Generated by DeepSeek V3.2| A Million Random Digits with 100,000 Normal Deviates | |
|---|---|
| Name | A Million Random Digits with 100,000 Normal Deviates |
| Author | RAND Corporation |
| Country | United States |
| Language | English |
| Subject | Random number generation, Statistics |
| Publisher | Free Press |
| Pub date | 1955 |
| Media type | |
| Pages | 400 |
| Isbn | 0-8330-3047-7 |
A Million Random Digits with 100,000 Normal Deviates is a seminal 1955 reference work published by the RAND Corporation. It is a printed volume containing a vast table of random numbers generated through a physical process, intended as a tool for statisticians, researchers, and engineers. The book became a foundational resource in the Cold War era for fields requiring high-quality randomness, including Monte Carlo simulations, cryptography, and experimental design. Its creation was driven by the needs of the United States Air Force and projects at the Los Alamos National Laboratory.
The project originated from the post-World War II research environment, where institutions like the RAND Corporation were heavily engaged in work for the United States Department of Defense. The demand for reliable random numbers surged with the development of complex systems analysis, nuclear weapon simulations, and early computer modeling. Prior methods, such as using published tables of mathematical constants, were deemed insufficiently random for these sensitive applications. Under the direction of researchers like George W. Brown, the RAND Corporation undertook the monumental task of generating a truly random sequence. The resulting data was first issued as a report before being formally published in book form by the Free Press in 1955, during a period of intense scientific competition with the Soviet Union.
The book's primary content is a list of one million random digits, each from 0 to 9, presented in a matrix of 50 rows and 100 columns per page. The digits are grouped into blocks of five for readability. The second major section provides 100,000 normal deviates, which are random numbers following a Gaussian distribution, essential for statistical methods like analysis of variance. The volume includes introductory material explaining its use and a set of instructions for researchers at facilities like the Lawrence Livermore National Laboratory to perform statistical tests. The austere, typeset presentation belies the complexity of its generation, serving as a physical artifact of pre-digital information theory.
The digits were produced not by an algorithm, but by a physical random pulse generator called the Electronic Random Number Indicator Equipment (ERNIE). This hardware device used a fundamentally random source—the unpredictable pulses of a Poisson process derived from an electronic noise source. The output was recorded on punched cards and subjected to extensive statistical analysis to verify its randomness. Teams at the RAND Corporation and consultants like John von Neumann applied rigorous tests for frequency, serial correlation, and poker hand distributions. This process was a landmark in the philosophy of randomness, distinguishing between pseudo-randomness and true physical randomness, a concern central to the work of Claude Shannon at Bell Labs.
Upon release, the book was rapidly adopted by government agencies, corporations, and academic institutions. It was crucial for designing unbiased samples in surveys conducted by the United States Census Bureau and for simulations in operations research. Within the aerospace industry, companies like Boeing and Lockheed Corporation used it for reliability modeling. The tables also found early use in the developing field of cryptography, providing key material for one-time pad ciphers. Its influence extended to the Social Science Research Council and the National Bureau of Standards, helping to standardize methodological approaches across disciplines and reducing reliance on inferior random number sources.
The book remains a celebrated icon in the history of science and technology in the United States. It has been reprinted by the RAND Corporation in 2001, highlighting its enduring curiosity. It is frequently referenced in discussions about data, probability, and information entropy, and has been exhibited at institutions like the Museum of Modern Art for its conceptual design. The work inspired later projects, including the RAND's Table of Random Numbers website, and is often cited in the works of authors like William Poundstone. It stands as a physical monument to a pivotal era when the management of chaos and uncertainty became central to national security and scientific progress.