Patterns, Breaks & the Limits of Predictability
View Comments: Newest First | Oldest First | Threaded View
Re: Predicting behavior from a gleam in the eye?
  • 4/16/2014 10:51:24 PM

I agree with Michael that the human (sentiment/vocal/past behavior) analysis is our best bet to be able to predict the actions of world leaders. It may not be solid yet, but it's close.

The objection that major military actions are rare is valid. But there is a way around it, which is to study how the leader makes the near-big decisions. if we watch, analyze and predict how Putin makes the almost-big decisions, we'll develop a good model for how he'll decide the big ones too.

Maybe what we need is a crystal ball
  • 4/16/2014 11:12:35 AM

On a more practical level, who do you think got fired or "re-deployed" on the State Department's Russian or Eastern European desks for not having predicted or foreseen the current situation with Russia and Ukraine? And how much would Mr. Putin's "pauses" have told us about his seemingly random but clearly strategic annexations? If I'm reading this right, Shore's model only helps us with where Putin will annex next, but doesn't really address where he might initiate a new kind of action and for what reason(s). 

Re: Predicting behavior from a gleam in the eye?
  • 4/15/2014 11:31:59 PM

Agreed, Lyndon. And obviously one of the biggest combatants of predicable analytics is entropy itself. That is to say that anything can happen. Being unpredictable is what separates human territory from animal territory, haha!

Predicting behavior from a gleam in the eye?
  • 4/15/2014 5:27:19 PM


Michael writes

I think, given the current state of emotion detection and sentiment analysis technologies, that it may be possible to read a leader based on quantifiable variables and historical patterns. The science of analytics is moving beyond predictions of iterative processes and into far more human territory. And even though no one method is foolproof, I'd bet there's a predictive model that reaches similar conclusions to those that Shore posits are reachable only via strategic empathy.


Well maybe. Do you mean predicting behavior from a leader's vocal tone or facial expressions? Or the "historical patterns" of other leaders?

Seems to me you'd need an awful lot of evidence derived from experience to validate a model derived that way.

Maybe the leader just read a tragic novel, or has heartburn, or a bad hair day ... and decides to order global nuclear attack? Could happen...