On the Necessity of Errors

I want to commend an excellent post by Jeffry Snider at RealClearMarkets to your attention. The post begins with the value of error in the decision-making process but actually is pointing out the necessity of errors in producing good decisions. Here’s a snippet:

There is efficiency in seeming randomness. In the natural world, there is tendency for physical forms to find the most efficient shapes possible. The sphere, for example, is the most efficient distribution of forces which the material in it can displace maximum volume for finite quantities. The soap bubble is a perfect example of that. Economies and other complex systems are in many important ways similar; always searching for the most efficient manner of distribution given finite quantities of everything. There is no way to find such efficiency and thus true progress at the direct exclusion or dismissal of error. The Federal Reserve, had it existed in 1879 and assigned itself the task under its current rules, would have told Edison to stop after only a few tries at the light bulb and instead use whatever he had at that point. In other words, monetary intrusion is a purposeful deformation of the efficient shape because economists have assumed they can, with no wisdom of the crowd, define better what is and is not error and prudence.

Read the whole thing.

We’d better hope that errors are a necessary part of good decision-making. We’ve made some whoppers in recent years. I hope we can benefit by them. I see few signs that’s the case.

2 comments… add one
  • michael reynolds Link

    Such an interesting piece. It calls to mind for me the sense I had on first encountering a digital catalog in a library. I was working for a DC law firm as a library grunt and occasional researcher. And I had without really understanding why, adopted the habit of flipping through a few wrong index cards – mistakes – before I got to the right one. If you’re looking for Br start your search in Bl because what the hell, you never knew what you might run into by accident. When you look at a painting you want to see the whole canvas and the frame as well. Probably some bit of code left over from a brain evolved to hunt and scavenge. Digital search felt too much like someone was forcing me to use a microscope without an opportunity to take in the whole scene first.

    It’s a bit like brainstorming in a group where you might have one guy who is just way off in left field, but you still feel a need to keep that guy around. Hearing the wrong helps you find the right. Maybe it’s the inverse of the famous Conan Doyle quote: “When you have eliminated the impossible, whatever remains, however improbable, must be the truth?” In this case, If you have excluded the impossible, you may have obscured the truth.

    Thought-provoking piece. I’d love to see more data on this.

  • Guarneri Link

    I was, unfortunately, not as enamoured with the article as he seems to conflate a number of statistical and mathematical issues to give a thin veneer of science to guessing born of experience and judgment. He could have simply said “this is the way markets work, kinda messy but it’s the best we got.” Not much new there.

    The falsity of the jelly bean example proving some magic to the technique of a large crowd guessing about total sugar load, in spite of or caused by error, can be illustrated by a revised experiment. This time, ask the very same crew how many jelly beans, but imbedded in the core of the container have six hidden golf balls, yet the appearance from the outside be unchanged. The error prone guessers won’t come similarly close. Rather, they would systematically overestimate the number of beans by the close packed jelly bean volume displaced by six golf balls. The real world analog would be unknown, uncontrolled for or immeasurable economic variables. But upon “spilling the beans” the error of the guessers’ assumptions would be revealed.

    I think we end up in the same place as to the observation of expected suboptimal results from meddling in markets, but it is not some freakish mathematical law of guessing.

Leave a Comment