Precision and avoiding error


in Geek stuff, Politics, Security, The environment

It is fundamental to the nature of truth as accessed by human beings that there is a trade-off between how precise a view we express about a particular subject and how certain we can be of avoiding error.

This can be expressed in a basic way by thinking about estimation. If we are asked to guess how many years Genghis Khan lived for, it is safe to say ‘between 1 and 1000’. It’s not very precise, but the real figure is in there somewhere. Every time we specify a narrower band, we increase the risk of missing the target.There is an inescapable connection between providing a more precise answer and running a greater risk of excluding the answer that is true.

This remains true when it comes to questions that are much more complex and abstract, such as “what is likely to happen in Afghanistan after NATO leaves” or “what are the likely consequences of climate change on international security”. In responding to complex questions, we probably need to acknowledge the limits of what is really knowable. We have limited information, and often a limited span of time in which to make choices. Dealing with that probably requires an awareness of the precision/certainty trade-off, along with a willingness to keep all possibilities in mind, even if they are unproven.

Quite possibly, we should be more willing to err on the side of caution when the level of uncertainty is high and at least some large credible risks seem to exist. When a nuclear reactor may be melting down, it may be a good idea to inject the core with seawater. Doing so ruins the reactor for future electricity generation, but reduces the risk of a terrible outcome in which a meltdown is coupled with a large-scale containment failure. Excluding the worst possibilities usually involves real costs of various sorts, but it is probably better to accept the certainty of a known loss to significantly reduce the probability of an unknown but potentially much worse outcome. In short, it pays to play it safe on important matters.

{ 1 comment… read it below or add one }

. December 12, 2011 at 12:19 pm

Nuke plant manager ignores bosses, pumps in seawater after order to halt

TEPCO’s credibility came under fire again when it was revealed May 26 that the head of the Fukushima No. 1 nuclear power plant continued pumping in seawater after company superiors issued instructions to stop.

Although the official turned out to be correct, such an important decision normally requires the president of the Tokyo Electric Power Co. to sign off on it.

Not only was there a breakdown in internal communications, but TEPCO officials have until now insisted that the seawater pumping was temporarily suspended.

At a May 26 news conference, Sakae Muto, a TEPCO executive vice president, was asked when company executives learned that Masao Yoshida, the Fukushima plant head, had gone ahead and continued pumping in seawater.

Leave a Comment

Previous post:

Next post: