Month: March 2021

Worldly Wednesdays: Nolan Bushnell and Spacewar!

This post is part of an ongoing series annotating my book They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I. It covers material found in chapter 5 on pages 68-70. It is not necessary to have read the book to comprehend and appreciate the post.

Now that we are through all the early experiments, its time to start examining the beginning of commercial video games. And its here the stories start to get more complicated. There really is not much disagreement in the sources about who did what in Spacewar! or Tennis for Two, but when it comes to the launch of Computer Space and Pong, there are three major participants and what feels like ten different stories. In writing the book, I tried to construct a single coherent narrative out of this mess, which means that I made certain choices about which information I found reliable and which I did not. In some cases, where that was impossible, I included footnotes explaining where the stories diverged. The single longest footnote in the book explained why I was confident Nolan Bushnell did not see Spacewar! while a student at the University of Utah in the 1960s. So let’s expand that footnote into an even longer blog post.

The standard story Nolan Bushnell tells about his first encounter with video games basically goes like this: in 1965 or 1966, soon after transferring to the University of Utah from Utah State, a buddy or a fraternity brother informs him there is something really cool he needs to see. Sometimes they just head on over to the computer lab, other times it plays out like the world’s geekiest heist story as they sneak in late at night. Either way, he beholds Spacewar! for the first time at the University of Utah Computer Center, where its running on a big mainframe, either an IBM 7094 or a UNIVAC 1108. Bushnell is entranced and is soon playing the game all night long whenever he can.

Also working as an arcade manager at the Lagoon Amusement Park at the time, Bushnell thinks to himself that the game would be perfect as a coin-operated product if it were not running on such an expensive computer, but alas, there is nothing to be done. He goes on with his life, playing his Spacewar! and even programming a few of his own games like a “Fox & Geese” program until he graduates in 1968 and moves to California. An avid go player, Bushnell meets a man named Jim Stein at a Stanford University go club, who takes him over to the Stanford Artificial Intelligence Laboratory (SAIL) and reconnects Nolan with Spacewar! Soon after, Nolan sees a sales sheet for a Data General Nova, realizes computers may now be cheap enough to run a coin-operated video game, and the rest is history.

Bushnell has consistently told this version of events for fifty years, and while some of the details get embellished as time goes on, the core story remains unchanged. I fully believe that today, Mr. Bushnell considers the above to be a true and accurate account, as memory is a funny thing and the more we relay a certain set of facts, the more real they become in our own minds. Sadly, this account is almost certainly false, save for the last part where Jim Stein introduces — rather than reintroduces — Nolan to Spacewar! at Stanford.

Nolan Bushnell playing a very different game from Spacewar! at the University of Utah.

So why lie? Why does it matter whether Nolan Bushnell first saw Spacewar! in 1966 at Utah or 1970 at Stanford? Either way, he played a complex computer game relatively early and took inspiration from it to play a key role in launching commercial video games. Well, the answer to that has to do with patent litigation. As has already been mentioned more than once in this series, when Ralph Baer and company developed the prototype of what became the Magnavox Odyssey, they filed patents relating both to the invention of video games generally and to the invention of a system in which two objects electronically rendered on a screen collide and one of them changes vectors specifically. Once those patents were issued in 1972, Magnavox, as the sole licensee of Ralph Baer’s technology, could theoretically claim patent infringement against any electronic game in which a human-controlled object and a computer-controlled object displayed on a screen collided and one of them changed vectors, which basically describes any of the dozen or so ball-and-paddle games that took the coin-operated world by storm in 1973 after Atari launched Pong. The best way to avoid paying damages for patent infringement in this scenario would be to show that your own invention predated the work of Baer and his team to some degree. As Bill Rusch came up with the Odyssey table tennis game in late 1967 and the collision patent was filed in 1969, suddenly whether Nolan Bushnell was inspired by Spacewar! in 1966 or 1970 becomes incredibly important.

Meanwhile, in Summer 1973, Nolan Bushnell gave his earliest known interview in which he discusses the origin of his fascination with computer games. The venue is a short documentary called Games Computers Play that chronicles the birth of arcade video games through Bushnell’s Computer Space and Pong as well as the fascination with Spacewar! at SAIL. In this documentary, a very young Nolan Bushnell states the following:

We used to play Spacewar a lot at the AI project at Stanford, which is a big computer complex, and one day it just hit, you know this is a lot of fun, you know I oughta be able to package it and sell it for a price.

No mention of Utah or amusement parks or long-standing dreams of a coin-operated video game empire. He was just hanging around Stanford and one day realized people just might pay money to experience this fun computer game. After Nolan reveals his inspiration, the documentary transitions from talking about Atari to talking about Spacewar! at SAIL, and the narrator begins the new segment with the following words: “The AI Project, where Bushnell first played Spacewar…” While this statement does not come straight out of Nolan’s own mouth, clearly the understanding of the documentarians based on their interview with Bushnell is that he first saw Spacewar! at Stanford.

By November 1973, Nolan Bushnell is singing a different tune. A profile of Atari in a trade publication called Systems Engineering Today recounts the following:

During the early part of the Spacewar craze, Nolan Bushnell was pursuing an EE degree at the University of Utah and working summers as manager of the games department at an amusement park. He was familiar with Spacewar, and from his amusement arcade background he believed that it would be a great arcade game. The only problem was the price: a computer, even a mini, would cost far more than any amusement arcade operator would spend for a game.

And there it is, the Spacewar! at Utah story fully formed. A story from which Bushnell would never deviate in all the years to follow.

So what changed between Summer and Fall 1973? Well, another November 1973 profile of Atari, this one in Business Week, provides the answer:

Magnavox Co., which introduced its Odyssey games that can be played on home TV sets last year, also plans to require companies marketing coin-operated TV games to obtain a license. But some people in the coin-operated game business believe that the Magnavox patents pertain primarily to the home TV market.

So by Fall 1973, Magnavox is making noises about enforcing its patents against companies producing coin-operated video games, and said companies are already trying to figure out their response to this threat. Interestingly, Bushnell takes his Utah story a step further in Business Week as well. Says Bushnell:

The idea for Pong really goes back to my time at the University of Utah. I used to go into the computer center late at night and think up games to play with the computers.

Now, Nolan is not just playing games in the mid 1960s; he is dreaming them up too. And, he claims this is the lineage of Pong, a lineage that is now being conveniently established in a time well before the Magnavox Odyssey existed.

Nolan Bushnell in Games Computers Play, the one time he indicated he first saw Spacewar! at Stanford.

In 1974, Magnavox finally pulled the trigger, suing Atari and several other companies for patent infringement. In the depositions that followed, Nolan continued to refine his Spacewar! story. From a deposition given by Nolan Bushnell on January 13-14, 1976:

Q: Were you personally involved in any activities prior to December 31, 1969 related to apparatus for playing of games which utilized cathode ray tube displays?

A: Yes, I was.

Q: What was the first such activity of that kind that you can recall?

A: I recall playing a game on the computer at the University of Utah.


Q: When did that activity occur?

A: I have been trying to pinpoint that. I think it was in the neighborhood of 1965. It was shortly after I came to the university of Utah.

Q: By shortly after you came to the University of Utah, how long a period to you mean by shortly?

A: I really don’t recollect. […] I had a friend in the engineering department that I used to play chess with that said, “There’s some great games over at the computer center.” And we went over one night and played.


Q: Could you describe the game which you saw on a computer at the University of Utah, this first game you saw?

A: Yes, It was a game which was called Space War.


Q: Using a computer?

A: Yes.

Q: What kind of computer was being used?

A: I’m not sure. That’s one of the things that I can’t put the time on it. It was either a Univac 1108 or an IBM 7094. The University of Utah changed computers while I was there and I’m not sure which it was, really.

This exchange establishes two things: that Nolan Bushnell is now swearing under oath that he saw Spacewar! at the University of Utah around 1965, and that his testimony is as vague as possible. He does not remember exactly when he saw it, and he does not remember on exactly which computer he saw it. In fact, he does not even really remember the person who showed it to him. From the same deposition:

Q: What was the friend’s name?

A: His name was Jim Davies, I think.

Q: And you knew him through your work at the University of Utah?

A: No. I knew him through the chess club.

Q: Do you know where Mr. Davies is located today?

A: I have no idea. I’m not really sure that Davies is his last name. In fact, just a second. I’m not sure that Jim Davies isn’t another guy. It’s Jim something, and it started with a D, but I’m not sure.

This vagueness would continue in a followup deposition given on March 2, 1976.

Q: Mr. Bushnell, in your testimony in January you referred to a Jim Davis or someone with a name like that and then referred to him as Jim D. Have you determined with any great specificity who that individual was?

A. No, I haven’t. In fact, I– No.

Q. This was an individual who you said showed you a game being played on a computer at the University of Utah; am I correct?

A. That’s correct.

Q. Have you made any effort to determine who that individual was?

A. Yes, I have.

Q. What have you done in that regard?

A. I went to the university and went through the rogues’ gallery. They have a listing of the graduates for each school year. I attempted to match a face with a name.

Q. Did you personally go to the university and do that?

A. Yes, I did.

Q. Where is the rogues’ gallery maintained at the University of Utah?

A. On the second floor in front of the Electrical Engineering department.

Q. In what office or department?

A. It’s in the hallway. It’s actually right in front of the computer center.

Q. Did you do anything else to try to establish who Jim D. was?

A. Yes , I did.

Q. What else did you do?

A. I talked to some of the people in the computer center attempting to find some people that had been around at approximately the same time.

Q. Were you able to find anyone?

A. No, I wasn’t.


Q. When did you make this trip?

A. It was in the middle of January.

Q. Of 1976?

A. Yes. Sometime shortly after the deposition.

Q. Did you prepare any report on your trip?

A. No, I did not.

Q. Any memoranda of any kind?

A. No, I did not.

Q. Were you able to find–

A. Well, memoranda? I wrote the names down in my notebook here.

Q. Nothing other than that?

A. No.

Q. Were you able to find any document that would support your testimony concerning Jim Davis showing you a game played on a computer at the University of Utah in 1965?

A. No .

Q. Did you find any documentation that would support your testimony that a game was played on a computer at the University of Utah in 1965?

A. No.

Q. Did you find anything other than documents tangible, that would support your testimony that a game was played on a computer at the University of Utah in 1965?


A. Excuse me just a second . No.


Q: In your earlier testimony you indicated that you weren’t certain whether the computer was a UNIVAC 1108 or an IBM 794 [sic] that you saw a game played on. Do you know which of those two it was?

A. No, I don’t.

Q. Did you make any effort to find out at the University of Utah?

A. No.

Q. Did you make any effort to find out what computer was at the University of Utah in 1965 at the time you say you saw a game played on a computer there?

A. No.

After this alleged first encounter with Spacewar!, Nolan claims he did not encounter the game again for some time. Back to January 1976:

Q: After this first occasion when you saw Space War shortly after going to the University of Utah what was your next activity with relation to the apparatus for playing games using a cathode-ray tube display?

A: Well, it was about, oh, somewhere around a year later and one of my fraternity brothers got involved in the computer center a little bit more and introduced me to several of the people and we got to talking about the games and I thought it would be kind of fun to learn how to program games.


Q: Did you see any Space War games between the first time that you saw it and the time approximately a year later when your fraternity brother got involved in the computer center?

A: No, I didn’t.

This second encounter supposedly led to more engagement with computer games at Utah and then to a senior thesis in 1967 in which he first put forth his ideas to create his own video game system that could be played at an amusement park. Again, from the deposition:

Q: Did the paper include any description of the types of games that might be played on it?

A: Yes, it did.

Q: What kind of games were described?

A: Space War.


A: Hangman, which is a word game. […] A baseball game.

Q: Any other games?

A: I think those were the only three that I described.

As with his claim of being introduced to Spacewar! by the elusive Jim Davies, Nolan was unable to provide any proof that this senior thesis ever existed. Per the March 1976 deposition:

Q: Mr. Bushnell, you testified about a paper that you prepared while at the University of Utah on games. Have you made a further effort to find a copy of that paper?

A: No, I haven’t.

Q: I think you said the paper was done in conjunction with a course of Professor Atwood; is that correct?

A: I believe that’s true.

Q: have you contacted Professor Atwood as you indicated I think that you were going to do?

A: No, I haven’t.

Q: Do you have any better information today than you had at the time of your last deposition session with respect to that paper or its contents or its location at the present time?

A: No.

I’ve thrown a lot of quotes at you, so this is probably a good time to summarize the historical record so far. In early 1973, Nolan Bushnell indicated that he first saw Spacewar! at the Stanford Artificial Intelligence Laboratory. Later that year, Magnavox publicly declared it was looking to exploit its patents by extracting licensing fees from all the companies making coin-operated video games, including Atari. As the patents were filed in 1969, an individual or company looking to avoid liability in a patent suit would need to prove that their video game activities started before that date. Nolan did not see Spacewar! at SAIL until spring 1970 at the earliest, however, a timeframe established in another part of his deposition. This would not do.

In late 1973, Nolan suddenly starts talking about all the game ideas he had at the University of Utah after seeing Spacewar!. In depositions given in 1976, he further explains that he first saw the game in 1965, first seriously engaged with it in 1966, and wrote a senior thesis all about playing computer games on a system tailored for use in amusement parks in 1967. If true, this new narrative would give his work priority over the work done by Ralph Baer and Bill Rusch. However, when pressed for details by the attorneys, Nolan was as vague as possible on the names of the people involved and the computer system on which he played the game. He was also unable to provide any documentation proving that he saw Spacewar! at Utah during the period in question or that his thesis on playing computer games in an amusement park ever existed.

That about does it for Bushnell’s own words on the matter and why he might have an incentive to distort the truth, but it still leaves us with the fact that he did testify under penalty of perjury that he first saw Spacewar! at Utah. So would he really lie under oath? Well, his answers were vague enough that he would probably avoid a perjury charge, and indeed the lawyers did not seem particularly interested in trying to catch him in a perjury trap, content to let his vague and flimsy recollections stand on their own. After all, if he was lying, his sparse recollections backed by zero documentary evidence would be unlikely to convince the judge in the case — the matter was not being decided by a jury — that Bushnell really was designing video games in the 1960s, but if he was actually telling the truth and an investigation did uncover additional documentary evidence, then it could be problematic for the Magnavox case. For the lawyers, it was far better to just let it be.

But would Nolan lie about his role in early video game history with or without the threat of legal consequences? Sadly, we know the answer to that question is almost certainly yes. While there is no conclusive proof that Nolan’s Utah statements were false since it is nigh on impossible to prove a negative, we do have other examples of him not being completely honest.

The prime example of Nolan’s aversion to the truth comes from another aspect of the Magnavox case: to wit that he saw the Magnavox Odyssey in Burlingame in May 1972 when it was being demonstrated ahead of release. Nolan admitted as much in his deposition — it was hard not to when Magnavox presented as evidence a guestbook from the event with his signature in it — and these days he even admits publicly that he told Al Alcorn to do a ping-pong game as a test project because he saw the Odyssey. He even tries to downplay any controversy thereof by claiming it was no big secret. For example, in a post he made on the Atari Age forums in 2010, Bushnell stated:

I saw the Burlingame demo of the odyssey and thought it was crap and it was. I signed the register with my own name and have never denied it. It did spark the idea that the ping pong idea could be an interesting game if it were done well.

However, while Nolan did cop to the visit in his deposition, in public he denied it for decades. From Video Invaders by Steven Bloom in 1982:

“Pong was no coincidence,” Baer says firmly. “Later on in the mid-’70s, when we negotiated with Atari to get them under license, it came out that somebody over there had actually seen Odyssey sometime during the course of 1972. I don’t know how they did it, but they saw it. So, Pong was a derivative of Odyssey – not the other way around, by any means. The coin-op games are derivative of what we did here back in the ’60s!”

Surprisingly, Bushnell barely counters Baer’s assertions. “It’s really hard to say,” he replies. “I think he can say that even though I had not seen an Odyssey game at that time. But if you do look at the time frame, Pong was actually on the market before Odyssey. I remember being quite surprised to see Odyssey.”

And nearly two decades later in Steven Kent’s The Ultimate History of Video Games:

What they’ve always alleged was that there was a meeting or a distributor show somewhere in the valley, and I should have, would have, could have been there. So it’s one of those pissing matches.

Nolan was never honest about this until Ralph Baer wrote his own book and included a copy of the guestbook signature for all the world to see.

The infamous guestbook, as presented in Video Games: In the Beginning by Ralph Baer

A second fabrication from these early days concerns Bushnell’s own history, or lack thereof, creating games before Computer Space. As far as I know, Nolan still maintains to this day that he was not just playing games at the University of Utah, but also creating his own. In particular, he is proud of a Fox & Geese game. Per another post in 2010 at the Atari Age forums:

Did you program a game in college called Fox and Geese? Could you tell us about the game and the experiance [sic] of programming the game?

I programmed several games but the one that played the best and was sent around the country was Fox and geese. In the game the fox could move in any direction at twice the speed of the geese there was only one fox. The geese could move up to the right or the left. If a goose was alone the fox could eat it. If there were two geese adjacent he could not. The object was for the fox to either eat all the geese or escape past them. The geese won if they trapped the fox against the upper part of the screen. The number of geese was variable. 3 geese almost always lost 7 geese always won.

Bushnell has also described creating this game to Kent and other interviewers as well. Under oath in 1976, however, a different picture emerges:

Q: Do you recall any games other than Space War specifically at the University of Utah?

A: Space War was really the only game I was interested in at the University.

So where does this Fox & Geese story come from if he was not involved with any other games at Utah by his own admission? The deposition answers this question as well:

Q: Do you recall seeing or playing any other games played with a computer and a cathode-ray tube display either at that conference [the Spring or Fall Joint Computer Conference circa 1969] or at the AI project or at the University of Utah prior to the time you left your employment at Ampex?

A: Yes.

Q: What other games do you recall?

A: There was a game called Fox & Geese. That was a game where the fox would chase geese and the geese would, you know, run away from the fox and you would attempt to trap a goose in a corner and eat it. I think the geese were hollow circles and the fox was a filled-in circle. That was definitely at the AI project [SAIL].


Can you fix the time of the fox and geese game any closer?

Q: A: I’d say somewhere between the spring and summer of 1970.

Q: How do you fix that time?

A: I fix it pretty much coincident with when Jim Stein was working there, who was my key to the door to get into the place.

So not only did Nolan not program any games at Utah by his own admission nor play any games there other than Spacewar!, but the Fox & Geese game was just a game he saw at SAIL. Notice this portion of the testimony also establishes he was not visiting SAIL before 1970, hence the need to concoct earlier encounters with the game at Utah to avoid patent infringement.

So now we have established that Nolan both had a motive to stretch the truth and a history of doing the same. What other evidence can we bring to bear?

Well, first of all, we can establish just what computer hardware the University of Utah owned during the events in question. If there was no system capable of playing Spacewar! between 1965 and 1968, the years Nolan attended the institution, then he could not possibly have played it there. Fortunately, Utah has put many of its internal university records online. One such record is a report made by David Evans in November 1966. At the time, Evans was just beginning the research that would put Utah on the cutting edge of graphical technology and provide breakthroughs in everything from polygonal rendering to virtual reality. According to the report:

The Computer Center has replaced the IBM 7044 computer by a UNIVAC Type 1108 computer and has undertaken the system programming required to interface with the graphics laboratory.


The major computing facility is the 1108 system, which was delivered during November and is undergoing hardware and software checkout. It is expected to be in service beginning early in December.

This report is interesting for two reasons. First, it does corroborate Nolan’s claim in his deposition that the Computer Center changed hardware while he was there and upgraded to an 1108 system. They upgraded from a 7044 rather than a 7094 as Bushnell stated, but that can easily be explained away by screwing up the similar model numbers in his head. More importantly though, it confirms that the only computer in the Computer Center in 1965 and 1966 when Nolan claimed to have played Spacewar! was an IBM 7044. As explained by Atari historian Marty Goldberg in a since-deleted blog post from 2014 in which he delved into some of the same issues, the 7044 was an older machine that could not be interfaced with a graphics terminal or monitor; it was strictly a teletype system. Therefore, it is simply impossible for Bushnell to have played Spacewar! in the computer center in 1965 or 1966 as he claims.

For the sake of argument, let’s say Nolan was mistaken about the timeframe. Could he have actually seen the game on the UNIVAC 1108 before graduating college in 1968? Again, this does not appear possible. While the 1108 was hooked up to a PDP-8 to create a graphical system, Marty Goldberg and a graduate teaching assistant at the University of Utah named Sarah Bell did a deep dive into David Evans’ project reports from that time period and learned that by the end of 1968, Evans was still struggling to get basic shapes up on his kludged-together system, making it highly unlikely anyone was running Spacewar! on it. Utah was on the brink of becoming a major center for advanced graphics research when Nolan Bushnell graduated, but it was not quite there yet.

Another strike against Spacewar! being at Utah in the 1960s is that no other faculty member or alumnus has ever claimed the game was playable at the university. In fact, in 2020 my colleague Ethan Johnson specifically asked noted Utah alum and Spacewar! enthusiast Dr. Alan Kay about his first encounter with the game. Kay stated he saw the game while a graduate student at Utah, but he had to visit MIT to do so. Kay was at Utah from 1966 to 1969, worked closely with Evans, and knew Spacewar! well enough that Stewart Brand quoted him on the influence of the game in his seminal 1972 article for Rolling Stone, yet even he had to travel elsewhere to see the game.

Finally, the game never appears in the University of Utah student newspaper, which is available online for the years in question. At both MIT and Stanford, the student newspaper reported on the game soon after it appeared, but the Utah newspaper is completely silent on any game playing on Utah computers before the 1970s.

Articles in the MIT student newspaper, The Tech, and the Stanford student newspaper, The Stanford Daily, announcing the arrival of the hot new Spacewar! game on April 25, 1962, and May 16, 1963, respectively.

Finally, I will leave you with another piece of evidence that I discovered in my own research. In describing his second encounter with Spacewar! in 1966, Nolan Bushnell states the following in his January 1976 deposition:

Q: What did you do as a result of your thinking that it would be fun to program games?

A: Well, I asked for a listing of the current Space War game, I think I wanted to understand how they had done what they had done, you know, and made some modifications.

Q: Who did you ask for this?

A: Randall Willey.

Q: Who was Randall Willey?

A: He’s the fraternity brother.

Q: How do you spell Willey?

A: W-i-l-l-e-y, I think.

So here we have a name of someone else involved in the supposed Spacewar! scene at Utah, one that feels far more tangible than Mr. Jim with a D. As soon as I first saw this deposition back in 2015 or so, I honed in on Mr. Willey as a person that might be able to shed some further light on the situation. Unfortunately, he appeared to have vanished without a trace.

Flash forward to 2018, when I realized the Utah student newspaper was available online. A search within revealed that Nolan was wrong: his fraternity brother’s last name was actually spelled W-i-l-l-i-e. Armed with the correct spelling, I soon managed to locate Mr. Randall Willie and conducted a phone interview with him. The following exchange comes from that interview:

Q: So at that early stage in the Utah Computer Center were there any games there, was any game stuff going on?

A: There might have been, I don’t recall a lot of that.  I don’t recall any games going on.

While its not an outright denial of any games at the university whatsoever, it is a complete denial of his own involvement with such. Its hard to believe that someone so enmeshed in that scene that according to Nolan he knew where to go to get game code would have no recollection of any computer games after the fact. It appears we may have caught Bushnell red-handed, though he needn’t fear, as the statute of limitations for perjury has long since expired. Regardless, while it is impossible to completely disprove a negative, I think the weight of the evidence speaks for itself. Nolan Bushnell did not encounter Spacewar! at the University of Utah in the 1960s; he encountered it at Stanford University in 1970.

They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I 1971-1982 is available in print or electronically direct from the publisher, CRC Press, as well as through Amazon and other major online retailers.


Worldly Wednesdays: The Cutting Room Floor

This post is part of an ongoing series annotating my book They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I. It covers material found in chapter 5 on page 68. It is not necessary to have read the book to comprehend and appreciate the post.

This week, as we reach Chapter 5 and prepare to delve into some of the first real controversial material that I had to wade through regarding the respective roles of Nolan Bushnell and Ted Dabney in the early days of Atari, I want to pause for a brief moment to discuss the writing process of the book a little bit and share a little material that did not make it into the final draft.

There is surprisingly little cut content in They Create Worlds. The goal of the book all along — a goal that my publisher supported — was to be an “everything but the kitchen sink and maybe that too” look at the history of the video game industry. Obviously, this does not mean that every last fact about every last person or game was going to be shoved into the thing indiscriminately, but it did mean that size was probably not going to be an issue. I had as many as 600-620 pages to play with in volume I, and I ended up coming in comfortably under that count.

The volume that I completed was actually my third attempt to write a book on video game history. I started this project way back in 2006 after reading Rusel DeMaria and Johnny Wilson’s High Score, 2nd Edition (my first video game history book, acquired and read in 2004) and Steven Kent’s The Ultimate History of Video Games. At the time, I was new to the topic and took both tomes at face value, but I was frustrated by their limitations. High Score was a coffee table book with a nice layout and interesting pictures, but was comparably light on narrative. This satisfied the objectives of the book’s authors, but not my personal desire for in-depth knowledge. Despite this, it had a lot of great information on early computer game companies like Sierra, Brøderbund, and Electronic Arts. Steven Kent’s book was narratively longer and richer, but it ignored computer games almost entirely. What if, I naively thought, I could smash the two books together, add a little embellishment from a website or two, and create something more comprehensive?

High Score and The Ultimate History of Video Games: Two great tastes that taste great together?

At first, I was not sure what that something more comprehensive would be. Probably not a book at this point, but maybe a Wikipedia article? I started toying around with a structure, but it quickly became apparent that what I was trying to do would result in the longest Wikipedia article in history by a fair margin, so I scrapped that idea pretty quickly. Then, I started writing what could generously be called a book. I still have that draft, last modified April 7, 2008, and no you cannot see it. Its 263 single-spaced pages in Microsoft Word and goes up to about 1998, which is a reflection on when I abandoned it as opposed to a cutoff in years. It really is just a mashup of High Score and Ultimate History of Video Games with a few tidbits from the Internet. There are no copyright violations — the words are entirely my own — but its incredibly shallow. As I did more research and learned more about the industry, I realized this would not do.

So I started over. By now, my research was becoming more sophisticated as I mined online newspaper databases and online interviews, and even conducted a few interviews of my own in 2009. I took the original manuscript and started rewriting it section-by-section to go into more depth and rely more on primary sources than Steven Kent. The good news was I felt it was morphing into an original story with new insights rather than kludging together the work of others. The bad news was that I had settled on a level of detail that I knew would never result in a single volume of publishable size. In 2013, this draft was abandoned as well (and no, you can’t read it either).

At this point, the idea of the three-volume history took shape. This approach would allow me to maintain the level of detail I wanted while not having a single work ballooning to 2000 pages. This time, I started over completely from scratch rather than rewriting a previous draft on the fly. I did not outline the work, but I had a vague idea volume one would take me through the crash. I also believed, laughably now, that the first volume would be in two roughly equal parts divided by the launch of Space Invaders. This was due less to a lack of imagination than to a lack of sources on 1970s video game history. As my research deepened, Space Invaders was pushed later in the text.

One advantage of my years of false starts was that by the time I started attempt number three, I had a good sense of the scope of the work before me and the level of detail I wanted to achieve. Therefore, the book changed very little between the first and final draft. Grammar and mechanics were tightened, footnotes were polished up, and the occasional late discovery such as Bouncing Ball as computer game was snuck in, but very little material ended up on the cutting room floor.

Most cuts were small and resulted from doubts that I really had the sources to back up a claim. A good example of that comes from chapter 3 and my discussion of the Nimrod computer. Originally, when discussing the exhibition of the computer, I wrote that Nimrod “premiered at the festival on May 4, 1951, and remained on display until the exhibition closed in October, after which it was displayed for three more weeks at the Berlin Industrial Show and made a stop in Bertie’s hometown of Toronto before being dismantled.” That last part about Toronto is no longer in the book. It was a claim I found on the Internet in a couple of places, but as I finalized my manuscript, I realized I had no good proof this had ever happened from contemporaneous primary sources, so I took it out. Of course at that point, the oft-invoked Ethan Johnson was poking around and found two sources reporting on Nimrod’s triumphant arrival in Canada. So that piece of cut content has transformed into a piece of errata. Such is life.

Nimrod, the second game computer demoed in Canada during its brief heyday as the computer game capital of the world.

Outside these small tweaks, there were only two large passages cut from the book. The first appeared in Chapter 5 and interrupted Nolan Bushnell’s story to provide a brief history of Silicon Valley. The material was not bad, but it was not exactly on point with the story being told in the chapter. Furthermore, its a story that is more about the development of the American technology sector rather than video games specifically. While many video game companies have called Silicon Valley home, the story of how the region slowly became a technology hub has little to do with video games and more to do with vacuum tubes, transistors, computers, and the Internet. Its a topic that deserves its own book, and my little summary did not really do it justice. Unlike my early book drafts, I will share this material in this blog post. I will probably do the same with the second large passage when we get there. For those keeping score at home, this section would have started on page 68 after the paragraph that ends with the sentence “Anxious to leave Utah, Nolan travelled to Northern California shortly before graduating in December 1968 to look for work among the high concentration of technology companies in a region that would soon be christened “Silicon Valley” due to the large number of semiconductor manufactures in the area”

In the 1920s, the San Jose Chamber of Commerce christened California’s Santa Clara Valley the “Valley of Heart’s Delight” to highlight the idyllic pastoral setting dotted with orchards full of plum and apricot trees that had become the largest fruit production and packing region in the world.  Even at this early date, however, the region had already experienced its first brush with high technology.  In 1885, businessman Leland Stanford decided to establish the Leland Stanford Junior University as a tribute to his teenage son, who had died the year before.  Opening for its first term in 1891 at a campus halfway between the cities of San Francisco and San Jose, Stanford University aspired to become the “Harvard of the West” and looked to attract top talent across all academic disciplines.  The study of electricity proved especially important to the region, as California found itself in the middle of a population boom in the early twentieth century and needed to electrify rapidly by transmitting power over greater distances than required on the East Coast.  Therefore, the Stanford electrical engineering school, established in 1894, worked closely with local businesses to develop better techniques for long-distance electric power transmission. 

Electric power transmission gave way to wireless communication in 1909 when former Stanford electrical engineering student Cyril Elwell established the Federal Telegraph Company  in San Francisco to commercialize the arc transmitter, greatly improving the efficacy of wireless transmissions.  The next year, Elwell hired Lee DeForest, whose work with vacuum tubes and electrical signal amplification at Federal Telegraph and elsewhere would prove instrumental in the development of not only the wireless industry, but also both the first national telephone network and early radio broadcasting.  Elwell and DeForest presided over a brief period of technological dominance in the San Francisco Bay Area that ended after World War I when the Federal government decided a nationwide wireless communication network was too important to entrust to a few upstart companies on the West Coast and therefore aided General Electric in acquiring most of the important wireless patents, after which GE established a new public company in 1919 called the Radio Corporation of America (RCA) to control the field.

Meanwhile, Stanford University continued to churn out capable electrical engineering graduates only to see the best of them leave the region after completing their studies to work for one of the big East Coast companies like GE, Westinghouse, Raytheon, RCA, and AT&T.  One professor at the school desired to change that.  The son of a Stanford professor himself, Frederick Terman earned a bachelor’s degree in chemical engineering and a masters in electrical engineering from the university before heading east to study with Vannevar Bush at MIT.  After Terman received his PhD in electrical engineering in 1924, he was offered a job at the Institute and would have most likely remained on the East Coast like so many electrical engineers before him if he had not suffered a string of serious illnesses while visiting his family back in California that led him to take a part time teaching position at Stanford instead.  Over the next thirty years, he rose from professor to electrical engineering department head to dean of the School of Engineering and finally to University Provost and vice president all while significantly improving the reputation of the school so as to attract top students that he worked to keep on the West Coast.

Even after GE and RCA took over the bulk of the US wireless industry, the Bay Area continued to house a small vacuum tube manufacturing industry spearheaded primarily by two firms, Eitel-McCullough and Litton Engineering.  Terman decided to harness this industry to transform Stanford into a center for vacuum tube research and therefore enticed Litton Engineering founder Charles Litton to join the electrical engineering faculty in 1936.  Litton, who specialized in manufacturing equipment used to shape the glass found in vacuum tubes, helped establish a vacuum tube laboratory on campus that attracted significant attention from East Coast companies and also provided a grant that helped Terman bring two of his favorite students back to California, Bill Hewlett and Dave Packard.

Without Terman’s intervention, Hewlett and Packard would have almost certainly departed the West Coast like countless electrical engineers before them, as Packard, who graduated Stanford in 1935, immediately took a job with GE, while Hewlett, who graduated a year earlier, matriculated to MIT to pursue a master’s degree in electrical engineering.  Armed with a $1,000 grant from Litton, however, Terman lured Hewlett back to Stanford after he completed his master’s in 1936 for a project to build a new type of oscillator.  Soon after, he brought Packard back on a leave of absence from GE to assist in the project, which led the two friends to explore going into business together.  On January 1, 1939, the duo established an electronic test equipment business called the Hewlett-Packard Company (HP) in the garage of a duplex they were renting.  Today, this event is considered the symbolic birth of Silicon Valley.

In the late 1930s, Stanford’s increasing expertise with vacuum tubes resulted in the development of a device called the klystron in 1937 by brothers Russell and Sigurd Varian, the first high-power vacuum tube that could amplify signals in the microwave range and therefore proved crucial to the emerging technology of radar.  During World War II, the Varian Brothers, Charles Litton, and Fred Terman consequently all ended up at various East Coast laboratories that were working with the new technology.  At the war’s conclusion, Litton and the Varian Brothers returned to California and each established a microwave tube manufacturing company, Litton Industries (1946) and Varian Associates (1948), transforming Northern California into a leading hub for the technology. 

Terman, meanwhile, returned to Stanford as the dean of the engineering school.  Realizing that government funding for scientific research would only increase after the key role science had played during the war, he initiated a master plan to secure government money for projects by Stanford students who could then continue their work through the growing private industry base in the region.  He decided to focus these efforts on microwave tubes, which continued to be important in the sophisticated radar and electronic countermeasure projects being developed by the United States military.  He therefore joined the board of directors of Varian and helped secure lucrative contracts for Litton Industries.  Terman’s efforts culminated with the establishment of the Stanford Industrial Park in 1951, a university owned office park catering to technology companies.  Hewlett-Packard and Varian Associates became early tenants.

The final important piece of the Santa Clara Valley technology puzzle was the arrival of Shockley Semiconductor in 1955, which established itself in Palo Alto so William Shockley could be close to his mother.  When the “Traitorous Eight” formed Fairchild Semiconductor two years later, they located their company in Palo Alto as well.  In the 1960s, Fairchild spawned its own group of spinoffs dubbed the “Fairchildren,” thus cementing the region’s role in the semiconductor manufacturing business.  This led to the moniker “Silicon Valley,” which first appeared in print in an article written by Don Hoefler in January 1971 for Electronic News.  For an ambitious young electrical engineer like Nolan Bushnell, there was no better place to seek employment.

They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I 1971-1982 is available in print or electronically direct from the publisher, CRC Press, as well as through Amazon and other major online retailers.

Worldly Wednesdays: A Timeline of Spacewar!

This post is part of an ongoing series annotating my book They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I. It covers material found in chapter 4 on pages 43-59. It is not necessary to have read the book to comprehend and appreciate the post.

After three chapters depicting a motley assortment of AI experiments, military simulations, and demonstration programs, chapter four finally arrives at the inflection point for the birth of commercial video games: the creation and proliferation of Spacewar!, conceived by Steve Russell, Martin Graetz, and Wayne Wiitanen and largely programmed by Russell with support from Graetz, Peter Samson, Dan Edwards, Alan Kotok, and other hackers hanging around the PDP-1 at MIT. While Spacewar! itself was not commercialized in its original incarnation, its direct influence on the first commercially released video game, Computer Space, and the first significant video game company, Atari, give it pride of place in any history of the video game industry. While the book already discusses the creation of the game, this annotation will explore what we know of the chronology of the game’s creation in a little more detail.

Before the last few years, there were three significant sources discussing the birth of Spacewar! The first was an article written by famed Whole Earth Catalog creator Stewart Brand for Rolling Stone and published in October 1972. While the intersection of computers and counterculture is somewhat overblown in some monographs discussing the early days of personal computing, there was significant collision between these two forces in the San Francisco Bay Area and the adjacent Santa Clara Valley, where the summer of love and the free speech movement mingled with the prestigious technical degree programs at UC Berkeley and Stanford. Brand himself was Stanford educated and embraced the counterculture values of individual empowerment as a tool for social justice and ecological renewal. The Whole Earth Catalog was one aspect of this empowerment drive, a how-to guide and catalog of useful items for self-sufficient living off the land, which was a key tenet of the commune movement.

For Brand, liberating the individual from the shackles of modern society was about more than farming and communal living; he also felt it important to expand the mind. An associate of Ken Kesey, Brand experimented with LSD and other mind-altering drugs in the 1960s as one means to this end, but by the 1970s he became convinced computers possessed even more potent mind-altering powers. To Brand, learning to use a computer was another path towards taking control of one’s own destiny, and nothing showcased the power of a computer like a game. As he put it in a 2016 retrospective about his 1972 article: “I saw them having some kind of out-of-body experience. Their brains and their fingers were fully engaged. There was an athletic exuberance to their joyous mutual slaying. I’d never seen anything like that.” Of course, he was referring to people playing Spacewar!

Playing Spacewar! may just be an out of body experience

By 1972, Spacewar! had become so enmeshed in university electrical engineering and computer science departments, government think tanks, and corporate research facilities that no less a personage than Alan Kay, then at Xerox PARC, opined that “The game of Spacewar blossoms spontaneously wherever there is a graphics display connected to a computer.” The game certainly did not escape the notice of Brand, who encountered it again and again as he toured computing centers in the Bay Area to come to grips with this exciting new technology. In 2016, he still remembered his enthusiasm: “I was intrigued at the quality of game design intelligence these guys had from the very start. You were balancing skill versus luck, and not only dealing with the threat of your opponents, but the threat of losing control and being slurped into the sun. And hyperspace was an astonishingly brilliant breakthrough.”

Brand’s love of the design for Spacewar! echoed his larger appreciation for hacker culture, which brought the same do-it-yourself mentality of the commune to the computer lab. He became determined to document the hacker movement for the general public using Rolling Stone as his vessel, and he chose Spacewar! as his — and the reader’s — entrée into the hacker world. He convinced Lester Earnest, the manager of the Stanford Artificial Intelligence Laboratory (SAIL), to shut down the premises for an evening to hold a competition he dubbed the “Intergalactic Spacewar Olympics” in which an assortment of students and researchers faced off in a multiplayer variant of the game that had recently been created by Ralph Gorin. After a series of 2v2 preliminary matches, a final five-player free-for-all determined the champion. Brand and photographer Anne Leibowitz were on hand to capture all the action, including the triumphant winner of the affair, Bruce Baumgart, standing next to a terminal with a huge grin on his face as he celebrated his victory.

Bruce Baumgart celebrates victory

While Brand focused most of his article on what may have been the first esports competition, he did spare a few paragraphs for the creation of the game, complete with quotes from Steve Russell. Brand captured the barebones story that Russell created the game in 1961-62 at MIT on a PDP-1 after being inspired by E.E. Smith’s Lensman novels, that Peter Samson added the “Expensive Planetarium,” and that Dan Edwards and Alan Kotok also pitched in, but he missed the equally important contributions of Graetz and Wiitanen to the whole affair. That would have to wait for the August 1981 issue of Creative Computing, in which Martin Graetz himself took a stab at detailing the origin of the game.

Of all Spacewar!‘s many fathers, Graetz was the most literary minded, having been an avid reader and writer of science fiction since his secondary school days and even managing to have a story published in a pulp magazine at one point. Therefore, he took on the mantle of semi-official Spacewar! historian. As Graetz revealed to me, he was an employee at DEC when company co-founder Ken Olsen set up a small company museum featuring a few old computer systems that soon after moved to a public exhibition space in Malboro, MA, and served as the genesis of what is now the Computer History Museum in Mountain View, California. Graetz helped restore one of the few surviving PDP-1 computers for the museum and made sure Spacewar! ran on it. He also contacted David Ahl at Creative Computing and pitched the idea of doing an article on the history of the game to commemorate the twentieth anniversary of its inception in 1981. Ahl loved the idea and commissioned the article.

In “The Origin of Spacewar,” Graetz gave the first complete telling of the Spacewar! story, which he based not only on his own recollections, but also on the memories of his fellow hackers and several of MIT’s computer custodians like Jack Dennis and John McKenzie. Here we learned for the first time of the “Hingham Institute” consisting of Graetz, Russell, and Wayne Wiitanen, roommates and co-workers at Harvard’s Littaeur Statistical Laboratory who were reading E.E. Smith novels and watching special-effects-laden b-movies out of Japan from Toho Studios. As Graetz tells it, he moved on from Harvard to take a job with Jack Dennis at MIT in summer 1961, at which point he learned DEC would be donating a shiny new PDP-1 computer to the university in the fall. Wowed by the graphical demo programs already extant on the forerunner to the PDP-1, the prototypical TX-0 housed in MIT’s Research Laboratory of Electronics (RLE), he convened the Hingham Institute to do them one better.

In Graetz’s telling, Wiitanen becomes the real hero:

With the Fenachrone hot on our ion track, Wayne said, “Look, you need action and you need some kind of skill level. It should be a game where you have to control things moving around on the scope, like, oh, spaceships. Something like an explorer game, or a race or contest…a fight, maybe?”

“SPACEWAR!” shouted Slug and I, as the last force screen flared into the violet and went down.

With that, Graetz informs us that Russell returns to MIT from Harvard, tells everyone about this neat demo idea, and finally gets to work on the game after Alan Kotok provides some sine-cosine routines directly from DEC that imploded the last major excuse Russell had been clinging to so he did not actually have to put in the work. Then Samson adds his starfield, Dan Edwards is given credit for bringing gravity to the match, Graetz himself programs a hyperspace feature, and Kotok and Bob Saunders, members of some organization briefly identified as the Tech Model Railroad Club, create some custom control boxes to make it all much easier to play. By an MIT open house in May, Spacewar! is done and makes its public debut.

Alan Kotok, Steve Russell, and Martin Graetz prepare to play a game of Spacewar! at a 1983 Computer Museum event.

Those wanting further insight into this “Tech Model Railroad Club” (TMRC) would have to wait three more years when the third and final significant source on the creation of Spacewar! was published: Steven Levy’s Hackers: Heroes of the Computer Revolution. Taking up where Brand left off, the book attempts to chronicle the birth and spread of hacker culture from the insular nerds at MIT to the more open counterculture crowd in the San Francisco Bay Area, and finally to the business-minded hackers like Ken Williams and Doug Carlston who established the computer game industry. While his version of the creation story largely follows Brand and Graetz, he spends considerable time fleshing out the history and motivations of the TMRC hackers and provides profiles on some of the supporting characters in the Spacewar! story like Samson, Kotok, and Saunders. While Kotok and Samson in particular are elevated in this account, Graetz and Wiitanen fade once again into the background, presumably because they were never really part of the TMRC hacker scene that Levy is so eager to chronicle. While the text indicates an awareness of the Hingham Institute, Levy joins Brand in focusing Russell as the key figure in not just programming the game, but also in conceiving it.

Hackers and “The Origin of Spacewar” defined the Spacewar! story in the historical record for the next thirty years. Steve Bloom’s Video Invaders, which predates Hackers, essentially regurgitates the Creative Computing story that had just come out the year before. Steven Kent’s Ultimate History of Video Games cribs from Hackers, supplemented by Kent’s own interview of Russell. As is typical, however, Kent cannot help but introduce a tiny error in the narrative when he refers to Russell’s love of “Doc Savage” stories when other sources clearly demonstrate this should have been a reference to the Lensman books written by Doc Smith. Rusel Demaria and Johnny Wilson return to the Creative Computing story in High Score as befits their own close affiliation with early computer magazines, while Tristan Donovan’s Replay hews closest to Hackers with its emphasis on Russell and his place in the larger TMRC hacking scene.

While the Brand, Graetz, and Levy renditions of the origin of Spacewar! place their emphasis on different people, they largely agree on chronology. In this tale, word of the PDP-1’s imminent arrival comes in the summer of 1961, the computer arrives in the fall, and Steve Russell finally gets to work on the game in December. In early 1962 other programmers like Samson and Edwards add a host of new features like an accurate starfield and gravity, and the whole affair is wrapped up in time for an MIT Open House in May. This general account, based entirely on the memories of the participants anywhere from ten to twenty years after the fact, appears to be generally correct, but as more documentary evidence has become available in the past few years, this timeline has been tweaked.

The first breakthrough in cracking open the Spacewar! chronology came from Austrian web developer Norbert Landsteiner. Around 1996, Landsteiner became a pioneer in Java game programming, releasing takes on several classic arcade games like Space Invaders and Pac-Man that could be played within a browser. In the early 2010s, he began collecting paper tape and original code of Spacewar! variants to recreate them on his website, and managed to get his hands on some of the earliest extant code for the game. In particular, he made the remarkable discovery of Spacewar 2B, which is now the earliest known “complete” version of the game, i.e. the version that appears to have all the pieces present at the reputed public debut of the game in May 1962. Two versions were discovered, one dated March 25, 1962 pulled off an original paper tape, and one dated April 2, 1962 reconstructed from .bin files. The two versions are largely identical save for a few minor settings parameters. These files show that the renderings of the ships, the basic controls, the gravity of the sun, and the starfield were all implemented as of March 25. They further give an exact date for the integration of Samson’s Expensive Planetarium, patched in according to the comments on March 13, 1962. There is also a space carved out to “put [in] more bells and whisles [sic], like hyperspace,” but the functionality is not actually present. Landsteiner did also discover the separate hyperspace patch in a version dated May 2, 1962.

A rendition of the “CBS Opening,” named because the path of the ships vaguely resemble the CBS logo, which illustrates the gravitational pull of the central star.

Landsteiner’s work has been invaluable for determining when the original version of Spacewar! reached a finished state, but it did little to enhance or corroborate the narrative around when the game was conceived and when programming began. Narrowing down the first point is where my own work took center stage. In 2015, I tracked down Wayne Wiitanen on the Internet and emailed him several questions that he was kind enough to answer. To my knowledge, this was his first contribution to the historical record outside his uncited contributions to the Graetz article, though he has given at least two interviews since. I had little hope he would be able narrow the timeline much after over fifty years, but I was proven wrong. As Wiitanen explained it: “It had to be summer, but before August 10th (the date when my recall orders issued) as I had to report to Ft. Bragg in October and didn’t have time to get involved in detailed design or programming.” So while we still don’t know exactly when in the summer the brainstorming session occurred, we do know it happened before August 10th, a hard date fixed by Wayne’s recall orders. This matches up with the recollections of both Graetz in his article and Alan Kotok in Hackers that the MIT community became aware of the impending computer donation at that time.

So when did Russell’s work on the game actually begin? Russell himself in a 2008 oral history with the Computer History Museum remembers the development environment, including display, being in place by fall 1961 and development beginning before the end of the year. Hackers further pinpoints the date to early December 1961. Graetz states “Slug produced the first object-in-motion program in January 1962,” which does not explicitly contradict the early December start date, but does cast it into serious doubt. Graetz also makes another curious claim that no other person, in articles or oral histories, has ever made, which is that the display “was scheduled to be installed a couple of months after the computer itself.”

This remained the state of our knowledge on the matter until 2017, when fellow researcher and frequent collaborator Ethan Johnson made one of his trips to the National Archives facility in Chicago. This facility is home to much of the original Magnavox patent suit, which was adjudicated in federal court in Chicago. As discussed previously, the defense to this lawsuit resolved around invalidating the Baer and Rusch patents through proving the existence of “prior art” in video games, which naturally made Spacewar! one of the focal points of the trial. Representing MIT in this case was John McKenzie, the engineer charged with keeping the TX-0 and PDP-1 in working order at the time of the events in question. Mr. McKenzie’s deposition in the case, taken in 1975, was still extant in the court records when Ethan pulled the files. Not only did McKenzie testify to his own personal knowledge of events, but he also brought with him the log books for the PDP-1. As time on the computer had to be signed out, this meant that he had with him a complete written record of who did what with the computer throughout 1961 and 1962!

John McKenzie (seated right) in 1983. Seated next to him is Martin Graetz, and next to Martin is Jack Dennis. Alan Kotok stands behind them (l)

The main dates McKenzie’s testimony firmly establish are the dates both the computer and its display were delivered. According to the logs, the computer arrived September 15, 1961, and the monitor was installed on December 29, 1961. So Graetz was right: the monitor really was installed a few months after the fact! This calls into question large portions of Russell’s recollections, namely that he was partially inspired to finally start creating the game by the so-called “Minskytron” graphical demo program and that he started coding the game in 1961. The Minskytron could not have existed before Marvin Minsky had a display to actually program, and Russell himself is unlikely to have begun programming in December without a display to work with. This appears to confirm the Graetz account that programming on the game commenced in January 1962.

Later on, McKenzie provides another gem:

McKenzie: On page 8, the middle of the page, it says “Monday 19 March 1962. Clock equals 2694. 7.” Some more beyond that, at 0345, “Power off. installed user’s IOT input to I/0 on IOT11.” That was initialed RAS, which would have been Robert Saunders.


Q: Does that entry have a meaning to you?

McKenzie: Yes; very significant. The students built up two control boxes; and at this time the control, some became optional with a sense switch setting, determining whether you wanted to take input from the earlier mentioned test word switches or from the control boxes. These control boxes, the state of the switches in the control boxes was, the term — the computer term is strobed or brought into the computer on the execution of an IOT11.

So thanks to McKenzie’s meticulous record keeping, we know the custom control boxes were installed on March 19! Note the installation was done by Bob Saunders. In “The Origin of Spacewar” Graetz credits the boxes to both Saunders and Kotok, but in a 2018 oral history with the Smithsonian, Saunders took sole credit for developing the boxes. The log book does not necessarily corroborate Saunders’ version of events, but it does provide support for the idea that at the very least he took the lead on this project even if Kotok helped him scrounge up some parts.

Artist’s Rendition of the original Spacewar! controllers, which have sadly been lost

McKenzie also remarks on another aspect of the game’s creation that has nothing to do with chronology, but is worth mentioning here. In describing the people behind Spacewar!, McKenzie says:

Some of the people who were most involved would have been Peter Samson, Daniel Edwards, Alan Kotok. Steven Russell I did not see as frequently. He was a Harvard student, and was not around during the day. I knew of him, I had met him; but not frequently.

Note he identifies Russell as being affiliated with Harvard. He was not a student — that’s a provable error — but this contradicts Graetz’s assertion that Russell had returned to MIT at the time he created Spacewar!. McKenzie’s testimony is not the only source of contradiction on this point, as Russell himself describes his career history in his 2008 oral history:

At some time around, I think it was 1961, late 1961, early 1962, I had gotten bored with working on Lisp and got a job at Harvard. And part of that involved my not having a deferment anymore, so I went into the Army Reserve. […] I had to go off for six months of active duty; I came back, management had changed; new management was not nice, and […] about the time I was convinced that I didn’t want to continue working for Harvard, John was moving to California, to Stanford, and offered me a job at Stanford, which I took. So I ended up at Stanford in the fall of 1962.”

Practically every account of the creation of Spacewar! describes Steve Russell as an MIT student at the time of Spacewar!‘s creation. This claim is so widespread I half expect his tombstone will read “Here lies Steve Russell, noted MIT graduate.” In fact, Steve Russell never attended MIT and never earned a degree from any college. He attended Dartmouth for four years, but failed to graduate because he never got around to writing a senior thesis. In the meantime, he had done some work for MIT professor John McCarthy on the side and secured employment at MIT in 1958 in the newly formed AI Lab. He joined TMRC in 1960 despite not being a student, thus further confusing the historical record, before gaining employment at the Littauer Statistical Laboratory at Harvard by mid 1961 (while Russell is unsure of the sequence of events, he was at Harvard at the same time Graetz was there and Graetz was back at MIT in summer 1961). He stayed there — minus six months of Army Reserve duty — until he departed for Stanford in the Fall of 1962. If you take nothing else away from this blog post, please, please come away with the knowledge that Steve Russell never attended MIT and was no longer even an employee of the university at the time he took the lead on creating Spacewar!

Anyway, the McKenzie deposition appeared to fix the rest of the Spacewar! timeline, and indeed these were the last new chronological revelations that I incorporated into my book. Alas, history is a frustratingly fluid discipline despite being stuck in the past, and Ethan Johnson made an additional discovery after I had locked my manuscript. It turns out the MIT student newspaper, The Tech, is available in its entirety online, and it eagerly announced the impending public unveiling of Spacewar! in the April 25, 1962 issue:

For Parents’ Weekend the PDP1 will run [Spacewar!] for parents and their Techmen between 12 p.m. and 6 p.m., with a special version designed for this purpose. As an example of the whimsical uses of computers this is something parents shouldn’t miss. As a simplistic preview of future war games, it might well be classified were this report to fall into the wrong hands.

At first glance, this does not seem to change anything. In fact, it further confirms that Spacewar! made its debut at that famous May open house that keeps coming up. However, if we delve further into the same issue, we learn that Parents’ Weekend is coming up on Saturday and Sunday. April 25 was a Wednesday, so Parents’ Weekend is actually April 28-29!

This may seem like a minor shift in the timeline, but it has important implications. The surviving version of Graetz’s hyperspace patch is dated May 2, 1962, three days after the Parents’ Weekend concluded. This strongly implies hyperspace may not have been available at the public unveiling. The May 2 revision is not the first implementation of this code if the version number is any guide, so it is possible an earlier version of the patch found its way into the game. This is unlikely, however, because a full description of how Spacewar! works is included in the April 25 article and hyperspace is not mentioned at all. Was it patched in over the next few days? Possibly, but I am inclined to believe it just missed the public unveiling. Alas, my book will need to be corrected.

So now that we have explored all the sources, here is the Spacewar! timeline as best as it can be figured at present:

Summer 1961: Members of the hacker community at MIT, which includes new university employee Martin Graetz, learn that DEC is about to donate a PDP-1 computer to the university. Graetz wants to create a new demonstration program for the machine and enlists his roommates, Harvard employees Wayne Wiitanen and Steve Russell, to brainstorm ideas. Sometime before August 10, they decide to create a game featuring spaceships shooting at each other inspired by their love of the Lensman series of novels by E.E. “Doc” Smith.

September 15, 1961: The donated PDP-1 is installed in the RLE.

Fall 1961: Russell, who still comes around to hang out with his TMRC buddies on evenings and weekends despite no longer working at MIT, starts bragging about the cool program he has dreamed up for the PDP-1 display. Everyone gets excited and starts pressuring him to implement it once the display arrives.

December 29, 1961: The display for the PDP-1 is installed.

January 1962: Steve Russell begins coding Spacewar!, and before the end of the month he has created the first motion object for the game “a dot which could accelerate and change direction under switch control.”

February 1962: Work continues until by the end of the month a barebones version of the game exists consisting of “just the two ships, a supply of fuel, and a store of ‘torpedoes’ — points of light fired from the nose of the ship. Once launched, a torpedo was a ballistic missile, zooming along until it either hit something (more precisely, until it got within a minimum distance of a ship or another torpedo) or its ‘time fuse’ caused it to self-destruct.”

Early March 1962: AI Labs programmer Dan Edwards, who was instrumental in helping Russell get the ships up on the screen, takes it upon himself to add gravity to the game to make it more strategic after Russell pleads lack of skill to make the addition himself.

March 13, 1962: Peter Samson’s “Expensive Planetarium” is patched into the game, providing a scrolling starfield that accurately depicts the actual position and relative brightness of the stars visible from North America [Note: As there is no exact date tied to the gravity addition, its possible that Samson’s contribution came before gravity. I have placed them in this order because Brand’s Rolling Stone article did so. Being written in 1972, it is the earliest source on the matter, meaning memories were presumably freshest.

March 19, 1962: Bob Saunders installs control boxes of his own design on the PDP-1 to make playing Spacewar! less arduous.

March 25, 1962: Spacewar! exists in a largely completed form by this date in a version designated 2B. Ships, torpedoes, gravity, and starfield are all present and operational.

April 2, 1962: Version 2B is slightly modified. Gameplay remains unchanged.

April 28-29, 1962: Spacewar! is played in a public setting for the first time during the annual MIT Parents’ Weekend.

May 2, 1962: Martin Graetz completes a patch adding a hyperspace feature to the game.

They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I 1971-1982 is available in print or electronically direct from the publisher, CRC Press, as well as through Amazon and other major online retailers.

Worldly Wednesdays: The First Real-Time Games

This post is part of an ongoing series annotating my book They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I. It covers material found in chapter 3 on pages 38-42. It is not necessary to have read the book to comprehend and appreciate the post.

So here we are again talking about firsts. Despite two out the first three annotations in this series dealing with the first this or that, who did something first is not really a preoccupation of mine or of the book. It can be fun to research firsts, and people certainly have fun learning about them if the amount of media devoted to firsts on a given topic is any guide, but at the end of the day being first does not say much about the world. As my post on the Father of Video Games should make clear, I am far more concerned with who inspired whom than in who might have done something first in a vacuum. That said, the first three chapters of the book are about the earliest video game experiments, so its only logical that the annotations pertaining to those chapters examine some of these firsts in a little more detail.

The point of this post, however, is not really to get to the bottom of who we should call the “father of real-time games,” but rather to illuminate how my book is a product of not just my own research, but of a wonderful collaboration among professional enthusiasts and amateur historians exploring video game history. Most of these individuals that I interact with are members of the Gaming Alexandria and Video Game History Foundation Discord communities, which I would encourage anyone interested in this history to join even if they are not active researchers or content creators. The GA discord is free to everyone, while the VGHF Discord does require a small donation to gain access. I assure you, the work Frank Cifaldi, Kelsey Lewin, and their team is doing to preserve video game history is worthy of your support. Anyway, on with the show.

The question of who created the first real-time video game, that is a video game in which the display updates quickly and continuously in response to user input to give the impression of seamless action, is of some consequence, for some people would argue that real-time action is what truly separates video games from earlier, pen-and-paper, board, and card games. One could make the argument that Bertie the Brain or the British Nimrod computer are just quaint adaptations of existing manual games and not harbingers of a new medium. Readers of my first annotation will already know I reject that notion, but even I admit that without real-time action, the video game medium would not have flourished. While there have been plenty of successful turn-based games, the history evolves in very different directions if Space Invaders is a strategy board game or Tetris is just a static puzzle like the pentominoes game that inspired it. So while not a critical question in video game history, it is an interesting one.

Pentominoes: an okay brain teaser, but not much of a video game.

So what was the first real-time video game? Well, for decades the answer to that question would have been Tennis for Two, the 1958 game by Willy Higinbotham of the Brookhaven National Laboratory in which two players engage in a tennis match. Its funny that such an early milestone would be a tennis game when Odyssey Ping-Pong and Pong jointly launched the video game revolution just over a decade later, but it is truly a coincidence. Higinbotham only displayed his game at two rounds of Brookhaven visitor days in 1958-59, and no one associated with Sanders Associates, Magnavox, or Atari came anywhere near those exhibitions. Indeed, the game looks and plays completely differently, with a side rather than a top view and no paddles or rackets visible on the screen. Instead, the graphics consist merely of a horizontal line representing the court, a shorter vertical line representing the net, and the arc of the ball. Rather than a dial to move a paddle up and down, the players spin dials to select the angle of their return and press a button to cause the ball arc to change trajectory. All of this action does happen instantaneously in response to input, so it qualifies as a real-time game.

Tennis for Two earned its pride of place in video game history because even though none of the pioneers saw the demonstration, a budding young electronics enthusiast named David Ahl did. Ahl’s career is covered extensively in the book, so I won’t go into all that here, but one of his many pioneering feats was founding Creative Computing, the first magazine dedicated solely to the personal, hobbyist use of computers. Ahl was a high school student in Malverne, New York, in 1958, a Long Island community just 50 miles away from Brookhaven. One of the perks of a scholarship he received was a trip to one of those Brookhaven visitor days, where he played Tennis for Two. In 1982, as interest in the history of video games was starting to percolate for the first time, Ahl sent one of his writers, John Anderson, to Brookhaven to interview Higinbotham based on that memory, and in the October 1982 issue of Creative Computing, Anderson boldly proclaimed Higinbotham the “Grandfather of Video Games” (take that Bushnell and Baer!). Thanks to this publicity, Higinbotham’s game was featured in nearly every video game history book to follow, even if it was usually treated as a footnote (quite literally in the case of Steven Kent’s Ultimate History of Video Games).

A recreation of Tennis for Two, which for decades was thought to be the first real-time game.

Tennis for Two remained the gold standard in early real-time gaming for over thirty years, at least for the general public. In truth, an even earlier real-time game had already been rediscovered in the 1970s during the Magnavox video game patent lawsuits. These legal contests pitted the Magnavox Corporation and its new parent company, Philips, against any and all individuals and companies that attempted to create a coin-operated or home console video game without paying a licensing fee. The Magnavox claim of primacy in the video game space hinged on the work of Ralph Baer in the 1960s to create a video game system at Sanders Associates. Baer and his team filed several patents relating to their video game technology and then granted Magnavox the sole right to exploit the technology. Anyone else who wanted to play in the new video sandbox was required to pay Magnavox for the privilege.

Contrary to popular belief, the Magnavox patents did not really make a defensible claim to the sole right to exploit video games generally, but they did advance a claim that Baer’s fellow engineer Bill Rusch had invented a system in which a player-controlled dot and a machine-controlled dot rendered on a CRT collide and one of them changes vectors. Therefore, any video game that accomplished this same feat was infringing on said patent. The best defense against this claim was to show that other inventors had done this before Rusch, thus invalidating the patents due to the existence of “prior art.” Lawyers and legal interns working for companies like Atari, Midway, and Williams poured through old patent filings and technical journals to unearth earlier examples of real-time graphics with collision detection.

One game discovered through this process was a pool game developed in 1954 at the University of Michigan to demonstrate the capabilities of a computer called MIDSAC. MIDSAC pool featured graphics for balls and a cue stick rendered on a CRT, though the table had to be drawn on to the monitor with a grease pencil. Once a player took a shot with the cue stick, the balls would bounce off each other and the sides of the table in real-time while exhibiting realistic ball physics. One of the student creators of the game, William Brown, was called to testify in the first Magnavox patent trial in 1977 and described the creation of the game in some detail. Ultimately, it was not found to be prior art, presumably because it did not render its images through use of a video signal.

The documents from this case remained buried in legal archives until Ralph Baer unearthed a trove of materials from the case that had, if memory serves me, been sitting in a storage locker owned by one of the law firms involved with the case. Baer shared these files with a few places, and in 2011 scanned copies were posted on the website of the Franklin Pierce Center for Intellectual Property of the University of New Hampshire School of Law. They remained relatively unremarked upon until 2013, when they were publicized by Keith Smith (no relation).

At the time, Keith was deep into working on a new edition of his still-yet-unpublished opus All in Color for a Quarter, which tells the history of the coin-operated video game industry from its inception to about 1985. Astute readers of my book will notice references to this work throughout, and I can safely say my own book would be sorely lacking without his research. By tracking down rare trade publications and interviewing obscure pioneers, Keith has crafted the most comprehensive examination of this subject ever attempted and has done much to correct a legion of misconceptions regarding the birth of the video game industry. I only hope all of you get to read it some day as well.

In 2013, Keith wrote a blog post in which he commented on all the early video games identified by the parties in the lawsuit, many of which were unknown at that time. This is the first significant mention of MIDSAC Pool that I am aware of in historical scholarship. Two years later, Keith tracked down Brown’s trial testimony, which for the first time gave us an idea of how the game actually worked. Some time after that, when the Chicago Tribune launched a complete digitized archive online, I unearthed an article reporting on the actual demonstration in 1954. I also discovered some pictures of the game that had been sitting in an online photo archive maintained by the University of Michigan. Tennis for Two, while still an interesting real-time game, was no longer first.

MIDSAC Pool, which for a few short years was though to be the first real-time video game.

After Tennis for Two‘s thirty-one year reign, MIDSAC Pool only held its title for four years. As I was writing my own book, I was forced to think more deeply about how to define a video game and what early programming experiments I should identify as such. I was hardly the only person on the Gaming Alexandria Discord thinking about these subjects, and in September 2019, a few of us had a discussion on all the early video games we knew about. During the course of this discussion, my friend Dale, who goes by QuarterPast, asked about Bouncing Ball. The Bouncing Ball program was known to me, as it featured prominently in the account of the creation of Spacewar! written by Martin Graetz in 1981 for Creative Computing. In the article, Graetz described how when a person ran this program on the Whirlwind computer at Lincoln Labs “a dot appeared at the top of the screen, fell to the bottom and bounced (with a “thok” from the console speaker). It bounced off the sides and floor of the displayed box, gradually losing momentum until it hit the floor and rolled off the screen through a hole in the bottom line.” Graetz reckoned it was the first computer program that displayed a moving object on a CRT, but he dismissed it as merely a demo. So did I. Dale did not.

In our September 2019 conversation, Dale asked me straight out what I thought of Bouncing Ball, and when I once again dismissed it as a demo, he pointed me to a 2013 book by a gentleman named Jon Peddie called The History of Visual Magic in Computers: How Beautiful Images are Made in CAD, 3D, VR and AR. Nestled in this unassuming textbook was a remarkable claim I had never seen before: “Charles W. Adams, assistant professor of digital computers at MIT, and John T. (Jack) Gilmore Jr., one of the first systems programmers in the Mathematics Group at Whirlwind, were intimate with Whirlwind. They generated the first animated computer graphic by creating a program that would generate a bouncing ball on MIT’s Whirlwind’s CRT in 1949. Adams expanded the program so the operator had to adjust the display’s controls such that the bouncing ball would find a hole in the floor and drop in. This was the first interactive computer graphics game.”

This was stunning. Out of nowhere, a new candidate had emerged for not just the earliest real-time game, but also the earliest digital computer game, beating Bertie the Brain by one year! At the time, I remained unconvinced that this was a true game, as there was no evidence of authorial intent to create an entertainment program in this floor-adjusting mechanic. I was also incredibly suspicious of the year, as other sources intimated that the Bouncing Ball was created in 1951, and Whirlwind itself was still incomplete and barely operational in 1949. Still, this claim needed to be run down.

A search on the Internet turned up another person making a similar claim about Bouncing Ball as game, celebrated computer graphics pioneer Alvy Ray Smith. In a 2016 article for the IEEE Annals of the History of Computing, Smith wrote a piece called “The Dawn of Digital Light” on early computer graphics in which he examined Bouncing Ball and revealed that “anecdotally, Adams and his colleague Jack Gilmore modified the bouncing dot (“ball”) animation into a sort of game, perhaps in late 1950. The players (or player) would interactively alter the frequency of the bounces with the winner being the first to make the ‘ball’ go through a hole in the floor—a gap somewhere along the horizontal axis. The published code doesn’t show that this was a cycling program that awaited the next player’s move—in other words, that it was truly interactive. It appears instead to have been a program that was restarted each time with a different three initial conditions.” This statement appeared to provide the proof of authorial intent needed to satisfy my definition of a video game, but the use of the word “anecdotally” gave me pause. Was there proof that this game variant existed or wasn’t there? Smith did cite a source for this claim, a 2008 book called The Engineering Design Revolution: The People, Companies and Computer Systems That Changed Forever the Practice of Engineering by D.E. Weisburg, but it provided no further insight. All it contained was an uncited assertion that “Adams wrote a short program that displayed a bouncing ball on the display. This was done by solving three simultaneous differential equations. A little later, probably in late 1950, Adams and Gilmore wrote the first computer game. It consisted of trying to get the ball to go through a hole in the floor by changing the frequency of the calculations.”

Finally, after a little more searching I found the source of all these anecdotes: the panel proceedings of the 1989 SIGGRAPH Conference. In a panel entitled “Retrospectives: The Early Years in Computer Graphics at MIT, Lincoln Lab and Harvard,” Norman Taylor, who was actually at MIT at the time of the events in question, told of how Adams and Gilmore created the Bouncing Ball in 1949 and that “a little later Adams and Gilmore decided to make the first computer game, and this was also in ’49. This is a more interesting display. You see that the bouncing ball finds a hole in the floor and the trick was to set the frequency such that you hit the hole in the floor. This kept a lot of people interested for quite a while and it was clear that man-machine interaction was here to stay. Anyone could turn the frequency-knobs.” He even provided a slide of the bouncing ball in action! This seemed to lay the debate to rest: Bouncing Ball really was a game. One lingering question remained, however: was it really created in 1949?

A reproduction of a slide from Taylor’s 1989 SIGGRAPH presentation showing Bouncing Ball in action, complete with the hole in the floor.

At this point Dale and fellow GA researcher and author Ethan Johnson dug even deeper into this question. Turns out that MIT has a lot of old Whirlwind documentation up on its Dome online archive. Searching through this material, Ethan discovered a project report from February 1951 announcing that a student named Oliver Aberth has created a new bouncing ball program. This not only locks in a date, but also gives us a creator, and its not Adams or Gilmore!

So why is the program always associated with Adams? Well, even if Adams was not involved in its creation, he sure ran with it, making the program a standard part of the courses he taught on computer applications. These classes were probably the first exposure many MIT students had to the program, and they presumably naturally assumed the course instructor wrote it. Sadly, Aberth died mere months before this discovery, so we could not obtain more information from this newly discovered pioneer.

So now we know when the program was created, but did it feature at its inception the all important hole in the floor that turned it into a game? Probably not. The program is described in detail in the programming manual for the computer released in July 1951, complete with a drawing, and there is no indication of a hole in the floor. Dale examined the code of the program and theorized the hole was added later due to a bug that would eventually cause the ball to fall through the floor and no longer bounce. By providing a hole, the ball can gracefully exit the stage before the bug takes hold.

So when was the hole added? Sadly, we cannot say for sure. The only evidence that we have right now is a series of MIT course descriptions that feature the game. Adams was using the program in his classes by early 1952, but the hole is not explicitly mentioned in descriptions until February 1953. Based on the limited documentary evidence, Ethan speculates that the game variant was created sometime between late 1951 and the last third of 1952. Even if it did not come into existence until the first documentary proof in 1953, however, it beats MIDSAC Pool, so its undoubtedly the first game with real-time graphics.

A drawing of Bouncing Ball from the July 1951 programming manual. Notice the lack of a hole.

Ethan, Dale, and I had these conversations in early October 2019. My book was released in late November 2019. Thankfully, while my manuscript had been turned in months ago by this point, we were still in the final stages of proofing it, so I was able to sneak these new discoveries into chapter 3. Because of the last minute nature of this addition, there is a small error where the creators of the game variant are just identified as “Adams and Gilmore” rather than by their first and last names. This is because they had previously been referenced earlier as the creators of Bouncing Ball before we discovered Aberth, and when I took that reference out it did not occur to me that they were no longer properly introduced in the text. Still, I am pleased we were able to get this info into the book at such a late date. The goal of the book is to be comprehensive while offering new insights into the history of video games, and being the first published video game history book to reveal the earliest real-time game application fit both of those goals.

They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I 1971-1982 is available in print or electronically direct from the publisher, CRC Press, as well as through Amazon and other major online retailers.

Worldly Wednesdays: Human and Analog Computers

This post is part of an ongoing series annotating my book They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I. It expands on material found in Chapter 1 on pages 1-7. It is not necessary to have read the book to comprehend and appreciate the post.

One of the goals of They Create Worlds is to place advancements in video games in the proper context of the technological and economic situations of their times. To help meet this goal, the book occasionally delves into general computer history. This is especially true in the early chapters of the book, for I felt it important to demonstrate why Spacewar!, the game that started us down the path towards a new pastime and entertainment industry, could not have been born and spread until the early 1960s. Still, the book is a history of video games, not a history of computers, so I had to be careful not to get too sidetracked by this more general history, fascinating though it may be. Therefore, this annotation will give a little more context to early computer history.

What we call a “computer” today bears only a small resemblance to what the pioneers in this field would have considered a computing device. As with most modern appliances, the computer was created to automate a task already being performed through manual labor, in this case the work performed by people known as computers, a moniker that has existed since at least 1646. These days, if one is aware of human computers, they most likely associate them with the book and movie Hidden Figures, which traced the exploits of several female African American computers who contributed to NASA’s mission to put a man on the moon. While a great story, Hidden Figures does perhaps unduly glamorize this profession and places too much emphasis on individual ability when generally a human computer was just a small cog in a much larger machine.

Not every computer gets to work with Kevin Costner

The computer job developed in tandem with the scientific and industrial revolutions that spread throughout Western Europe between the sixteenth and nineteenth centuries. Its rise is most closely associated with the logarithm, a mathematical concept first espoused by Scottish mathematician John Napier in 1614. The introduction of logarithms essentially allowed complex multiplication and division of incredibly large numbers to be accomplished through the addition and subtraction of small numbers instead, thus greatly reducing the time needed to complete these calculations and allowing mathematical modeling of a whole range of phenomena that had previously been too complex for existing algebraic or trigometric functions. This led to advances in numerous fields, including not just pure mathematics, but also navigation, astronomy, and surveying.

Divining the log of a number quickly in a time before calculating machines required the use of mathematical look-up tables. Mathematical tables had existed for centuries before the introduction of the logarithm, with the first trigometric tables dating back to ancient Greece, but the use of logarithms allowed for more precise tables than previous efforts and turned the production of tables into a literal cottage industry. One example is the Nautical Almanac, a book of navigational tables commissioned by the British government in 1766 and updated annually ever since. The creator of the Almanac, Nevil Maskelyne, employed retired clerks and clergymen across the country to complete the tables in their own homes.

Because creating tables involved performing computations, the individuals doing these calculations were called computers. These were not mathematicians boldly coming up with new formulas or proofs, scientists discovering how the universe works, or even engineers using mathematical and scientific principles to solve real-world problems. They were more akin to bank clerks hunched over a ledger entering figures while preforming basic arithmetic. They had to be somewhat quick of mind, but required only a small amount of specialized training. Working in computing was really little different from, say, spinning cloth save that it required mental rather than manual dexterity. Therefore, just as cottage industry in the textile industry was displaced by factories during the Industrial Revolution, so too was it ultimately displaced in computing.

Victorian-era bank clerks, which are like computers, but handle money

The first noteworthy computer factory was created in France in 1791 by Gaspard de Prony, who was engaged by the French National Assembly to create an extensive set of mathematical charts to aid in a new national survey of all the land in the country. A mathematician who believed the division of labor principles of Adam Smith’s Wealth of Nations could be applied to mathematical table creation, de Prony gathered his computers in a single building and divided the labor into three parts. At the top were a handful of brilliant mathematicians who specified the formulas and basic parameters of the tables. Next, were a group of lesser mathematicians who performed the most important calculations, started each table, and instructed the third tier of employees, the computers, how to proceed. These computers then completed the tables using nothing more than rudimentary addition and subtraction.

While de Prony’s computers were drawn from all walks of life, many were either unskilled laborers or professionals in fields that did not involve math. For example, many were hairdressers left unemployed as the aristocratic heads upon which they practiced their trade were lopped off by Madam Guillotine at an ever increasing rate. As the work progressed, de Prony discovered there was no correlation between level of intelligence or education and accuracy and that computing had therefore truly been reduced to a basic unskilled task. This does not mean all computers henceforth were unskilled laborers. As the problems tackled by computers became more complex in the 20th Century, more and more education was required to do the job well. Indeed, Katherine Johnson, one of the principle protagonists of the aforementioned Hidden Figures, was a college-educated math prodigy who’s early career arc would have probably been very different if she had been a Caucasian man instead of an African American woman. Still, most computers were not Katherine Johnson, and the job was generally considered akin to low-level clerical or secretarial work at best.

The first machines that we would today call computers were conceived as part of an effort to automate table making in the same way the spinning mule automated thread production. Indeed, Charles Babbage, the mathematician and gentleman scholar who envisioned the first such machine, was well acquainted with de Prony’s factory. Babbage felt that England had fallen behind mainland Europe in mathematical sophistication and visited Paris multiple times beginning in 1819 to consult with esteemed members of the French Scientific Academy. He also oversaw an astronomical table project in his native country beginning in 1820 that operated in much the same manner as Maskelyne’s Nautical Almanac through the use of freelance computers in a cottage industry model. Frustrated with the process, he resolved to build a machine that would automate the table making and printing process based on the principles of de Prony’s factory. Once he completed a design on paper for this “Difference Engine,” he started work on a more general-purpose mechanical calculating machine called the “Analytical Engine” that could perform all four basic arithmetic functions and store numbers in a form of memory during the process. Though never built, this Analytical Engine idea would influence early computer designs in the mid twentieth century.

Charles Babbage was unable to build any of his computing devices both due to a lack of funding and due to a lack of mechanical parts with sufficient precision to perform the processes he required of them. This latter problem ceased to exist by the late 19th Century due to advances in manufacturing technology. This led to a new category of machines that we would call analog computers today, though at the time the term “computer” was usually only applied to a person rather than a machine. These analog computers evolved in parallel with human computers rather than replacing them and were used to simulate real-world phenomena when mathematical equations and models were not sufficient for the task at hand. As stated in my book, perhaps the most celebrated analog computer of the nineteenth century was Lord Kelvin’s Tide Predictor, which used a system of levers, pulleys, and gears to simulate tidal forces and allowed for the creation of far more accurate tide tables for seaports around the world than previous methods.

A modern recreation of the Difference Engine, which Charles Babbage was not able to build in his own time

By the beginning of the 20th Century, the computer profession was expanding as differential equations were increasingly used to generate mathematical models of phenomena that could not be easily observed with the naked eye such as electromagnetic wave transmission and atomic structure. As these equations became longer and more complex, armies of human computers ran numbers through these formulas to aid in a variety of scientific and engineering fields. Analog computing continued to develop in tandem, for when an equation proved too difficult or too complex to solve through computing, a physical model of the phenomenon being studied would be created instead. That way, even if the math behind a physical process was not fully understood, it could still be simulated and measured to solve a variety of practical problems.

As touched on briefly in the book, by the early 20th century, analog computing had become an indispensable tool in building power grids. Power transmission was a particular area where the differential equations were so complex that it was easier to simulate processes physically over a small area and scale up the results rather than solve the equations mathematically. This is how computing machines became inextricably linked with university electrical engineering departments in the United States before the establishment of computer science as a separate discipline rather than with math or physics departments. The vast geographic distances in the American West required solving a plethora of problems in transmitting electricity that were not being encountered anywhere else in the world at the time, and analog computers provided many of the solutions to these problems.

Electrical engineer Vannevar Bush was perhaps the first person to steer analog computing more directly into the realm of the human computer when he completed the first practical differential analyser, so-called because it uses mechanical parts to solve differential equations via integration. The theory behind the machine was developed in 1876 by Lord Kelvin’s younger brother, James Thompson, but he proved unable to build a working model. Many of his principles were incorporated into his brother’s Tide Predictor as well as several other analog computing devices, but these remained special-purpose machines tuned for specific tasks. Bush, an electrical engineer working on power transmission problems at MIT, built a device called a product integraph in 1924 that simplified the solving and plotting of first-order differential equations and then expanded it between 1928 and 1931 in conjunction with Harold Hazen, who suggested the device could be improved to solve second-order equations as well. The resulting differential analyser was a general-purpose device that could solve a wide variety of differential equations. Before long, engineers in other parts of the world had constructed their own differential analysers, setting in motion the eventual replacement of human computers with machines.

Vannevar Bush’s Differential Analyser

Most of the early digital computer projects, in which mathematical modelling completely displaced physical simulation, were started to solve differential equations. This included Howard Aikens’s Harvard Mark I, John Atanasoff and Clifford Berry’s unfinished prototype later dubbed the Atanasoff-Berry Computer, or ABC, and the ENIAC at the University of Pennsylvania. Indeed, ENIAC, short for Electronic Numerical Integrator and Computer, is the device which best brings together the disparate threads of this article. The co-creator of ENIAC, John Mauchly, worked in weather prediction and yearned for a device that would allow him to solve complex equations. After viewing an electric calculator displayed by IBM at the 1939 World’s Fair, he believed he could build a similar device for his purposes and immersed himself in the study of electronics. After giving a lecture on his ambitions at the American Association for the Advancement of Science in Decmeber 1940, he connected with John Atanasoff, who invited him back to his home base at the University of Iowa and showed him the work he had done on the ABC. Soon after, Mauchly took a position at the University of Pennsylvania and bonded with a graduate student named J. Presper Eckert, who was interested in developing high-speed calculating devices using vacuum tubes.

Meanwhile, Mauchly’s wife was running a training program for human computers that were being funneled to the U.S. Army’s Ballistic Research Laboratory, which was harnessing a differential analyser and an army of human computers to create artillery tables that would allow artillery commanders at the front to compute the proper trajectory for their guns to hit targets miles away without the need to solve complex trigometric equations on the fly. Just as Charles Babbage saw a need to automate the creation of astronomical tables in the 1820s, Mauchly envisioned speeding up the complicated and time-consuming process of producing artillery tables using an electronic computer that worked in a similar manner to the differential analyser only much faster. Though completed too late to be of much use in World War II, ENIAC, reduced the time needed to generate an artillery table to just 30 seconds as compared to 15 minutes for a differential analyser or 20 hours for a team of human computers. While ENIAC did not exert much influence on future computer design, it was the first completed general-purpose electronic computer that was publicly revealed, thereby playing an outsized role in stimulating further computer research at other institutions.

Human computers reached their apex in the 1940s as scientific advances played a key role in ending World War II. Because so many able-bodied men were serving as soldiers, the field came to be dominated by women. Furthermore, since educated women had few options for careers at the time, this also meant that a greater percentage of computers had advanced degrees in mathematics than in the time of Gaspard de Prony. Computers continued to play an important role in performing calculations into the 1960s, as Hidden Figures plainly attests, but as electronic computers continued to become both more sophisticated and cheaper, the occupation eventually faded away. Analog computers continued to play their part into the 1960s as well, and even into the 1980s in specific fields, but were likewise eventually obsoleted by their digital brethren. Electronic computers, meanwhile, continued to expand beyond mere calculating devices to encompass a wide variety of tasks. Much of this was due to the work of artificial intelligence pioneers like Claude Shannon and Alan Turing, which is where early computer history and early video game history intersect through the tic-tac-toe, nim, checkers, and chess programs of the 1940s and 1950s that are the subject of the first two chapters of the book.

They Create Worlds: The Story of the People and Companies That Shaped the Video Game Industry, Vol. I 1971-1982 is available in print or electronically direct from the publisher, CRC Press, as well as through Amazon and other major online retailers.