Can Authoritative Studies Be Trusted
Abstract
A study known as Strat-X, published in 1967, was probably the broadest and most influential study ever conducted on strategic options. Its results have been used or quoted countless times over the years, the latest instance being MX Missile Basing, a study conducted last year by the Congressional Office of Technology Assessment (OTA). In the course of examining the possibility of ship-basing the MX, I found mathematical errors in Strat-X, one of which is quite significant. It has nothing to do with the assumptions, only with a faulty understanding of the ship-trailing process being modeled. This Strat-X error is a simple example of averages being taken too soon, a frequent malady of expected-value approaches. In this case, it takes little extra effort to get the right answer. Further, a more detailed Markov-chain trailing model that also includes trail reacquisition is not difficult to build. In addition, the Strat-X formula for the required number of Soviet trailing ships, duly retranscribed and used in the OTA MX study is not explained, could not be rederived, and does not make sense for certain values of the parameters. A correct expression is also not difficult to find.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1982
- Accession Number
- ADA145909
Entities
People
- M. M. Mizrahi
Organizations
- Center for Naval Analyses