Saturday, October 30, 2010

WWII Enemy Tank Estimates by Sampling Serial Numbers

Background


Imagine you are an Allied General during World War II. How valuable would it be to know the number of tanks that the Axis Powers can manufacture each month? Tanks first came into their own during WWII where they were integral to blitzkrieg strategy (note that the Luftwaffe were not used in blitzkrieg) as well as Patton's campaigns. Knowing this, the number of tanks becomes sensitive, strategically significant information with a variety of applications in military intelligence. This information would improve estimates of tanks held in reserve by the Axis Powers, aid Allied tank production planning, and assist battle planners.

How the Allies Used Math Against German Tanks


An article appeared in Wired magazine's Autopia section a week ago, entitled: How the Allies Used Math Against German Tanks. It is short and worth reading in its entirety, detailing how traditional espionage/interrogations indicated that 1,400 German tanks were made per month. Disbelieving this number, the Allies thought of an elegant alternative which involved collecting the serial numbers of all enemy tanks captured or destroyed and analyzing the information. "Allied intelligence noticed each captured tank had a unique serial number. With careful observation, the Allies were able to determine the serial numbers had a pattern denoting the order of tank production." Someone then called in a statistician, who determined that this was the appropriate formula to use:


and he promptly estimated German tank production at 255 per month. After the war, internal German data showed that monthly production was 256 per month... almost unbelievable in its accuracy. Furthermore, it was more efficient to collect serial numbers and bring in a statistician than to interrogate captured soldiers, try to intercept Axis communications, and then to consolidate all of this information. If quantitative methods were considered first, who knows what other information the espionage and interrogation networks may have garnered?

Implications



Predictability Can Be Preyed Upon: Looters deliberately headed to New Orleans before Katrina hit. Bomb-makers build IEDs triggered by bomb-detection equipment. Germany circumnavigating the Maginot Line to invade France. The criminals in Die Hard requiring the FBI's standard operating procedures to be executed for their plan to succeed.    

Less Effort Required for Quantitative Methods: The collection of digital communications by the NSA is habitually referred to as SigInt (Signals Intelligence) while espionage and interrogation are referred to as HumInt (Human Intelligence). I conjecture that,
Consider Quantitative Methods earlier in the process: Given the cost differential, it behooves decision makers to bring QM into the process earlier so that efficiencies can be gained.

Networking: QM familiarity and creativity can yield surprising opportunities, but only produces results if the QM practitioners are not walled off from their organizations.

Information Sensitivity Evaluation: Does your organization freely distribute any quantitative information that benefits your competitors? Have you ever investigated if this is the case?