In Football Analytics: Now and Beyond, Daniel Memmert, Jurgen Perl and Robert Rein noted that when computers became available in the late 1970s “more complex data recording and analysis became possible” and “connections to mathematical game theoretical models became feasible”.
This introductory comment sent me off to investigate the literature relating to the arrival of computers in sport.
Wikipedia (link) observed “computers in sports were used for the first time in the 1960s, when the main purpose was to accumulate sports information. Databases were created and expanded in order to launch documentation and dissemination of publications that contained any kind of knowledge related to sports science”.
Morris Collen (1986) (link) noted “the earliest references to any applications of electronic digital computers in medicine appeared in the 1950s in biophysics,bioengineering and biomedical electronics publications”.
The first all-electronic digital computer was invented by Presper Eckert and John Mauchly (link) and their co-workers at the University of Pennsylvania in1946. The computer was called the Electronic Numerical Integrator (ENIAC) (link) and Calculator; it used 18,000 vacuum tubes as its active logic elements and had wired program plugboards and programming switches for storage. In 1945 John von Neumann (link) devised a method for storing programs, that is, instructions as distinguished from data in a computer by an electrically alterable memory capable of storing both the instructions and the data to be used in calculations. In1951 Presper Eckert and John Mauchly used John von Neumann’s stored program technology and built the Universal Automatic Computer (UNIVAC) (link) for the US Census Bureau for use in the 1950 census (link). This was the first computer to handle both numeric and alphabetic data, and it used magnetic tape drives to replace punched cards as a storage medium.
The second generation digital computers, in 1958, used transistors instead of vacuum tubes and added magnetic-core storage memory designed by Jay Forrester (link) at the Massachusetts Institute of Technology, Boston. The third generation computers appeared in 1963 in the form of solid-state integrated circuits consisting of hundreds of transistors, diodes and resistors embedded on tiny silicon chips, a process called large-scale integration (link). This permitted the construction of a hierarchy of computers of different sizes.The IBM360 (link) series, introduced in 1964,was one of the earliest third generation main frame computers. Minicomputers, smaller in size and designed to do generalized though more limited tasks, were successfully marketed in the 1960s, especially by Digital Equipment Corporation as their Programmed Data Processor (link) series.
The first personal computers were built in 1971 by John Blackenbaker (KenbakI) (link), and by the Hoff Intel Corporation (link). The fourth generation of computers, in the 1980s, exploits very large scale integration containing thousands of components on very tiny silicon chips. These computers greatly increased performance despite the smaller size and lower cost.
In terms of programming languages: FORTRAN was developed in 1957 by John Backus (link) and his colleagues at IBM for scientific and numeric calculations. COBOL (link) was created in 1959 by a joint committee of computer manufacturers and was partly based on previous programming language design work by Grace Hopper. BASIC was developed in 1965 by Thomas Kemeny and John Kurtz (link) at Dartmouth College as a language for introductory courses in computer science. LISP was developed in the late 1950s at the Massachusetts Institute of Technology by John McCarthy (link).
For Morris Collen’s discussion of the origins of informatics see his 1994 paper (link). In it he notes:
In the early 1940s ‘computer’ was a job title. A person was given a calculator and a set of formulae and then called a computer. By the late 195Os, however, when an electronic device carried out the arithmetic functions of addition, sub- traction, multiplication, and division, or the logical functions of “and,” “or,“ and “not,“ the device was called a computer.
A digital computer required a central processing unit with a primary or main memory to hold the data being processed, a program of instructions for processing the data; and circuitry to perform arithmetic and logic operations and to control execution of instructions. Peripheral equipment included secondary or auxiliary storage devices (such as magnetic tapes and disks); data input devices (such as keyboards, card and tape readers, and direct input from secondary storage devices); and data output devices (such as displays, printers, and plotters). It was the computer software-the computer languages, programs, procedures, and documentation that made the hardware usable for applications. Programming thus evolved from first-generation machine languages to fourth generation applications-oriented languages. By the 198Os, many of the most commonly used programs were commercially available, and most computer users did little programming themselves. During the same four decades, computer commu- nications moved from copper wires to fiberoptic cables.
John Billings (link) suggested in 1880 when discussing census data “there ought to be a machine for doing the purely mechanical work of tabulating population and similar statistics . . He thought of using cards with the description of the individual punched in the edge of the card”. In 1882, Henry Hollerith invented a paper card with 288 locations for holes that were punched out by a hand-operated machine (link). He went on to build machines for electrically punching and sorting these cards. The 1890 census data on 62 million people were processed in one month using 56 of Henry Hollerith’s machines. Morris Collen argued that John Billings laid the conceptual foundation for the development of digital computers in the United States.
The world’s first fully functional, program controlled, general-purpose, electromechanical digital computer was completed in 1941 by Konrad Zuse (link), in Germany. Howard Aiken (link), at Harvard, working with IBM engineers, built the Mark I electromechanical computer in 1943. The Mark I was a program controlled computer that was based on the decimal system. All machine operators were performed electromechanically by wheel counters and relay switches (link).
Morris Collen notes that credit for the invention of the first generation of all-electronic digital computers has been given generally to the mathematicians Alan Turing and Max Newman (link) and their colleagues at the Bletchley Research Establishment in England. Their machine, called the Colossus, was installed and working in December 1943. The first electronic digital computer built in the United States was completed in 1946 by John Mauchly, and Presper Eckert, and their co-workers, at, the Moore School of the University of Pennsylvania. It was called the Electronic Numerical Integrator and Calculator (ENIAC). A precursor to ENIAC had been constructed by John Atanasoff, a physicist at Iowa State University, between 1937 and 1942 (link).
Bo Dahlbom (1996) (link) suggested:
The first computing machines were built during the Second World War. At first they were simply thought of as automatic versions of the mechanical calculating machines used in offices and retail stores at the time. In a request for funding in 1943 to the Army Ballistics Re-search Laboratory, John Mauchly described the machine he wanted to build, the Electronic Numerical Integrator and Calculator. One of his primary objectives was to build more efficient calculators to produce mathematical tables, in particular ballistic tables for military use. Such tables had been computed by people using calculators, but with the rapid development of weapons during the War, these human “computers” were unable to keep up. Efforts had been made to rationalize computing work by organizing it on the model of the typing pool, but these computing pools were now to be replaced by machine computers that were claimed to be faster, cheaper and more reliable. All through the 1950s, this original use of computers, as computing machines, continued to dominate. Computers were automata that were fed algorithms in order to make large computations. To program the machines meant turning calculation tasks into algorithms that the machines could handle. At that time to become a programmer you had to master the science of calculation, and numerical analysis. When, in the early 1960s, computers began to be used as information systems, it was their capacity to handle large sets of data that became the focus of attention. Computerised information systems were made possible by development of memory mechanisms.
Bo Dahlbom added that attempts were made all through the 1970s to introduce home computing, but when eventually personal computers really became a commercial success, it was due to their use in offices, as spreadsheets, word processors, and desktop publishing tools. The 1980s became the decade of the PC and the use of computers again shifted its center of gravity.
Computers in Sport
In this blog, I have written about Donald Knuth at the Case Institute of Technology in Cleveland, Ohio in the period 1956-1960 (link). The Computer History Museum‘s biography of Donald includes notes:
Knuth’s lifelong love affair with computers began as an undergraduate when he discovered the IBM 650 computer system at Case. He quickly mastered the inner workings of the machine and developed a novel program to automate coaching of the school’s basketball team.
I have written about Anatolij Zelentsov and Valerij Lobanovs’kyj in a golden age of cybernetics in Russia in the 1970s (link). They first met in 1968. They lived near Kyiv, the centre of the Soviet computer industry. An early prototype of a domestic computer was developed there in 1963. Jonathan Wilson observes “it is no great surprise that Lobanovskyi should have been gripped by the spirit of technological optimism”. They worked together for four years at Dnipro and then, in 1974, Valerij was appointed coach of the Dynamo Kyiv team. Anatolij moved to Dynamo Kyiv with him. As their partnership developed, so too did their interest in the ‘functional readiness’ of players. This led Anatolij to develop a computer program to collect and analyse data about players’ physical, cognitive and affective behaviours to inform and support Valerij’s quest to optimise the team’s performances in training and competition.
It strikes me that both posts underscore that Donald, Anatolij and Valerij were situated to take advantage of the availability of computers. Their professional experience enabled them to be open to digital computing. Donald and Anatolij were particularly skilled in using computers. Donald had experience of programming and Anatolij was Dean of the Dnipropetrovsk Institute of Physical Science.
Daniel Link and Martin Lames (2009) (link) point out that “the term Informatik was mainly associated with questions of technology”. They add that Informatik was regarded as “the science of the systematic processing of information, in particular the automatic processing using digital computers”. This definition includes “mathematical activities, which deal with algorithmic processes for the description and transformation of information and also engineering activities, concerning aspects of the development and application of computers”.
In his summary of computers in sport, Adrian Lees (1985) (link) noted “computers are becoming increasingly common in sports applications particularly since the development of the low-cost microcomputer”. He added “they have numerous roles to play but can be identified as a major mechanism for easing the burden associated with data gathering and information processing”. He reports the use of computers in the 1960s and the importance of computers for biomechanical modelling.
Peter Dabnichki and Arnold Baca (2008) (link) edited a book that investigated computers in sport. They observe “applications of computers in sport (to be understood in its broadest sense)have been reported since the mid 1960s. Statistical computations, numericalcalculations in biomechanical investigations and sport documentation tasks werecarried out” (my emphasis). They add “progress in hardware (processor speed, storage capacity,communication technologies), software (tools), information management concepts(data bases, data mining) and media (internet, e-Learning, multimedia) are of essential importance”.
Arnold Baca (2006) (link) notes “in 1975, an international congress entitled Kreative Sportinformatik was organized in Graz, Austria. He adds “1967 as the year, where an automatic processing of sports documentation was first demonstrated at an IBM computer in Graz, Austria”. Hristo Novatchkov and Arnold Baca (2015) (link) report that computer science flourished with “the appearance of the first microprocessors in the mid 1970s, allowing the realization of computers with higher capacities and increasing processing power”.
In a foreword to a collection of articles about computers in sport (Arnold Baca, 2014) (link) Larry Katz notes that data has been derived from technological innovation that has included information about athletes’ skill levels, tactical, technical, physical and emotional performance. In the book, Arnold uses a particularly German approach to computers in sport as an interdisciplinary endeavour. This approach is clarified in the book by Daniel Link and Martin Lames (link). They explore three decades of computers in sport as a growing discipline.
The literature about computers generally and computers in sport specifically provides an important context for our understanding of analytics and informatics. In all countries where computers were developed, there is evidence of sport taking early advantage of digital developments. In a separate post, I note the biography of Neil Lanham. Neil combined his professional involvement as a chartered surveyor and estate agent with his passion for sport. This kind of integration is also evident in the biographies of Donal Knuth (link) and Anatolij Zelentsov and Valerij Lobanovs’kyj (link).
Kenbak-1 (Adwater and Still)
WIMU (Real Track Systems)
Neil Lanham (Neil Lanham)
I was delighted to see Garry Gelade’s tweet (link) about this post:
“I have used 80 character punched cards you submitted by hand, paper-tape I punched myself, and 10MB external hard drives that could only be read by the machine that wrote to them. All those moments will be lost in time, like tears in rain”.
I thought this was a fascinating observation and it made me think about how many biographies I might need to explore to build a story about computers in sport.