One of the topics discussed in Fooled by Randomness was the tendency of people in general, but journalists and business economists in particular, to attempt to spin any random series of events into some sort of coherent story. Rather than being content to report an event in its own right there is a compulsion to explain even if the link is tenuous.
He gives some typical “Bloomberg” examples (from the widely used news/trading service). Comments such as “Dow is up 1.03 on lower interest rates”, when there is little evidence that the two events are related, particularly when a 1 point move on the Dow is essentially it remaining unchanged. Essentially they are trying to explain noise as something real and this goes wider than the reporting on markets.
Other classic examples are journalistic explanation of polling data. A drop of a percent or two in the figures of one poll is read as having been the voter’s perceptions. This article in the SMH from earlier this year is a classic example:
The downturn in support for the Government over the past month indicates the tax cuts have failed to assuage public dissatisfaction over the rise in interest rates and petrol prices, and the fiasco following the death in Iraq of Private Jacob Kovco.
How much of a swing? 3 percentage points, which as Andrew Leigh explained at the time, once the margin of errors are taken into account, is a result not significant enough for any credible social scientist to draw a conclusion from.
Taleb’s contention is that in situations like this that humans are poorly equipped to deal with randomness but evolution has conditioned us to attempt to induce some reasoning in a classic post hoc ergo propter hoc fallacy. This is evidenced by “gambler’s ticks”, the desire to repeat behaviour that you have taken to believe assisted you in getting the desired outcome previously. Whether this be lucky shoes, or screaming out the desired card and banging on the table like one annoying woman I had the misfortune to play at a blackjack table with.
One bird was conditioned to turn counter-clockwise about the cage, making two or three turns between reinforcements. Another repeatedly thrust its head into one of the upper corners of the cage. A third developed a ‘tossing’ response, as if placing its head beneath an invisible bar and lifting it repeatedly. Two birds developed a pendulum motion of the head and body, in which the head was extended forward and swung from right to left with a sharp movement followed by a somewhat slower return.
I was disappointed to read later researchers have disputed the conclusion that what was observed was superstitious behaviour and attributed it to standard foraging behaviour, which funnily enough makes Skinner’s conclusion a case of reading meaning into noise.
Bloggers too of course fall into this trap, chasing every piece of news as something of significance and trying to build a case of it or explain it as a general narrative. This post on the Liberal party’s “war on science” or this post on gun crime to cite a couple of recent ones I read. I’m sure I do the same from time to time.
One of the benefits of drawing everything into a grand narrative is that it makes things easier to remember. Our brain it seems is very good at remembering things if they fit into some pre-established patterns, and so chess masters for example can remember complicated board layouts in a game, but struggle to remember ones where pieces are laid out randomly. Another example that occurred to me was the Australian aboriginal practice of having “songlines” a song which described the terrain, landmarks and animals in a particular journey across the land. Disregarding the spiritual significance, the songs provided narrative and a way a large amount of essentially random landmarks could be remembered and recalled to navigate across the country.
Trying not to rationalise every event which occurs can actually be difficult, particularly working out what to rationalise and what not to as they are occurring particularly the small stuff. Still its worth bearing in mind that you don’t have to, and that much of the time causal explanations people come up with for losely related day to day events are complete crap.