The Park Box

THE SIMPLE RULES OF COMPLEXITY

We use rules a lot. Games have rules. The rules constitute them. Driving has rules. If we did not obey the rule to drive on the left or the right we would not get very far in one piece. It is not a good idea to interpret that rule liberally. Better to just stick to it, even if you think the other side of the road is clear. The rule itself is arbitrary. In some countries you drive on the left, in others on the right. Neither is better than the other. But it is vital to choose one, and vital that everyone follows it, so it is embodied in law.

Other kinds of rule are not arbitrary, but ways of guiding action. Rules of this kind used to be called ‘maxims’.
The philosopher Immanuel Kant referred to them in his first major work of moral philosophy, the Grundlegung zur Metaphysik der Sitten or Grounding of the Metaphysics of Morals, published in 1785. Ethics, he wrote, defines the laws governing freedom. Kant distinguished a ‘practical rule’ from a ‘moral law’, the latter being derived a priori. He defined a maxim as ‘a subjective principle of the will’, whereas an objective principle is a ‘practical law’ which can be understood as an ‘imperative’. A hypothetical imperative is a means to an end. It claims that the action it articulates is a good way of realising an intention. A categorical imperative on the other hand, articulates action which should be followed for its own sake. Kant claims that a categorical imperative is: ‘act in such a way that the maxim you are following could be a universal law’. It claims universal validity. Philosophers have been pondering that ever since. Managers might profit from pondering the other imperatives Kant identifies, the subjective hypothetical imperatives which define the actions needed to achieve an intention.

In strategy we are not dealing with universal moral laws, but with deciding how to do the right thing in a particular context. In giving guidance about that we are formulating maxims which constitute hypotheses about the practical imperatives which should be followed in order to achieve certain specific ends. Rather than determining the rules of a game or laying down rules that should be followed no matter what, we are formulating principles which are in need of interpretation in any given situation: subjective principles of the will which have the status of hypothetical imperatives. We are exercising freedom of choice in deciding what to do, and good maxims guide our choices. If we get them right and everyone follows them, people can make better decisions when information is limited and time is short, and whilst all have the freedom to act flexibly depending on their specific circumstances, the totality of everyone’s actions has coherence.

Maxims have long been used by generals. Napoleon formulated a whole set of generic ones to help his marshals when they went to war. Historians have identified over 100 of them, and they form a set of heuristics which explain his methods, such as ‘never unite your forces close to the enemy’, ‘when on the offensive, pursue it to the utmost’, ‘the strength of an army is its mass times its velocity, so march rapidly and press on’, and so on. They are designed as general principles for conducting warfare successfully in the early C19th.
The Prussian Army of the later C19th also formulated maxims in sets of field service regulations which made up what we would now call ‘doctrine’ – i.e. ‘how we do things round here’. They include things like ‘march divided, fight united’, ‘always support your comrades’, ‘sins of omission are worse than an error in the choice of means’. The German Army which emerged from the Prussian Army retained many and added to them: ‘the person on the spot knows best’, ‘if you lose a position, always counterattack as soon as you can, even if you only have weak forces, before the enemy consolidates’. They all emphasised speed, decisiveness and initiative. They made the German Army formidable at operations, and brilliant at execution. They lost the two world wars of the C20th because they executed losing strategies.

Some of the guidance in Napoleon’s maxims and the German Army’s field service regulations are straightforward rules of the kind ‘always do this, never do that’. Others cover specific contingencies: ‘if A happens, do X’. Others require more judgement: ‘in situation B, try to do Y’. In this case, the maxim-user needs to make a judgement about whether the situation is in fact B or something else and work out the best way of doing Y. Simple rules come in many different kinds.

Don Sull and Kathy Eisenhardt have been exploring simple rules for a long time. They first published a well-received article about them in HBR in January 2001. Now they have published a book called Simple Rules which builds on that article, modifies a few of its ideas, but above all extends its scope, covering a huge number of examples from just about every aspect of life, including online dating. Simple rules, it seems, are everywhere. The book is quite entertaining and a stimulating read.

The lure of the subject is that simple rules offer a mechanism which can help us to turn strategy into action in the uncertain, fast-changing, complex environment most businesses face today. They are a powerful empowerment tool, enabling individuals in an organisation to act autonomously but in alignment with each other, making them effectively self-organising. Self-organising systems have evolved in nature, and evolution is pretty smart, so it is certainly something to take seriously.

In May 2001, just four months after Sull and Eisenhardt’s article appeared, scientist Eric Bonabeau and consultant Christopher Meyer published another simple rules piece in HBR called ‘Swarm Intelligence – A Whole New Way to Think About Business’. The article links systems thinking and biology. The question is how creatures of very modest intelligence like ants and termites can create nests which are marvels of engineering, find food under adverse conditions and sustain reproductive success over countless generations, making them today some of the most successful species in the competitive environment of nature.
Curiously, they seem to do so without any charismatic leaders. Rather than rushing around issuing instructions and making speeches, their queens just sit at home being fed and laying eggs. Ants and termites don’t need a leader because they possess ‘swarm intelligence’, and scientists can model how swarm intelligence works by using simple rules. Complexity arises out of simplicity.

Ants’ foraging behaviour can be explained by assuming they follow just two rules: lay pheromone and follow the trails of others. Researchers have used their behaviour to create programmes for routing telephone calls and designing freight routing for airlines. Similar business applications can be derived from how bees allocate labour and have proved to be superior to more complex systems. Sull and Eisenhardt cite the flocking behaviour of starlings as an example of what they call co-ordination rules: avoid collisions, head in the same direction as your nearest neighbours and stay close to your nearest neighbours. Each individual starling can follow the rules using only local information, but the flock can respond to unpredictable local conditions, such as power lines blocking their path.

Now clearly, there is a difference between programmers being able to model the behaviour of insects or birds with just a few rules and Kant or Napoleon’s maxims. Evolution works by trial and error – mainly error – and takes millions of years. Kant and Napoleon worked things out, and they did it rather more quickly than the blind watchmaker of nature. And a rule like ‘lay pheromones’ is not a maxim to guide ants when exercising their freedom. I don’t know if ants have free will, but they are not noted for perspicacious judgement. Individual insects are not great at modifying their behaviour or adapting to specific situations, as you will know if you’ve ever watched a bee banging its head time after time against a pane of glass rather than flying through the window you’ve opened a few inches away to allow it to escape from your room.

In their book, Sull and Eisenhardt cast their net so wide that sets of quite different things, covering hard rules, heuristics and rules of thumb, guidelines, injunctions, operating principles, key success factors, boundary conditions and constraints, decision-making criteria and instructions are all called ‘simple rules’. Semantic over-stretch becomes acute when the failure of Scott’s expedition to the Antarctic is interpreted as a failure to change some simple rules. I doubt that Scott had a simple rule that his sleds would be pulled by men rather than dogs. He just decided not to use dogs, which turned out to be a lethal mistake. Having become so adept at using the hammer of simple rules, the authors see nothing but nails.
However, though I think the interesting cases they are talking about really involve maxims rather than rules, I am prepared to give way on terminology. We don’t talk much about ‘maxims’ these days, and no-one outside philosophy departments apart from a few oddballs like me has much time for Immanuel Kant, virtuoso thinker though he was. So let’s stick with ‘simple rules’. However, before saying goodbye to old Immanuel, let’s draw on some of his thoughts to describe the sort of simple rules I think we should be interested in when it comes to executing strategy: they are hypotheses, created in a specific situation, about an imperative, which enable people to make good decisions and act on them when they have to make difficult choices in the face of limited information and limited time.

Sull and Eisenhardt have a taxonomy of simple rules. It is not about their different status and function, but what it is that various rules address. In their original HBR article they identified five types of rule: boundary rules that help you make yes or no decisions; priority rules that help you rank alternatives; stopping rules that tell you when to reverse a decision; timing rules that tell you when to do things; and how-to rules that spell out how a process is to be executed. In the book they add a sixth, the coordination rules exemplified by the starlings.
How helpful are these distinctions? Well, formulating good rules is difficult, and this gives a handy list of factors people are likely to need to know about if they are to be able to make good decisions on their own. Also, people tend to absorb rules in a particular order: boundary and how-to rules first, stopping rules last. So if you are creating some rules it is quite helpful to check that you have covered boundary conditions, priority-setting, timing and so on.
In the chapter about using simple rules in strategy, the authors suggest a three-step process: firstly work out what will increase the gap between the two needles which describe the willingness of customers to pay and the cost line of a business, or, as they put it, ‘identify the critical choices which will drive a wedge between revenues and costs to increase profits and sustain them over time’. The second step is to identify a bottleneck, ‘a decision or activity that is preventing the company from improving profitability’. The third step is to craft the set of rules which ‘when applied to the bottleneck, improves profitability’. That sounds fair enough, though I am not sure it is always a matter of removing bottlenecks. A bottleneck suggests something like ‘lack of qualified sales people’ or ‘R&D capacity’. Later on the authors define ‘writing’ as a bottleneck in making successful comedy shows. That sounds to me like another piece of semantic over-stretch. Successful comedy shows are well-written. Writing is surely something those of us fluent in management-speak might call a ‘key success factor’. No matter.

When it comes to the really interesting bit - crafting the rules – things get a bit vague. We are told that it should not be done top-down, but by the rule users themselves. My bet is ‘sometimes, and sometimes not’. It depends who is in a position to identify the patterns out of which good rules originate. Then, they say, having generated some rules, you need to test them – as Kant pointed out, they are hypotheses. I’ll buy that.
However, the difficult bit is deciding what the rules should be. Here the guidance is thin, for good reason: crafting the rules is a creative act determined by insight and judgement. That is no surprise because crafting strategy is also a creative act determined by insight and judgement. The art of strategy has occupied some of the finest minds in history. The historian Martin van Creveld has characterised Napoleon as ‘the most competent human being who has ever lived’. Whether or not we agree with van Creveld, even his most severe critics all admit that Napoleon was a pretty bright bunny. Getting to be good at strategy is a serious challenge, and so is turning it into simple rules.
One gets the impression from the book that simple rules drive all forms of life on the planet, from ants to internet dating. Curmudgeonly old bugger that I am, I am not sure that is the whole story. But there are circumstances in which they are almost the only way to go. It comes down to complexity.
My delightful friend Stephen Carver, who teaches at Cranfield Business School, has made a special study of project management. After war, encounters between Mars and Venus, and giving your bank a discretionary mandate to manage your money, project management is surely the greatest cause of human misery. Steve has boldly gone where angels fear to tread. He believes that the reason project management is one of life’s greatest disaster areas is complexity.

He distinguishes two kinds of complexity; structural and dynamic. Structural complexity occurs when there are lots of discrete but interlinked things to manage; dynamic complexity occurs when those things themselves and the relationships between them keep on changing. Structural complexity implies co-ordination; dynamic complexity implies adaptation. They pull the poor project manager in different directions. The two variables can be visualised in a matrix and the high-high box is where you have not projects but programmes and the rate of change is high. Steve and I have done some sessions about this. We called the high-high box on the matrix ‘The Park Box’. Let me explain.

In the 1930’s the Royal Air Force took on the challenge of solving an apparently intractable problem: how to protect a country against attack from the air. A British Prime Minister, Stanley Baldwin, believed the problem could not be solved: ‘The bomber will always get through’, he told the House of Commons. One Air Marshal, Hugh Dowding, sought to prove him wrong by creating the world’s first integrated air defence system. It was an information gathering, processing and dissemination system which constituted the world’s first intranet, a command and control system conceived as a network.
Air defence was structurally complex, involving the co-ordination of defending fighters in different areas, AA guns, barrage balloons and air raid warnings, and had to be responsive to any raids any enemy might launch from anywhere, distinguishing which aircraft were friendly and which hostile, and among the hostile raids, which were the greatest threat, where they were and where they were heading.
It was also dynamically complex. Information about raids was updated every two minutes, with the information coming at first from radar stations, then, after the raiders had crossed the coast, from the 30,000 members of the Observer Corps. Within the system, the role of the highest level (Fighter Command) was to filter and sort the information and pass it on; the role of the second level (Groups) was to decide which forces to deploy against which raid; and the role of the third level (Sectors) was to control the aircraft being deployed to bring about a successful interception and then talk their boys home. Nobody knew in advance what their opponent was going to do, but they had to be able to counter any move. They had to take rapid decisions under high uncertainty, and the situation changed by the minute.

The largest of the four Groups making up Fighter Command was 11 Group. Covering London and South East England, it controlled about half of Fighter Command’s fighters. Its commander was a New Zealander called Keith Park. He mastered the challenge of the high structural and high dynamic complexity box of Steve’s matrix by using simple rules. He called them ‘Instructions to Controllers’.
The defence system faced its most severe test 75 years ago, from the beginning of July to the end of October 1940, in a campaign known as the Battle of Britain. Over this period, Park issued over forty instructions to controllers. Here are a selection:

- Do not engage pure fighter raids;
- Engage all bomber raids before they bomb;
- Scramble early enough to gain height and position;
- Do not pursue stragglers over the sea;
- Aim to break up formations;
- Use the squadron as the basic unit of deployment;
- Use pairs of squadrons when facing very heavy raids.

These rules are determined by Park’s rare insights into the nature of airpower in 1940 and how to conduct air defence. They show him turning the strategic intent of Fighter Command into operational reality. That intent could be expressed as: ‘oppose all serious raids in order to deny the enemy air superiority’. They are logically rigorous deductions about how to conduct defensive fighter operations in order to achieve that intent.
They were constantly updated in view of the evolving situation. Park examined the numbers to identify patterns and understand what was going on: claims and losses, number of successful interceptions, time to intercept, time from issuing an order to scramble to time the last aircraft leaves the ground, and so on. The system provided him with a vast amount of information at the end of every day through something called the Defence Teleprinter Network. Air defence was a fast-moving business even in 1940. Park worked at the sort of pace familiar to retailers watching daily store performance.
The data was important, but not enough for Park. Every evening he flew to the hardest hit airfields in his personal Hurricane to see for himself what was behind the data, and to talk to pilots about what was going on in the air. He got them to talk to each other about tactical best practice (for example, which formations worked best in the air), connecting them across squadrons and bases, and when he got back he sent out instruction to his station commanders about what his pilots needed most to help them. He also used the insights he got from his pilots to adjust his rules.
Each rule had a logic. The first five never changed.

‘Do not engage pure fighter raids’ - They are harmless and engaging them will waste resources in terms both of aircraft and time. This was in fact what the Luftwaffe wanted him to do. Noticing that he didn’t was the first sign some of the Luftwaffe’s wiser commanders had that winning was going to be tough.
‘Engage all bomber raids before they bomb’ – The strategy was to be able to challenge raids indefinitely. In order to do that Park had to protect his vital infrastructure. In order to do that he had to break up raids before they bombed so that their bombing would be ineffective.
‘Scramble early enough to gain height and position’ – Park knew that height and position (ideally up sun) trumped numbers every time. He wanted to give his pilots a competitive advantage in the air fighting, so he needed to get them off the ground as soon as a raid had been identified as a threat to give them time to reach altitude, but not so early that they would run out of fuel and have to land before engaging the raiders.
‘Do not pursue stragglers over the sea’ – He was less concerned about pilots getting confirmed kills than in inflicting some damage and getting home. He wanted his opponent to pay a price every time he entered British airspace and wanted to be able to exact a penalty the next day, so he needed to conserve his forces. Pilots baling out over the mainland would be picked up and be in action again on the morrow. If they came down over the Channel they would probably be lost.
‘Aim to break up formations’ – Doing this would disrupt bombing. The best way to do that was by conducting head-on attacks. Carrying them out was dangerous but often effective, and Park was ruthless enough to encourage them (though he never ordered them). In themselves they did not inflict many aircraft losses because the closing speeds were so high that aiming was difficult, but if the formation was broken up, inflicting losses was much easier.
The last two rules changed as the Luftwaffe sent over larger formations. Using single squadrons as a unit was a corollary of the primacy of time. He did not want to waste time forming up. It removed the need for coordination which would have overloaded radio frequencies. It also gave him flexibility and allowed him to engage raids continuously because each squadron would fly off, engage and return within an hour and be ready to be off again, as others did the same thing at slight intervals, giving him constant cover. Instruction No 10, issued on 5th September, specifies the use of pairs of squadrons against large raids with the faster climbing Spitfires taking on the top layers of German fighter cover and the slower Hurricanes the lower flying bombers. The instruction was reinforced in a signal five days later and elaborated on in Instruction 16 of 11th September.

So it was that the Battle of Britain was won using simple rules.

Park could be absent dealing with other matters, confident that his controllers could handle things by applying them. He thus rendered himself superfluous day to day, but he was fighting a campaign, not a battle, so he needed to keep on adapting his rules to the situation and to make sure they were being applied.
He worked them out himself, so it was top-down. Park had little time for pilots like Douglas Bader who thought they knew better. He had better information and made better judgements. But he shared his information widely, so as to create a shared context among his distributed leaders. His instructions to controllers were routinely copied to station commanders, squadron leaders and some to anti-aircraft command.
They worked, for him and the country in 1940.
Maybe they could work for you and your business today.


Notes:
1: A dictionary definition of a maxim is ‘a succinct formulation of a fundamental principle, general truth or rule of conduct’.