, “how the player walked and talked”, “what others said about him”. Morey saw conventional “wisdom” for what it was – bullshit!

He moves on to introduce us to Daniel (Danny) Kahneman and Amos Tversky. The main gist of the story starts with their impact on the Israeli Military. One of the changes they made that persists until today is they stopped senior officers in the Air-force from beating up on pilots when they did badly (“because they always did better afterwards”), and praising them when they did well (because they almost always did worse after the praising). Kahneman and Tversky simply saw things for what they really were – regression towards the mean. The more times a pilot flew the more times he did roughly equally well and equally badly around an average performance! They just had some good days and some bad days!

How often do we see this happen in the business world today? I was working with a large infrastructure management organisation and a senior manager always thought that because he gave his team a “good talking to” when they did badly, they always did better because of his rant! And this person was promoted into an even larger infrastructure management company based on his “results”.

We then get treated with some of the simple experiments they carried out on students and professors they knew in the academic world, and then extended those experiments into the real world – with incredibly bizarre and unexpected results. They blew the lid off the existent wisdom that the human mind was a rational calculating machine. Instead they offered a small number of critical biases and heuristics that played tricks on pretty well EVERYONE’S mind on this planet. So as not to spoil the book, I’ll quote only one of these highly challenging (and when you get it, highly amusing) examples Lewis refers to towards the end of the book. It’s now known as the “Linda Problem”.

Linda is 31 years old, single, outspoken and very bright. She majored (you will have to work with the American view of the world!) in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.

To what degree does Linda resemble the typical member of each of each of the following classes:

1.       Linda is  teacher in elementary school

2.       Linda works in a bookstore and takes yoga classes

3.       Linda is active in the feminist movement

4.       Linda is a psychiatric social worker

5.       Linda is a member of the League of Women voters

6.       Linda is a bank teller

7.       Linda is an insurance salesperson

8.       Linda is a bank teller and is active in the feminist movement.

Almost every person who took this test made the same mistake. I’ll leave it with you for now and reveal the mistake when I review Daniel Kahneman’s book.

But for all its examples of what breath-taking discoveries Kahneman and Tversky made about the human mind and decision-making, the story is really one of two human beings who worked and lived very closely together to make a potentially massive and positive impact on our world. It is now our job to go out and make that potential a reality.

David Anker

Leading by Alex Ferguson with Michael Moritz

“My job was to make everyone understand that the impossible was possible. That’s the difference between Leadership and Management.”

This book isn’t really about leadership in the formal business sense, but it is about how Sir Alex Ferguson led the various football teams he was involved with over his career. There are some great anecdotes and insights and you don’t have to be a football fanatic to read this book, although you are going to have to dig a little into his story and reflect on how what he tells you helps you think about leadership.

The book ends with a great set of statistics, which I think really should be right up there at the beginning. It includes his record of cups won, internationally and domestically, and how that compared with some of the other great managers of our time. When you see this you will understand how exceptional Ferguson was as a manager.

He also shows the goals scored and let in for each minute of all the Manchester United games. You can clearly see “Fergie time” (where goals were scored in the final moments of extra time), but he also has diagrams of the team composition by age and country as well as a chart of how well the academy boys did in coming through to the first team.

So what of the insights? His comments on how he changed his coaching style, by removing himself from the session to watch others run it, are interesting. “When you are a step away from the fray you see things that are a surprise,” and he goes onto say that this was “The most important decision I ever made about the way I managed." How often in business do we let someone else chair the meeting, or sit at the back and just watch what is going on? I know some good managers and leaders who do just that and that ability to separate yourself from the immediacy of the situation and to see the wider reaction of others is something we sometimes miss if we don’t take Sir Alex’s advice.

“The only time to give up is when you are dead!” – a great quote, but the book also shows how much fear of failure drove him and the team. When he needed inspiration he just played back the injury time in the Champions League final against Bayern Munich from 1999. I watched the game live on TV and I can still remember it well.

Ferguson's approach to targets and goals is different too. He didn’t set himself up to fail. Instead he was rather vague about what the club would win at the beginning of the season. And when things really went wrong, Ferguson focused on the next small steps to get  back on the path again. It’s not “we have the next 45 minutes to score 4 goals” but “let’s try and score the next goal and see where that takes us.”

There is material about buying and selling players, making deals, and remuneration too, but the lasting thought from the book for me was how “long term” his vision was. The youth academy was something for the future. The scouts were looking for very young youth signings and then looking at talent, often over several years, before the club made a move. Ferguson also understood the issue of tampering when he discovered at the beginning of a season that another club had developed a better way of pre-season preparation. What did he do? Nothing that year, it was too late, but he made sure they were up to speed 12 months later. That takes courage, but sometimes you have to go with what you already have and make the changes at the appropriate time.

Mike Bourne

Black Box Thinking (BBT) by Matthew Syed

This review follows the order in which the various topics arrive in the book, and is simply a collection of the messages that jumped out at the reviewer. It is not meant to be a comprehensive review of the totality of the book, and we look forward to further contributions from our readers.

1.       Open vs Closed Loop (page 15)

Up until now, in working with our customers we have used the term “closed-loop control system” to mean a system where feedback is taken from monitoring the output of the system under consideration and making changes to effect improvement, as well as monitoring (and possibly tweaking) those changes to ensure they are indeed delivering improvement. (We also refer to several feedback mechanisms where, for example, the system requires additional data – and feeding back this requirement to IT to collect that data). This is perhaps a service-based and consistent extension of a typical engineering approach to the subject.

Matthew Syed has now potentially polluted this definition (and he does own up to this in the book):

“(the terms ‘closed loop’ and ‘open loop’ have particular meanings in engineering and formal systems theory, which are different to the way in which they are used in this book. So, just to re-emphasise, for our purposes a closed loop is where failure doesn’t lead to progress because information on errors and weaknesses is misinterpreted or ignored; an open loop does lead to progress because the feedback is rationally acted upon).”

Going forward, we may need to change our usage of the term (but to what?), or we explain what we mean and hope the customer gets it.

2.       Learning from Mistakes (page 27)

There is a very nice quote from Eleanor Roosevelt: “Learn from the mistakes of others. You can’t live long enough to make them all yourself”.

We’ll use this quote in some of our material going forward. 

3.       Access to the Error Signal (page 51)

The book raises the idea of catching the “error signal”. This seems to be in two parts:

3.1               Monitoring the Impact of Corrective Action

This is something our organisation promotes and does really well, whereas the point the book makes is that most organisations aren’t equipped (or don’t even think) to do so. In some cases, we have encountered organisations where multiple recommendations/actions have been proposed, but when we suggest they be monitored for their (hopefully positive) improvement impact, we are stared at like aliens! “A decision means a decision. No need to review it, I’m hardly going to change my mind am I?” This may be when they work in a world of personal influence, but not in a joined-up, systems-thinking world.

What is obvious to us can be confusing to the customers we work with.

3.2               The full ecosystem

It is implicit, if you read between the lines, that you need to collect data for the main causes in a “Cause and Effect” system. So you have to keep on improving the data collection as well as the system. However, it is sometimes easy to back-slide – where data is problematic or costly to acquire there is a natural tendency to overlook it.

Our organisation strongly promotes holistic thinking. See point 6 below. 

4.       Cognitive Dissonance (pages 69 - 93)

This is a nice piece about how people want an assertion to be true and (often against all available evidence, or from one single anecdote) move heaven and earth to “make it true” by simply repeating it as often as they can. We see this in every company we work with and often in our own organisation as well. The Government does it all the time.

We see time and again that people make an assertion, and selectively collect any evidence to support it, and selectively ignore evidence that clearly contradicts the assertion.

Our organisation’s consistent and strong promotion of evidence-based decision-making is our main counter to this.

Cognitive dissonance is a well-known phenomenon in the Organisational Design (OD) fraternity, along with Generalisations, Distortions and Deletions, coming about as we try to make sense of the world. There are related subjects such as Cognitive Bias etc. Therefore, may we look to OD for ways to combat these issues?

A neat corollary to the above problem is the observation made by journalist Nicholas Barrett in the FT in summing up the Brexit vote:

“….. Thirdly and perhaps most significantly, we now live in a post-factual democracy. When the facts met the myths they were as useless as bullets bouncing of the bodies of aliens in a HG Wells novel”.

How long have we actually been living in a post-factual democracy? One of our consultants was faced with the following scenario in 2008, when applying statistical evidence to where to place ambulances on standby points based on a call being within a short drive from that point:

Them: “If you put the ambulance there, the chances are 9 out of 10 that the next call will come from the adjacent town.”

Us: “So what if we put the ambulance in the adjacent town?”

Them: “Then the chances are 9 out of 10 that the next call will come from the first town.”

Us: “Are you serious?”

The ‘so what’ is that we should expect rational arguments to be augmented and modulated by the perceptions of people who do not have a technical mind. This does not make them unintelligent, they just think differently.

Related to this is something another colleague in New Zealand picked up listening to a radio discussion about handling crises.

The particular incident was contaminated water in one of the towns in New Zealand and they were interviewing a US professor of psychology who had been involved in the Flint water supply crisis about how public bodies could get their message across. He said that the neurological studies had shown clearly that it was all about trust. If there was no trust there was no chance of getting the message across. He summarised it in the following way:

“It’s only if they know they can trust you that they will trust what you know”

5.       Narrative Fallacy (pages 146 - 148)

This links closely with the above point. It is also extremely well-practiced by politicians, and some of the wily managers we come across in our customers. We see economists on the news telling us why markets changed (often using dreaded binary comparisons) – but only after the events – why couldn’t they tell us what was going to happen before the event if they really understood the system they were referring to?

This is an area studied by the Nobel Prize-winner Daniel Kahneman: it refers to our propensity to create stories about what we see after the event.

We see this in the police, where every improvement initiative is doomed to succeed.

Our organisation again enables these fallacies to be exposed by presenting what is going on in the ecosystem as best as possible by “measuring the right things” and “measuring the right way”, along with visually projecting, with a good degree of confidence, what is likely to happen in the future in a way that is very different from normal forecasting techniques. The essence is not to just explain what caused the last result, but what the next result is likely to be! 

6.       Minimum Viable Product (page 153)

This is about getting started with what you have, rather than trying to make everything perfect before starting. Our organisation promotes this strongly and is counter-intuitive to the IT systems designers who try to specify everything before starting.

Can we better impress on the organisations we work with that this approach is likely to yield more and sooner than the typical “let’s test this in a PoC followed by a Pilot before we get going in earnest months or years later”?

We see organisations suffer from this effect in multiple ways – and in some cases we cause the problem, for example:

1.       Internally, in some cases we may say we need the customer to specify their business requirements, whilst we know full well the customer has no idea what they’re going to do with the system. This is disingenuous on our part. They expect us to be the experts.

2.       We do a ‘quick and dirty’ initial system build to get it done rather than perfect, then when the customer does want to go on, the issue becomes convincing the customer to deploy it to get early benefit, rather than spend time perfecting it while not getting business benefit!

7.       Rapid Development / Fail Fast (page 156)

He talks about rapid practical development vs lots of theory. Rapid loops of develop, test, learn, redevelop, test, etc.

Fail-fast is rapidly catching on at last. Find out what fails as soon as possible, so you can eliminate those parameters and move on to eventually finding success.

Our organisation promotes this strongly in the way we build signalsfromnoise® models, the way we enable drill-down to blind-alleys quickly so they can be eliminated, and so on.

It may be useful to talk with other organisations that promote this approach.

8.       Ballistic Approach vs Guided Missile Approach to Success (pages 157 - 158)

Syed compares the ballistic approach: Identify target, and design sophisticated plans to hit the bulls-eye (e.g. design perfect rifle, design perfect flow model (fluid-mechanics) to account for wind-speed and gravity etc., perfect bullet etc.) or plan the hell out of everything approach to project management and all will be well;


Guided missile approach: Get the missile in the air and the use its guidance system to dynamically locate target, dynamically control thrust and ailerons to dynamically change course as the target changes its course, external conditions change, etc. (agile or rapid development approach to systems development).

He compares before the event reasoning to the after the trigger adaptation.

We see the perfect example of the former on the Major Projects Authority website Annual Report 2015. They say:

“But perhaps the most significant intervention that we plan to make over the next twelve months is to improve the initial set-up of projects. The success or failure of projects is most often determined at their earliest and most formative stage. Projects that have crystal clear objectives, well-defined benefits, appropriately detailed plans, the right level of financial resource, the right people with the ability to understand and manage key stakeholders, and the right leaders – these are the projects most likely to succeed.”

Given that these projects last over many years, it says nothing about adaptability to changing requirements and changing technology – a changing world!

Our organisation promotes the “get going with what you’ve got approach”.

He also brings up “Unintended Consequences”. He doesn’t say much about these, but our organisation’s approach in getting going with what you’ve got, testing improvement actions and then identifying gaps in the measurement system (that measures the ecosystem under consideration) and all the while building a better (but not complete or perfect) understanding of cause and effect – this underpins the point he is making.

Can we learn more from studying the available literature on Complex Adaptive Systems to better promote this guided missile approach with our customers.

9.       Randomised Control Trials (RCTs) (pages 166 - 180)

Great stuff about RCTs, which our organisation supports with signalsfromnoise. Also, Syed recognises that not everyone can do these as they can be impossible or expensive – see next point.

Our organisation supports the RCT approach with the run-of-8 signals indicating sustained improvement (rather than the binary comparison of before and after the trial). 

10.   Marginal Gains (pages 185 - 205)

Great stuff where RCTs are not possible or actually required.

Our organisation and signalsfromnoise support the Marginal Gains approach by looking for statistical runs of results consistently one side of the previous average (of what is being measured) or the other indicating sustained improvement before starting on the next improvement.

Supporting RCTs where appropriate, and supporting Marginal Gains where appropriate, seems to be the ideal position to take?

 11.   Transform or Tweak (page 205)

Great discussion of transform vs tweak – or both. This is exactly why we developed the ideas of Luke Hoebeke to Sense & Respond, Process Improvement and System Improvement.

Transformation and / or Tweaking where appropriate – where to strike the balance?

12.   Understanding Complexity and Blame (page 246)

This is wonderful stuff:

“When we are dealing with complexity, blaming without proper analysis is one of the most common as well as one of the most perilous things an organisation can do. And it rests, in part, on the erroneous belief that toughness and openness are in conflict with each other. They are not.”

In our organisation, we have been experimenting with the phrase:

“Our organisation and signalsfromnoise enable us to understand the simplicity on the far side of complexity”. We’ve seen one global and highly respected consultancy not get this in a global blue-chip UK-based Telecommunications company, and in the same company, we’ve seen 6-sigma black belt consultants from a global confectionery manufacturer not get this. And so on.

There must be literature out there around the dangers of over-simplification? 

13.   Western Attitude towards Numeracy (pages 288 - 289)

Incredible to find all this good stuff in one book!

In the Sunday Times, Dominic Cummins wrote a very powerful article about our poor educational approach to numeracy.

“In the Programme for International Student Assessment (PISA) league table, which measures attainment among 15 year-olds, China rates first and Japan seventh in maths. The UK and US lag in 26th and 36th places resp.

In the UK and US, maths is widely considered to be something you either can or can’t do.”

In some cases, I have heard very senior people proudly declare “I just don’t do numbers!”

This flies right in the face of evidence-based decision-making.

The challenge our organisation faces is “How do we communicate? Or more correctly, how many different ways do we need to communicate, to help our customers access and consume evidence”? 

14.   Power of Project Pre-Mortem (page 311)

An interesting idea that our organisation may want to combine in our project kick-offs.

How could pre-mortems help our customers’ projects to be even more successful?


The author, Dr David Anker, is a Co-founder and Director of Lightfoot Solutions, and is a Visiting Fellow of Cranfield University. The opinions expressed above are the author’s and do not necessarily represent those of Lightfoot Solutions.

Lightfoot Solutions provides Business Improvement consulting and implementation services, along with its Performance and Change Management Platform signalsfromnoise to Emergency Services, Healthcare and Corporates in the UK, Australia and New Zealand.

Lightfoot Solutions and signalsfromnoise (sfn®) – Smarter Insights, Connected Thinking