Sunday, October 22, 2006

Sources of Power

I'm currently in the middle of a really interesting book entitled Sources of Power: How People Make Decisions by Gary Klein.

It's an interesting read, because it dispels the notion that the best decisions are always made after listing all the possible options, weighting and rating each option, adding up the ratings, then selecting the highest-scoring option.

As it turns out, highly experienced people in high-stress, time-sensitive roles (think firefighters, air force pilots, emergency room doctors, ICU nurses) who have to make rapid-fire decisions with very little information on hand do not go through the motions of enumerating and then rating their options. Instead, they use what the author calls a Recognition-Primed Decision (RPD) making model.

With RPD, decision-makers take in the situation, immediately come up with a candidate action, check mentally if it will work, and implement it right away if they believe it has a chance of success. If they're not convinced their candidate action will work, they just come up with another, evaluate it, discard it, and so on until they find one which they believe it will work.

There's no weighing of multiple options to find the 'best' one. There's simply weighing one option at a time and going with what (given the information and time available to decide) looks like a decision that will work.

I'm not sure yet how this is supposed to apply to me in real life, but then, I'm only on p.58 at the moment. Should be interesting to read it through all the way to the end.


Ants said...

This is why training pays dividends, specially if it's good training. To quote Richard Marcinko's SpecWar Commandments: "... the more thou sweatest in training, the less thou bleedeth in battle." I think that if I am trained to quickly recognize patterns and implement decisions which leave lots of options open towards the goal, the better off I am under stress conditions.

M said...

^^ Agree. The book also talks about the value of simulations as a training method. Although I did see from the table of contents that a later section also talks about the pitfalls of using simulations. Haven't reached that part of the book yet, though. :D

Another interesting point that the author makes -- the RPD model really works well only with experienced people. For the less experienced, it's still much better to use the traditional approach of listing all your options, scoring them, then ranking them --> precisely because you don't have the experience to discern the right choice.