Bits of Books - Books by Title

Adapt: Why Success Always Starts with Failure

Tim Harford

Problems like national health or education seems intractable - best way is to experiment with many different approaches - try something different, but accept that some of the new approaches won't work well. But this is politically difficult, partly because politicians need results within 3 year electoral cycle, and partly because most of the schemes will publicly fail and the pols will get blame. That is our fault as well, because we should be encouraging pols who will test drive their ideas with pilot studies.

Jamie Oliver did nearest to that by trying to get healthy meals into school cafeterias. An accidental (almost) controlled experiment, and analysis showed that less fat and salt and more fruit and veges translated into better health and ed results. Tony Blair, then British PM, fell over himself to endorse campaign. He had been in power for 8 years at the time.

More books on Food

Next big problem is feedback, particularly the informal anecdotes from people on the frontline. In a hierarchy, everyone sugar-coats their opinions when talk to powerful people. There is no reward for straight talking unvarnished truth. A traditional hierarchy is good at broad strategy problems but bad at dealing with local idiosyncracies, or at listening to the inconvenient dissent from junior staff.

Looked at professional poker. Can be played rationally, but too often becomes emotional, particularly after a player has lost a big bet. Problem that the brain refuses to register that the money has gone - the player makes over-aggressive bets that he would normally be smart enough to avoid to win back 'his' money. Given that commentators often describe the share market as a casino, it should come as no surprise that investors do exactly the same thing - they hang on to shares that have plunged in the hope that things will turn round.We are determined to avoid crystallizing that loss or drawing a line under a decision we regret.

More books on Games

We've all heard about psychology expt where subject and a group of other people are asked to judge length of a stick. Subject doesn't realize that others are all stooges who have been told to over-estimate. The subject moves his estimate up to match the group 'consensus', even though he can clearly see that it's wrong.

More books on the Mind

But the follow up expt also important. This time one of the stooges gives a correct answer - the subject is now not the only one with a non-standard view of the world. This one non-conformist was enough to free the subject to give an honest estimate. Further testing showed that this liberation occurred even if the non-conformist emphasised that he was nearly blind. Moral of story is that dissent is crucial. A group of the very smartest agents aren't as successful as a more diverse group of dumber agents. Even when "different" means "wrong", trying something different has a value of its own.

Any large organization faces a basic dilemma between centralization and decentralization. Frederick Hayek article published 1945 (eventually led to a Nobel Prize in economics) explained that a complex world is full of information that is localized and fleeting. He said that decisions taken at center could be valuable in allocating resources, but decisions taken at fringe of an org are often better because take into account the particular circumstances of time and place.

More books on Business

Armies and businesses often seem to operate on the assumption that better computers and better communication gather all the information together in one place so that a central planner could make the key decisions. But the opposite is true. It's the guys on the frontline who can see the immediate needs and can tell (by quick trial-and-error if necessary), what is needed.

The Spitfire. Up until 1930 the accepted military wisdom was that bombers could not be stopped. Only way to counter them was with Mutually Assured Destruction (although not till nuclear era that that phrase popularized). In 1931 the British sent out specs for a new fighter. No aircraft company could get close, but the Supermarine company suggested a radical new design. An enterprising civil servant found £ 10,000 in budget to order a prototype as 'a most interesting experiment'. It broke all the rules - conventional wisdom was sure that a plane could not exceed 260mph; the Spitfire hit 450 in a dive. The Spit only fired it's guns forward, whereas most assumed needed a two-seater with a gunner in a turret.

More books on Inventions

In 1938, a year before the war began, one expert stated: "We should now build, as quickly and in as large numbers as we can, heavily armed aeroplanes designed with turrets for fighting on the beam and in parallel courses ... the Germans know we have banked upon the forward-shooting plunging Spitfire whose attack ... if not instantly effective, exposes the pursuer to destruction." The expert was Winston Churchill. (He was the boss so they built the plane he wanted, as well. It was the Boulton-Paul Defiant, and you've never heard of it because it was a sitting duck for German fighters).

It is easy to say, with benefit of hindsight, that official doctrine was all wrong. But that is the wrong lesson. It's not that British nearly lost war with misconceived strategy. It is, given that misconceptions were all but inevitable, they somehow managed to commission a wild card anyway. Even when such prototype expts were regarded with derision.

The idea of allowing several ideas to develop in parallel runs counter to our instincts. We naturally ask "What is the best option?" and concentrate on that. But given that life is so unpredictable, what seemed like an inferior option may turn out to be just what is needed. In an uncertain world, we need more than just a Plan A.

The Spitfire was a long way down the alphabet from Plan A, not least because of the unlikely characters associated with it. Supermarine was run by Noel Pemberton Billing, a rich politician who hated lesbians. He once won a libel case where he accused an exotic dancer of spreading "The Cult of the Clitoris". The Spitfire's designer was a cantankerous engineer named Reginald Mitchell who refused to work with the bureaucracy when Vickers, a big defence company, bought Supermarine. The company let him run a 'skunk works'. Then there was Fanny Houston. In 1929 and 1930 Spitfire's predecessor won the Schneider Trophy and held the world speed record. But the government withdrew funding for the plane's development as the Great Depression bit. Lady Houston had become the richest woman in Britain after marrying a shipping millionaire and inheriting his fortune. Furious at the lack of official support, she wrote a cheque to Supermarine that covered the entire development costs of the Spitfire predecessor.

More books on Individuals

The PlayPump sounded like a great idea - a roundabout connected to a pump to get water for villages in Africa. Idea being that kids would play on it and water a side effect. But turns out the only time kids play on it is when a foreigner with a camera appears. Rest of time it's an inefficient pump. A traditional hand-pump took 28 seconds to fill a 20L bucket; PlayPump took over 3 minutes of strenuous running around. Now accept that PlayPumps work in large primary schools but nowhere else.

Big problem that aid groups get carried away by projects which look good - free textbooks or flip charts for schools etc. But when someone actually measured the programs, found that had little impact, and best outcome came from de-worming schoolkids. The aid projects typically fail because of 'god complex' - they are some outsiders view of what is best for the Africans. To be honest, they should compare (with randomised trials) the pet project with just giving the same amount of money to the villagers, and seeing which gotbest results.

Medical trials instead of subjective opinions. 300 years ago a sceptical Belgian scientist challenged quacks to prove blood-letting actually did any good. He suggested a wager of 300 florins, offered to take 500 people with any sorts of ailments, divide group in half. He would cure his group without bloodletting, quacks could do theirs, and "we shall see how many funerals both of us shall have." But the blood letting continued for another 3 centuries. 1747 Royal Navy tested various suggestions to prevent scurvy ranging from vinegar, cider and sulphuric acid to citrus fruit. Divided sailors into 6 groups and found that oranges and lemons gave best results. RN ships started carrying fruit and many sailors owed their lives to the test.

Idea of charter cities like Singapore, Shenzhen, Dubai prove can build city anywhere. Difference is they have different rules to surrounding areas. New Songdo completely-from-scratch new city S Korea. Paul Romer suggests that Canada take over Guantanamo Bay - US gets rid of a PR disaster, Cuba gets a gateway to 21st C capital, and Canada gets a Hong Kong in the Caribbean. Economically plausible but politically impossible.

John Endler classic evolution expt. He studied guppies in Venezuela and noticed that those in pools at bottom of waterfalls or cataracts tended to be drab while those in upper pools were brightly colored. Suspected that it was because the guppy-eating pike cichlid couldn't climb waterfalls. The drab guppies were surviving because better camouflaged. So set up a trial. Filled 10 pools with guppies, put the dangerous cichlids in some pools and either gentler or no predators in others. Some pools had large pebbles, some had fine gravel. Within 10 generations, the guppy populations had adapted. In the dangerous pools, only the drabbest guppies survived. And their camouflage fitted the design of the pool, with larger patterns on pebble-pool fish and smaller patterns on the gravel-pool ones. Evolution driven by failure - some guppies were eaten, while others survived to produce future generations. The guppy population evolved camouflage through trial and error, but no individual guppy did. Each was born with a good enough camouflage or not.

3 essential steps to adapting in both business and everyday life.
First, try new things, expecting that some will fail.
Second, make failure survivable - have safe places to expt (and fail) or move forward in small steps.
Third, make sure you can tell when you've failed, or you will never learn

Our brains are very good at stopping us from recognizing failure:
1. Denial isn't just a river in Egypt. Cognitive dissonance between our self-image and reality. Correct way to deal with anomaly is to say "I am a good and capable person but nevertheless I made a mistake."

2. Chase our losses to try get money back

3. 'Hedonic Editing' (term coined by Richard Thaler in Nudge). Think of the 'praise sandwich' you get at evaluation time. Boss says "This is excellent work. I think it wd be great if you [impt feedback here]. But overall it's great work." What you hear is "It's basically excellent." So a writer, facing a wall of criticism for his latest effort, finds a few people who like it, and convinces himself that the views of this discerning minority should be given greater weight.

Expt where subjects asked to rank 6 Monet paintings from (liked) best to worst. he researchers then offered the subject a choice of two spare prints they 'happened' to have. The prints offered were always number 3 and number 4 on preference list. Naturally they chose number 3. The researchers then came back later and asked the subjects to rank the same 6 pics. The rankings changed - the selected print was now 1 or 2; the rejected print was now 5 or 6. The really interesting part was that the expt was with severe anterograde amnesiacs, people incapable of forming new memories. And the researchers didn't come back 6 months later; they came back after 30 minutes, by which time the subjects had forgotten everything. They had no memory of seeing the prints, or even of meeting the researchers, before. Yet they still strongly preferred the print they chose, even though they had no conscious knowledge of seeing it before.

Somehow you have to develop the skill of asking yourself at every stage whether you might be wrong, and if there is any way to test it. Self-employed happier than salaried workers. Obviously a bit more in control of their destiny, but that can be wiped by lack of security. But the self-employed get a tick of approval every time they get an order or when they pay their invoice, whereas wage workers get much less regular or meaningful feedback.

More books on Work

John Kay in his book The Truth About Markets uses term 'disciplined pluralism' to describe the way markets work: exploring many new ideas but ruthlessly cutting out the ones which fail.

Metaphor of the fitness landscape - a vast and ever-changing geography consisting of troughs of failure and peaks of success. Evolution explores this landscape with a serendipitous mixture of wild leaps and small steps. The wild leaps usually end up at the bottom of some chasm, but sometimes they land in the foothills of some totally new range of mountains. The small steps perhaps lead only to the top of a molehill.

In life we tend to notice the people who take the wild leaps, the decisive moments when nothing is ever the same again. Some artists create best work when very young - the intuitive leap of Sergeant Pepper's Lonely Hearts Club Band. But others slowly perfect their craft - Mondrian's greatest paintings when he was in his 70's.

More books on Art

Basic message of book is that it is vital to keep experimenting, despite our inclination to avoid mistakes. A single successful expt can become a Spitfire - it can transform our lives for the better in a way that a failed expt will not transform our lives for the worse - as long as we don't deny or chase losses.

More books on Meaning of Life

1. Try new things
2. Try them in a context where failure is survivable
3. Avoid the brain tricks that stop us learning from our failures

The brain tricks:
1. Denial - based on 'cognitive dissonance'. We have trouble holding on to two apparently contradictory thoughts at once: "I am a capable, respected ......." and "My latest effort is a cock-up".

2. We chase our losses. (The poker players who make more reckless bets to try to recover "their" money)

3. 'Hedonic editing' ( a term coined by Richard Thaler in Nudge). We try to convince ourselves that the failure doesn't matter. Which is fine, as far as it goes - you feel better, but you don't get better.

(The 'praise sandwich' where evaluate someone with "You're doing a great job. You could improve by doing ..... But like I said, we think you're going well." - What do they internalise? "You're great." Full Stop.)

Self-doubt is a difficult skill to acquire - the ability to ask yourself "What if I'm wrong?" David Petraeus, US general in charge in Afghanistan, and as of 2012, boss of CIA, was notorious as a young officer for being unable to ever admit to being wrong. His mentor, Jack Galvin, gave him job of being devil's advocate, teaching him that an organization that suppresses dissent and ignores internal criticism soon makes disastrous errors.

"What you need are people with good judgement in other parts of their lives who care about you and will give you their honest opinion with no strings attached."

One of the unappreciated merits of the marketplace is that it supplies most of the elements of this validation squad. A purchase is a 'thumbs up' for the entrepreneur.

In our personal projects have to find other ways to get feedback. One way is to videotape your performance and watch it from viewpoint of someone else.

Evolution: success emerges from failure in nature - ceaseless generation of random mutations, tossing out the vast majority which make organism worse, and preserve the tiny few which improve. Do this enough and apparent miracles emerge.

David Brooks column on this book here

Books by Title

Books by Author

Books by Topic

Bits of Books To Impress