Bacon’s Rebellion Now Online

The May 9, 2005, edition of the Bacon’s Rebellion newsletter has just been posted online. Fifteen columns and features plus a shameless plug for Barnie Day’s new book, “Notes from the Sausage Factory”.

Click here to view it.


Share this article



ADVERTISEMENT

(comments below)



ADVERTISEMENT

(comments below)


Comments

  1. Ray Hyde Avatar
    Ray Hyde

    Congratulations, again.

    I particularly liked your coment that “Warner Wonks recognize that Knowledge Economy businesses are more likely to invest in a region where they can recruit employees with highly specialized skills and engage the support of universities, suppliers and other enterprises with specialized capabilities.”

    I wrote my masters thesis on the history of war gaming, and it goes back a lot farther than 1990. Some of the first computer applications ever were devoted to analyzing (gaming) war scenarios. For four years I worked with simulations for various topics including modeling toxic fumes in the event of a chemical attack and the activities of dismounted infantrymen, down to the level of their heart rate, body temperature line of site, and probability of kill.

    There is a lot of fantastic stuff out there, but there are also major dangers. 1) A simulation is only a simulation, it is not the real event. I recall reading a novel in which the Soviets simulated a war game, which they won. They sent the analysis to the U. S. president and demanded that he surrender, on the basis of the simulation results. After studying the simulation, U.S. experts concurred. The rest of the novel concerned itself with the idea that the results of an entire simulation can turn on one faulty assumption.

    2) In some cases it may be more realistic and less expensive to perform a test, rather than a simulation. However, the BMD program shows that thelines between a test and a simulation can be blurry. Microsoft Project can be considered a simulation of the events, schedule, and costs of a real project. However, if you take the time and trouble to enter every activity in the real world into the model, then the model will cost more than the project.

    3) All the glitz, displays, and programming power in the world means nothing if the model is either wrong or unwieldy. In the end, the real grunt work in modeling and simulation comes down to what is called validation and verification. Verification asks the question does the model perform correctly, as designed. Validation asks the question does the model accurately represent real world events. In the end, both of these questions involve judgement that must be made by ordinary human brains.

    Such brains are unlikely to choose to live in the kind of world postulated by EMR.

Leave a Reply