Tuesday 6 September 2011

'Multi-layered' problems are best solved with a touch of Obliquity

On a recent flight across Europe I read John Kay’s recent book Obliquity. Short and easy-going, the book focuses on the idea that solving problems using ‘direct’ methods based on logical deduction and algorithms often results in an imperfect solution. In many cases, he points out, an indirect or ‘oblique’ method arrives at a better solution.

As an example, someone travelling across London may, out of habit, look at the Tube map and travel from their starting point to the station closest to their destination. However, this direct method may be inappropriate in the cases where it is actually quicker to walk than take the Tube*. Kay relates how he once took the Tube from Paddington to Lancaster Gate because he didn’t realise how close the two are above ground (Londoners reading this should be smiling at this point).

He applies this ‘Obliquity’ principle to many problems with varying degrees of abstraction. For example, he suggests that ‘the most profitable companies are not the most profit-oriented’ and ‘the wealthiest people are not the most materialistic.’ He goes on to attempt to explain reasons why indirect solutions beat direct ones so often, bringing the concepts of complexity, incompleteness, abstraction and pluralism into the narrative.

To Kay’s list of reasons why direct, supposedly-rational approaches to problem solving fail I would like to add one contributor: the ‘multi-layered’ nature of most tricky problems. Often we approach a problem with one line of thinking, but as we investigate further we find further layers of complication in the problem. In that sense, tricky problems are like onions. However, as you peel away layers of an onion you can tell when you’ve found the core. Problem-solving is not like that: you never know whether you have reached the ultimate level of understanding or whether there is another level of complexity beyond your grasp.

One of my favourite examples of the ‘multi-layering’ of problems is the debate over energy efficient light bulbs (abbreviated to EE bulbs henceforth). Let’s start with the proposition ‘anyone who switches to EE bulbs will save energy.’

At the most basic level one might conclude that this is true, as by definition EE bulbs use less energy to produce the same amount of light as normal bulbs. But let’s look a level deeper. The additional energy used by non-EE bulbs is released as heat. This is normally considered to be wasted. But what if you get your heating from electrical heaters anyway? If you do, then switching to EE bulbs will mean you need to put your electric heating on more, thus you do not save energy. You may even end up using more.

I’ve seen this debate play out several times. Going a level deeper, one could argue that lightbulbs tend to be near the ceiling, and heat released at ceiling height just stays there, and is therefore wasted. But going a level deeper, one could argue that heat is mostly released by radiation and reaches all parts of the room. Ultimately the debate never gets resolved.

As John Kay points out in the conclusion of ‘Obliquity,’ the recent past is littered with examples of poor decision making by people in power. Clearly there is a long way to go before we really understand where we are going wrong. In the meantime, the pursuit of better problem-solving methods is a worthy goal.

*I love this innovative re-design of the Tube map, which depicts relative geographic lay-out of stations more accurately.

1 comment:

  1. The argument you used in EE lighting is bit of a moo point. It applies specifically to a building that uses electric heaters and that requires heating all year long. Any building that requires heating for only some days of the year would reduce its total energy consumption by switching to EE lighting. Furthermore, heat dissipated from light bulbs cannot be controlled. Hence even in a building that has electric heater and requires heating all year long, you would still in most circumstances want to switch to EE lighting.