U.S. Strike on North Korea? Prediction Update

The article NBC: U.S. May Launch Strike on North Korea Nuke Test, uses Franklin’s decision making method, in which the elements of a list of pros and cons are given equal weight. Although Franklin’s method was devised as a personal tool, it can also be used to analyze the decision making of others.

It’s been reported that the C.I.A. is working on prediction computers. Unless the developers rely entirely on unfathomable neural networks,  they must grapple with choosing the right method for the problem, and when to switch.  It is a fascinating problem, because, unlike Bayesian probability theory , the sample space — what you might call the “possible outcomes”, is not rigorously defined. Whether the sample space itself is correct becomes a random variable.

What does this mean to somebody who is nosy about the future? It means that you better not stick to your guns too long. As new information comes along, you should revise your prediction without embarrassment. This is why Franklin’s method is so acceptable here.  If new information morphs the problem, switch without embarrassment to another method. But for now,  Franklin’s method remains the choice, with several changes to the lists:

The Pros

  • Trump’s vow to solve the problem of North Korea.
  • His recent use of force in Syria.
  • The enthusiasm of China state media with the Xi-Trump meeting, in spite of China’s awareness of the above. Hence, the “trading material” reference.
  • Shared dislike of “Fatty Kim.”
  • Possible awareness by South Korea of a grim future with the North.
  • The conventional wisdom that force is off the table. Conventional wisdom is always vulnerable.
  • Two more THAAD batteries ( Reuters: four more launchers) have been deployed, at the very substantial cost of U.S. $1BN each, footed entirely by the U.S.
  • Stalled delivery of the promises of the Trump presidency, with a search for “fungible” alternative achievements.

The Cons

  • A possible attack by the North on the South, with all the ramifications. Countered  by the addition of THAAD batteries.
  • Refusal by the South to face up to the growing threat.

In the mind of the decision maker, the con, “A possible attack…”  is weakened  by the pro, additional THAAD batteries. Since we are trying to fathom a mind, this is entirely subjective. The list modification has nothing to do with actual cost reduction. But neither does it go against actual cost reduction. The domain of prediction is entirely mind.

In Xi-Trump meeting; Long Range; North Korea, it’s suggested that Trump’s concept of achievements is that they are fungible — replacement of unachievable objectives by others that are:

Since Trump’s concept of achievements is that they are fungible, he reconsiders the South China Sea.  There are things you want to keep, and things you want to trade. It’s key to streamlining a business.

In NBC: U.S. May Launch Strike on North Korea Nuke Test, I concluded,

Since the pros have it, 6 to 2, the estimate is that an attack has significant chance. It may have strange aspects.

I did not assign a probability, an “XX percent.” As a member of the Forecasting World Events team, my numbers were weighted with many others — a “transverse ensemble”, so it made sense to do so. But Franklin’s score, formerly 6 pro/ 2 con, is now 8 pro / 2 con.

Hence the chance of a strike on North Korea has increased.

The cheapest secret of prediction is to set aside all the discursive thoughts and discussion that might appear in a news-with-story article. Should the gamble be taken? Is it worth the price and the danger? Would it buy us security? Important as these issues are, they are needless distractions to the predictor.