Sunday, February 15, 2015

Annotated Atari Depositions - Part 4

Today, we continue with Nolan Bushnell’s January 1976 deposition in the Magnavox v Bally et al case. This time, I don’t have as many comments/notes or pictures. It’s mostly just straightforward text.


MR. WILLIAMS: Q. Mr. Bushnell, as I understand it, Documents 40-2, 40-3, 40-6, 40-7, 40-10, 40-11, 40-15 and 40-17 all relate to circuitry which you intended to use with the monitor in association with your system for playing games

A. Correct.

Q. But that there is, as I understand it, circuitry which you had also intended to use for that monitor which is not shown in any of these documents?

A. That's correct.

Q. Did you personally draw the diagrams of 40-2, 40-3, 40-6, 40-7, 40-10, 40-11, 40-14?

A. Yes, I did.

MR. HERBERT: Before I start questioning on that, prior to the recess I had indicated that we would, after finding the pre-production drawings for Pong and another game, Space Race, zero in on the earlier drawings on this to match up with what else goes here. During recess Mr. Bushnell told me that very probably all of the drawings that are missing from this package were left at Nutting. So we don't really expect to find them.

MR. WILLIAMS: Q. Do you believe that the drawings missing from Exhibit 40 relating to your monitor are at Nutting?

A. Yes, I do.

Q. Why were they left at Nutting?

A. Well, they were all essentially source documents which were later used to build the Computer Space machine which I sold to Nutting and since I had licensed them to build that machine exclusively they are obviously entitled to all the documents that have to do with that particular machine.

Q. Do you know when you drew these documents which I just enumerated?

A. I'd say it was probably around July or August 1970. It might have been as early as February for some of them, but I think the ones that I drew in February were rougher. These are more detailed as to interconnections.

Q. Since only a portion of the circuit of the monitor is shown on these drawings, it might help us if you could draw a block diagram of the operation of that monitor if you are able to do so?
A. Okay.

Q. And I might say that we will probably mark it as an exhibit. You will be forewarned.
MR. WELSH: Now, this is of the monitor system?

A. Incidentally, this is generalized terminology for data bus architecture. I am indicating there are several lines. The exact number of lines depends on the resolution that you wish. This is approximately with some simplifications.

MR. WILLIAMS: I would like the Reporter to mark the block diagram that Mr. Bushnell has just draw as Atari Exhibit 41.
(Drawing made by the witness was marked Atari Exhibit No. 41 for Identification.)

MR. WILLIAMS: Q. Mr. Bushnell, will you just give us a description of the operation of this system shown in Atari Exhibit 41?
A. The oscillator runs a sync chain which essentially counts down the oscillator frequency into a horizontal and vertical component. There’s a number associated with each picture element in both the horizontal and a number coincident with each line in the vertical direction. These numbers are fed into a compare circuit which is compared to a number which is the shift register. If you can visualize a television screen, zero zero being in the upper left-hand corner and 256 by 256, the number 256 by 256 is the number of the lower right-hand corner. Then you can see that there is a series of ordered pairs which describe every point on the TV screen. Everyone follow that?

Q. I follow it.
A. The sync chain will count every one of those numbers in one frame. So that first the vertical counter is at zero and the horizontal counter then counts up to 256 at which time it gives the TV screen a sync pulse and the scan is reset and now the vertical scan counts to 1 and again to 256, and then the vertical goes to 2, 3 until it's scanned the whole time.

Now, supposing that we wish to display an object at point 20/20. That would be one inch to the right and one inch down from the upper left-hand corner. We would then put the number 20/20 into the shift register. Upon 20 being compared--

Q. Excuse me. You mean load 20 into the shift register under the block marked “Sync H”?

A. Yes, and 20 into the shift register under the block "sync V." This comparator, all it does is look for a comparison. It says when is one number equal to another . As soon as it does, it goes, ''Hi," turns on the scanning matrix. The scanning matrix then says, "Okay, I'm ready to scan," and it counts—I should say it is hocked into the oscillator or to the sync H and V.

We'll just for ease put the oscillator here (indicating).

It says, "Okay, I'll display a rocket ship at that point," and it counts through and displays a rocket ship at Point 20/20.

Now, what the comparator does is it feeds that number 20 into the shift register. Now, next frame it says, "Okay, I want the rocket ship to move downward and to the right.” So the next frame it will load into it number 21/21. The same thing happens.

This time, though, the rocket ship is moved slightly. It's no longer displayed at 20/20, it's at 21/21. And successively each frame. So in that way the rocket ship appears to be

traveling in a downward and to the right velocity because the eye integrates the motion. It's just like a series of cartoons and you display it at slightly different places each time and the picture appears to move.

Now, the computer, of course, is keeping track of one, if the control is being pushed, say, in the forward direction and it’s thrusting, it’s saying, “Okay, if I want the rocket ship to go faster I’m going to say instead of moving it from 20 to 21 I might move it from 20 to 22. That gives the appearance of a faster motion. Or if it wants to move slower maybe it says, “I’ll keep it at 20 for a couple of frames and then I’ll move it to 21 for another couple of framed and then to 22 for another couple of frames.” So it’s going half as fast.

If you wanted it to go straight up and down all you are doing is you leave the horizontal counter fixed at 20 and you just increment or decrement the vertical count and the computer keeps track of all these numbers and feeds a new number out each frame which places the rocket ship anywhere.

Now, obviously, if you wanted to, you could make the object jump anywhere you want to once each frame. But generally by making a piecewise continuous function you can have the appearance of smooth motion, but it’s not constrained to that.

Q. The computer which you refer to is not shown in Exhibit 41, as I understand i.

A. No. It’s a data bus here. It’s out here (indicating). It comes in on the bus. I should have put that down, “data bus.”

Q. There are two boxes in the lower left-hand corner. What does the label on the upper one of those boxes say?

A. “Interface.” It’s essentially the part of the circuit that’s described in 40-7.

Q. And the lower box in the lower left-hand corner is labeled “IO.” Is that correct?

A. Yes. That’s essentially the problem that we were talking about before. Whether you do that on an interrupt basis, or whether you do it just putting data into memory. Now, you can do it either way. That IO is either a direct memory access channel or it’s an interrupt channel.

Q. So the player’s controls are located within the box marked “control”?

A. No. Those are control switches and coin slot.

Q. Over on the lower right-hand corner?

A. Right.

Q. And that information goes through the box marked “control” to the box marked “IO”?

A. Yes. Input-output.

Q. So the box marked “IO” was actually only an output channel, is that right, the way you have drawn it?

A. in a computer there is never really an output without an input because there’s parity checks. It talks. You send signals in two directions and for a multitude of reasons, but it essentially can be looked at, it says, “Hey, I’ve got some information.”

The computer says, “Okay, I’m ready for the information. Go ahead and send it. “ So it sends it down. The computer says, “Okay, I’ve got it.”
And the guy says, “Is it right?”

And the computer says, “Yes, it was.” And they do these handshaking things all the time. It’s just the way the architecture is.
Q. But the basic purpose of the IO block is to get information from the monitor to the computer, as I understand it.

A. Yes.
Q. You described the way in which one rocket ship is show during one scan of the cathode-ray tube.

A. Right.
Q. In the system you were building in playing Spacewar I assume that you wanted to show at least two rocket ships.

A. Yes. That adds a whole new level of complexity. That’s kind of that this is. Because these are shift registers. I have two sets of shift registers, one labeled A and one labeled B. I am referring to Document 40-2. It’s necessary that you have two data boards, one the location of the rocket ship A and one the location of rocket ship B, and under the control you can switch from one to the other.
Q. What does the circuitry shown in Exhibit 41 provide? What is necessary to display two rocket ships?

A. Well, it depends on how much intelligence you ascribe to the control module. Like it’s possible that the control module or the computer is smart enough to serially order the information in shift registers, in these two shift registers, so that it always hits the first shift register of the first object in the scanning sequence, and then the minute it sees that then it dumps that information out, grabs another piece of data from the computer and says, “Okay, this is rocket ship No. 2.” That could do it. But I think that 40-2 is a better approach, It’s a little bit cleaner.
Well, this is again the problem that I ran into. If you do it this way the computer has to be very smart and it has to be fast because it has to have that information ready for the second rocket ship very, very quickly because the minute the one rocket ship is done, if the other rocket ship is very close to it it has to have that information in a big hurry or you’re going to lose it, the rocket ship will disappear if it gets close. So what you can do is you can say, “Okay, I’m going to make the monitor smarter and I’ll just dump the information out at one time,” say during frame scan or frame reset in which there’s a lot of time, and that way the computer doesn’t have to be as smart. This design was an afterthought of this kind of architecture.

Q. You are saying that the document of 40-2 is an afterthought to the architecture shown in 41?
A. Yes. I want to keep this simple so that you can understand it. You see, the more and more smarter I made the monitor, the less power I had to have in the computer itself until finally I said, “The hell with it, “ you know, “let’s just build the hardware unit.”

Q. And the computer system which you intended to use with the apparatus you have shown in the block diagram form in Exhibit 41 was the Nova 1200 series computer; is that correct?
A. That is ultimately. I think this was a general-purpose design. I’m not sure which one this was, how late it was. But I originally designed the general purposely so that it could adapt to essentially any 16-bit machine.

Q. Did you have any requirements on the memory capacity of a 16-bit machine with which this could be used?
A. Yes. As small as possible.

Q. What memory capacity was required for a game system using, for example, four monitors?
A. I felt that in my original thinking I could get by with four K.

Q. For four monitors?
A. Yes. And memory was never the problem in the design. It was always update speeds.

Q. Can you identify Exhibit 40-15?
A. Yes. It says, “System Input, one coin box to initialize particular CRT and program.” I think at that time we were talking about having it to be a situation where you could not only choose whether you wanted to place Spacewar, but whether you wanted to play any other game that we had in the program. That was kind of a question of memory. We thought that it would be interesting to have the switch selectable so that you could play a multitude of games. So that was No. 1 as far as system input.

“No. 2, counterclockwise rotater input on fixed-time increment and rotate counterclockwise one unit. Unit equals question mark degrees.” I don’t know.
No. 3 says, “Clockwise rotator SPP2.”

No. 4 is. “Accelerator input on fixed time increment and add velocity increment to VX and VY.”
No. 5 is, “Fire control causes missile to shoot at fixed speed relative to rocket in direction rocket is pointing. Output to CRT done by sorting position, line and data. Words into output area. An interrupt will be generated at end of each field to indicate.”

That’s it.

Q. Did you write the document of 40-15?
A. No, I didn’t.

Q. Do you know who did?
A. No, I don’t. I have been asking myself that. It could have been a guy named Larry Bryan who was going to do the software at that time.

Q. Do you know what the list of five items under the heading “system input” is?
A. Well, yes, I think it’s essentially all the things that we wanted to put into the system, you know, to make sure that we had enough input ports to play the game.

Q. Did Mr. Bryan generate this document 40-15 as a result of a special review?
A. If he, in fact, was the one that generated it, and I think he was, yes.

Q. Do you know when it was?
A. It was probably during the summer.

Q. Of 1970?
A. Yes.

Q. Was anybody other than Larry Bryan assisting you in the construction of your apparatus?
A. Ted Dabney.

Q. What part did Ted Dabney play in the construction, in the development, of that apparatus?
A. He was a good circuits guy. He ultimately designed most of the sound circuitry and the video amplifier.

Larry Bryan was the software man.
I was the hardware man and Ted was the analog man.

[NOTE – This is a major point of contention between Ted and Nolan. While all acknowledge that Ted did the sound and video circuitry, Ted maintains that he (Ted) did the motion circuitry described above while Nolan claims that it was his (Nolan’s) work. The motion circuitry is what Nolan’s patent covered – though as we shall see if I ever get to it, he claims that his patent was not based on work done for Computer Space.]
Q. Can you identify Exhibit 40-16?

A. A timing diagram to a Nova 1200 computer.
Q. Can you identify Document 40-8?

A. I think it was a pin designation of input and output for the interface unit to the computer. Yes, it’s a pin assignment.
Q. Did you write the document?

A. Yes, I did. Any of the documents that you can’t read are probably done in my handwriting.

Q. Can you identify Document 40-18?

A. It’s a timing diagram.

Q. Is that also of the Nova 1200?
A. I think so. I don’t really know what the difference between the two documents is. One might have been for the Nova 850 which was a faster machine. When we started getting into problems I thought it might have to go to a faster machine.

Q. Would you please identify Document 40-9?
A. It’s a Xerox of a Signetics integrated circuit, and why I’ve got it there I have no idea.

Oh, I know. That’s an interface chip. It’s the interface unit. I wanted to make sure that what I was feeding to the computer wasn’t going to blow it up. Yes, that’s what it was.
Q. Can you identify Document 40-5?

A. It looks like it’s part of the technical manual for the input-output bus structure for the Data General computer.
Q. Was that for the Nova 1200?

A. I believe it was.

Q. Can you identify Document 40-4?

A. Yes. That’s the pin-outs of the bus connections for the Nova 1200. These are all Xeroxes because before I had a chance to start talking to the Nova people I was scrounging around for a manual on it and there was only one in the plant that I could get ahold of.
Q. One in which plant?

A. Ampex.
Q. What do you mean by the term “pin-outs”?

A. Well, the bus is essentially the input ports and it tells what part of the computer is connected to what pin. You know, like is It an input port, an address port of interrupt port.

As an example, all I could use were the ones that were vacant, you know, upon the direct memory access. These are General Data bus structures.
See, you put input and output in terms of a word. You see these data A8, All, each one of these represents a bit in the data word.

Q. Can you identify Document 40-13?
A. Yes. It’s a page describing how the data channel transfers work with this particular computer. It’s necessary in designing the thing to really have that stuff well scoped out.

Q. Would this particular computer be the Nova 1200?
A. Yes.

Q. Can you identify Document 40-12?
A. This is an OEM discount schedule which tells you what your price would be depending on how many units you buy. I was trying to find some way to get them to believe me that I was going to take 200 units the first year so I could get a 40 percent discount, but I didn’t quite have that much guts. But I was projecting that if the item did very well there would be considerable savings in the computer.

Q. I believe you testimony is that sometime between January 26, 1971 and February 16, 1971 you decided not to use the central computer system?
A. Correct.

Q. What did you do after you decided not to use the central computer system?
A. I put my time into designing a very inexpensive and complex exerciser, if you would, that would essentially do the calculations and hardware. At that time I had had a very complex exerciser already going, but it took me quite a bit to get it up to the point where it controlled two objects. It was only good for one object at that point.

Q. “That point,” being the time you decided not to use the central computer system?
A. Correct.

Q. Did you eventually build that exerciser?
A. Well, I guess you can say that—What’s to say when something is complete? It was complete when it went into production at Nutting Associates. I mean, that was the first commercial result of that. But I could move objects around the screen before that time.

Q. What time did it go into commercial production?
A. We sold our first units in December of January of the following year. I guess that would be December ’71 of January ’72.

[NOTE – Numerous sources give a release date of November 1971 for Computer Space, but almost none of them give a source for the date. One that does is Michael Current’s excellent Atari timeline website, which lists the source as “Cash Box 11/27/71 ad p54; 12/4/71 p45”. I do not have either of these issues, but it looks like the November reference was an ad for the game. Ads/flyers typically first appeared in trade magazines about a month before a game was released. It is unclear if the December reference was an ad or an actual release announcement for the game. ]
Q. Did it go into production approximately the same time it was first sold?

A. Oh, yes.
Q. How long before it was first commercially sold would it have gone into production?

A. Well, I think we were trying to get some units out as soon as possible. We showed it at the show I think it was October-November, and as soon as we—we were hoping to have production units ready by then, but the just weren’t and the production units weren’t really ready to ship until that December.
Q. At what show did you show it?

A. Music Operators of America.
Q. Where was the show held?

A. It was a hotel in Chicago. The Palmer House, I believe.
Q. When you say you showed it at the show, was there an operative game there at the show?

A. yes, there was. But it was a lash-up. I carried the computers in my suitcase to the show and we had shipped the cabinets ahead and brought the computers in and installed them and babied them through.
Q. When did you commence your employment with Nutting Associates?

A. I think it was March or April of ’71.
[NOTE – According to Goldberg and Vendel, Nolan started working at Nutting in March of 1970, not 1971. To me, the 1971 date seems more plausible given the information in this deposition (i.e the fact that they were still writing letters to Data General in January and February of 1971) and assuming its accurate. It also seems a bit implausible to me that the worked on the game for Nutting for almost two years before releasing it, but it was sort of a side project and Nutting was inexperienced, so maybe. The claims in the deposition seem to fit better with Nolan visiting his dentist in February 1971 rather than February 1970 as reported in Business Is Fun. OTOH, Marty and Curt generally have solid documentation to support their dates, and it may be that they have such information in this case and maybe Nolan is a year off in his dates.]

Q. I think you testified that you took the computers in a suitcase. What computers are you referring to that you had in your suitcase?
A. That ones that were built for Computer Space.

Q. When was the first time that you had a completed apparatus on which you could play the Computer Space game?
A. You say the Computer Space game. There was a lot of variations and modifications to it.

Q. When was the first time at which you had an apparatus completed on which you could play any version of Computer Space?
A. Oh, it was probably April or May of ’71.

Q. In connection with Exhibit 40-15, I think you said that you wanted to make the system so you could play Spacewar or other games. What other games did you have in mind at that time?
A. Well, I had in mind, you know, various sports games, various arcade games that I had seen in school, you know, when I was at the amusement park. I was thinking particularly of baseball. I was also thinking of hockey.

Q. Do you have any documents which would show the games that you contemplated using with your system at that time?
A. Yes, I do. Now, these are some of the files, some of which are missing, and I don’t know why I’d have these and why I don’t have the others or why I have any at all. I think most of the others are at Nutting. I also have my book in which I have just essentially some of my cost estimates on the Nova and the PDP-8. This is the company that I rented some time on a 16K Nova.

Q. Before you go any further, these files, I gather you were pointing to four files, the first one marked “File No. 9
Q. Was that the same agreement as the agreement relating to Computer Space?

A. No. It was an employment agreement that Nutting had.
Q. Was there a separate agreement from the employment agreement which dealt with your retaining rights to video technology?

A. No. That kept my rights to video technology.
Q. That is, the employment agreement?

A. Right. The agreement that specified the rights that I was conveying to Nutting was in a separate agreement which spelled out the payment terms and things for Computer Space. I was listing each game individually.
Q. Were there any agreements on any other game than Computer Space?

A. Yes. We had an agreement on a game called Two Player Computer Space.
Q. Were there any other agreements relating to games with Nutting?

A. No, there weren’t.
Q. Can you describe for us the game Two Player Computer Space?

A. It was essentially two rocket ships fighting one another in a star field. It’s much close to Spacewar than Computer Space was because it didn’t have the computer-operated flying saucer. Or it did have it. It was one or two-player. You could play against the computer or you could play against the other rocket ship. Computer Space was just a single-player game and could only be played by one person.
Q. Did Nutting ever commercially manufacture the Two Player Computer Space game?

A. Yes, they did.
Q. Do you know when the commenced this manufacture?

A. I think it was shortly after I left. Not shortly after I left, I think it was the following fall.
Q. For how long did they manufacture that game?

A. I have no idea. It was my impression that the game was a mistake. I didn’t think it was a good idea. It was one of the items preceding the disagreement on which I left. I think history bears me out that I was right on it.
Q. Did they manufacture it for a period or months of a period of years or—

A. I have no idea.
Q. Do you know how much they sold that game for?

A. I think it was $1500 or something like that. Very expensive.
Q. Nutting, I assume, did commercially manufacture the Computer Space game?

A. Yes, they did.
Q. Do you know how much they sold that for?

A. Yes. They started out at $1,295. Or was it $1,195? Something like that. It was either $1200 or $1295. I think they later dropped the price to $950.
[NOTE – This is one of the few semi-contemporary references I’ve seen to Computer Space’s sales price. I’ve seen some accounts that indicate a price of around $3,000 or more, which seemed way too high to me. Benj Edwards (normally very accurate and one of the best writers on early video game history out there) at one point seems to use a price in this range when trying to estimate Nolan and Ted’s Computer Space royalties.

Nolan’s figure of around $1200-1300 seems more plausible to me Bear in mind, that he was likely referring to the distributor price, not the operator price. Distributors generally marked a game up by 30% or so when selling to operators. At some other point in the depositions, it was mentioned that Nutting had a brochure at the MOA promising a price of under $2000.]
Q. I think you testified earlier that they started their commercial production in either December of ’71 or January of ’72?

A. Correct.
Q. Do you know how long that game was in commercial production at nutting?

A. I think they produced that through the following fall. I think they produced Computer Space up until they got Two Player Computer Space into production.
Q. Do you know you many units of Computer Space they sold?

A. I think it was about 13 to 15 (hundred) units. Since I got a royalty on it I probably have the figure around somewhere for sure.
[NOTE – the actual transcript says “13 to 15 units.” There is a handwritten word about the “13 to 15.” It is illegible, but looks like it says “hundred.”

The number of units Computer Space sold has been variously reported as 500-2200. A number of sources, including Goldberg and Vendel, report a figure of 1500, which I consider to be the most reliable figure. Some sources report a production run of 1500, but indicate that Nutting actually sold less. The 1973 student documentary Games Computers Play also cites a 1500 figure (probably the first instance of that figure, or any figure, being cited), but I don’t remember if it was the number produced or the number sold.]
Q. Do you know how many units of Two Player Computer Space they sold?

A. I have no idea.
(22 lines missing?)

Q. After you left Nutting, what was the first video game that you think you worked on?
A. A game called Asteroid.

Q. Was it known as Asteroid at the time you started working on it?
A. That’s what we called it around the company.

Q. Was that similar to the game that was finally sold under the designation Space Race?
[NOTE – This is quite interesting to me. As most of you probably know, Midway’s Asteroid was basically the same game as Space Race.

As revealed in depositions from Bally/Midway executives (John Britz and Hank Ross), which I hope to post in the future, Asteroid/Space Race is actually the game that Nolan/Syzygy delivered to Bally in fulfillment of its video game contract. Contrary to what some have claimed, when Bally officially turned down Pong (though Midway later licensed it), it did not void Nolan’s contract. Instead, he ended up fulfilling it with Asteroid/Space Race. Midway used the original name when releasing it. When Atari later came out with Space Race, Midway indicated that this might have constituted a breach of contract on the part of Nolan and they ended up dropping the 3% royalty specified in the contract as a compromise.]

A. That’s correct. You will find in our papers that we often have an in-house code name that doesn’t always come to market under that name.
Q. Is that name also known as VP-2?

[NOTE that Pong was VP-1]
A. Yes, it is.

Q. What was the next game you started working on after Asteroid?
A. It would have to be the game which is now called Pong. Maybe for classification here there were three of us that were technical.

Q. “Three of us” in what that were technical?
A. Well, three of the employees of the then Syzygy Company were technical and we each had our projects. Mr. Dabney had the pinball projects which was part of the contract engineering for Bally Corporation.

I had the Two Player Computer Space design for Nutting as well as the Asteroid design. The Asteroid design, incidentally, had been actually started before Computer Space because of the star field and all the other stuff. We thought that the first game should be Computer Space, but it was an easier game to do and we probably should have done that as our first entrance but we didn’t. So it was just really picking up on that design and rejuvenating it.
Mr. Alcorn, when he came aboard, his first project was to build a simulated tennis game. I only did about two days’ work on Space Race because I got bogged down in administrative details and running the company other than design and was able to finish up the Two Player Computer Space for Nutting, but Mr. Alcorn ultimately finished the Space Race design.

[NOTE – Bushnell here claims that Space Race was not only started before Computer Space, but also that it was the first game he worked on after leaving Nutting, though he only did so for a few days. Maybe it’s just me, but I find that interesting – especially since I’ve never seen it mentioned anywhere else before.]
Q. When did Mr. Alcorn come on board?

A. I don’t know. I can check the records. It’s in the spring. It was shortly after leaving Nutting.
MR. ETLINGER: What year would that be, ’72?

MR. WILLIAMS: Q. Shortly after you both left Nutting?

A. Yes.
(half a page missing?)

When you are a little company you think that model numbers are kind of window dressing.

Q. So the numbers were assigned sometime after the work on the machines actually began?

A. Yes.

Q. Did you give Mr. Alcorn the assignment of designing a simulated tennis game?

A. Yes, I did.

Q. How did you give him the assignment, was it orally or in writing?

A. It was oral.

Q. Do you know when you gave him that assignment?

A. The day he came to work.

Q. Can you state what the assignment was?

A. Well, I told him to make a tennis game. I wanted the ball to go back and forth horizontally. I wanted two men, two little men with rackets to move around the play field controlled by a joy stick with a button on top and when the button was pushed the little racket in the man’s hand goes like that (indicating).

[NOTE that Bushnell says he initially wanted little men on the screen, not just paddles, and that he originally wanted to use joysticks.]

Q. Indicating a striking motion?

A. Right. And that after a point is scored the ball would appear on the screen and you would have to move your man behind it to serve and bat the ball to the other side; that each time a point was scored you would hear a sound of a crowd of thousands cheering, which is an electronic circuit that you can make that does sound like "Hurray," you know, applause, and I wanted a distant "pop" when the ball hit and I wanted the ball to make a different sounding "pop" when it hit the floor or the sides.

Q. Was Mr. Alcorn successful in developing a game as you have just described?

A. It's hard to say. We worked very closely at the time and the game came together. Designing a game is kind of like drawing a picture and you initially make the big outlines and then the game is refined and refined and refined sort of like coloring in the sections.

I would say the first thing that's done is the sync generator is built and the ball-motion circuitry is put together. After that the paddle control is put in. Well, in an XY joy stick it's just a linked potentiometer so in a lab environment you generally don't go right to a joy stick. You go to two pots. Before you go to two pots you go to one pot.

We looked at the first thing that we had up on a screen which was essentially a rectangular blob which would later be cut by a diode matrix into the little man and the ball. But you could also--you know, it's very easy to make it so that when the ball and the paddle intersect instead of waiting for the computer to detect the hitting motion, that it just automatically bounce off.

That's the way we did the initial one. It didn't play badly, you know. We played it a little bit and found that the game was kind of fun. The problem we had was that the ball speed was very high at the time and we had trouble returning the serve. So we said, "Hey, let's play this a little bit more. Let's slow-the ball down."

Mr. Alcorn slowed the ball down and we played it some more and now we could get the serve back, hut the game was kind of dumb. I mean, it wasn't that much fun, you know.

Oh. I'm leaving out one thing. In this kind of a hitting motion we wanted the racket to do—

Q. The striking motion?

A. The striking motion. If you struck the ball when your paddle was in this direction--

Q. That is angled upward?

A. Angled upward, we wanted the ball to go up. If you hit it with the paddle perpendicular we wanted the ball to go straight over and if you hit it · while it was in this thing obviously the ball would go down (indicating).

Q. That is with the paddle angled downward?

A. Right. So we had various angles that the ball could have would be selectable. So we just selected which angle it bounced based on where the paddle was. That was in the game, but later it was going to be refined to detect coincidences of when the paddle moved, you know, where it was so that it was not just a, you know, get-in-front-of-the-ball kind of game, but a ball hitting the paddle, you know, where it was. So that was in the game. It played pretty fun, you know, it was pretty good. But, again, the ball now was too slow, and we said, "Well, if it's too slow, you know, if it needs to be slow to return the serve, but it needs to be faster after you get good to be fun." I said, "Well, why don't we just count the volleys and speed the ball up as a function of volley increase." So that's how that came into being. That was not part of the original design specification that I gave to Mr. Alcorn. So we put that in and it was fun. It was a good game.

Then we got into a big hassle. Mr. Alcorn didn't want to put the crowd of thousands in. He thought that it was a waste of time. He says, "Why not just a nice raspberry sound, sort of like (demonstrating) you know," and he said he could do that a lot cheaper.

I said. “Okay, put in the raspberry sound when it misses.” It was my idea that I wanted to cheer on the winner rather than badmouth the loser. But he prevailed on me. So the honk sound was put into the game on a miss. Digital scoring was put in. The game played pretty well. So we said, thinking in the back of our mind, "Hey, we’ve got this. We did it in a hurry. Let's give this to Bally satisfying their contract, their contract engineering. Then we can get off and get doing some of our own stuff.”

[NOTE – Pong was strictly a two-player game, so when one player scored a point, the other player missed. I think the fact that Nolan wanted to accentuate the positive while Ted wanted to highlight the negative says something about their personalities, but maybe that’s just psycho-babble.]

So this was a full six months ahead of schedule from when we were supposed to do it. So I thought, "Gee, this is great. The money is still rolling in and we will have satisfied our contract and happiness and bliss will reign in California." So I hopped on an airplane with the prototype, took it to Bally, showed it to Mr. Britts [sic – it’s actually “Britz”] and Mr. Lally who is, I guess, the vice-president of engineering at Bally.

Neither one of them liked it. The contract was so written that they could refuse--you know, that I had to provide to them an acceptable game, something that they accepted. So they said, "Aw, you have to have two people to play it. Who's going to pay a quarter to play ping pong on a TV screen," so on and so forth, "Go back to the drawing boards, Nolan."

So I did. I climbed back on the airplane very dejected because I thought it was a great chance to get off. I said, "Well, hell, we've got this game, it's designed. Let's put it in a cabinet and see how much it earns."

[NOTE that this is another bone of contention. Ted claims that it was his (Ted’s) idea to become a manufacturer, or at least that he was the main one pushing for it, while Nolan and Al were reluctant. Nolan and Al both say that it was Nolan that pushed for it while Al and Ted were reluctant.]

We did that. It earned very well. We all jointly made the decision that we were going to hock everything we had and go into production. So we figured out exactly how many units we could buy the parts for and hopefully have them sold by the time we had to pay for the parts. We had developed a little bit of credit in the valley at that time and so we made our first order for 75 units which at that time represented about five times as much money as we had or had hoped to even get. We made sure that the parts came in all on the same day so that we could essentially get them all built in a very big hurry and out and sold.

[NOTE that other sources, including Nolan himself in future interviews, claim that after the prototype, Syzygy initially produced around a dozen units. After they were sold (or at least 10 of them), they scraped together everything they had and produced 50 more.

Also, I got the impression reading Business is Fun that the units all coming in at once was by happenstance, not design.]
We did it and we were successful in being able to sell the machines, and with that money we made a re-lease for I think 300 at that time which was out of sight because we were in, you know, 1500 square feet of building. We ended up doing an awful

lot of assembly out in the parking lot. But that's essentially what happened.

MR. WILLLIAMS: Let's take a short recess.

(Short recess)



That’s it for this time. Next time, we’ll hear about Nolan’s famous visit to the Odyssey demo in 1972.

In the meantime, here are a couple of goodies.

First up are these two photos of Nolan from the 1963 Utah State yearbook (“The Buzzer”).


Next is a letter that Nolan sent to Bally’s John Britz on July 10, 1972.

Two things that jump out:

1) Note Nolan’s claim that he will be delivering a hockey game in fulfillment of the video game portion of the contract. From the description, this is not just his term for what became Pong, but is a far more sophisticated game that involved actual goaltenders, the incorporation of ice effects etc.
The interesting part is that this goes seems to contradict the standard story that Nolan originally intended to deliver a driving game and only assigned the tennis game to Alcorn as a training exercise and warm up for the driving game. OTOH, the standard story is very well attested by all three principals (Nolan, Ted, and Al). Nolan himself doesn’t recall writing this letter so we’ll probably never know exactly what was going on with the hockey game.


2) Note that Nolan uses the term “video game” twice in this letter. This is the very first instance I’ve found of the use of that term, which is normally dated to 1973. Of course, etymology’s usually rely on public uses of a term, not uses in private letters, but it still mildly interesting.


  1. As good as Marty and Curt's research usually is, I think their history of early Syzygy developments is off by a year. In addition to the discrepancy you point out here, I noted in my latest blog post (shameless plug) that they give an October 1969 date for the first meeting between Nolan, Ted, and Larry Bryan to discuss forming a partnership. Once again according to the excellent timeline by Michael Current, the Syzygy partnership was formally established in January 1971 (this, says Current, comes straight out of Atari's 72/73 financial statement rather than from a personal recollection, so it sounds pretty definitive), so an initial meeting in October 1970 makes a lot more sense. This also jives better with Nolan's recollections that he did not discuss a video game with Dabney until 18 months after starting work at Ampex (Kent and Ramsay) and that the original minicomputer concept was abandoned around Thanksgiving 1970 (Donovan, though of course the depositions indicate that he at least carried on a dialogue with Data General past that date). Current reaches the same conclusion and places the original meeting in October 1970, taking the month from Goldberg and Vendel, but explicitly rejecting the year.

  2. Alex, the date comes from Ted and the problem is that even the documents you mention are based off of Nolan's memories and ascertations at the time vs. formal documentation kept in what was an informal partnership. For example, the other Atari financial document from that period I sent you and Keith claims a December 1970 for the formal creation of the partnership, which also contradicts Michael's timline. Ted stated it was about a year after Nolan started at Ampex when the idea of a coin op video game came up (before that it was animatronic bear pizza place idea they were pursuing). That's what put it in Fall of '69. There were several meetings between the three and time for research into possible mainframes before that was abandoned and Larry was dropped. Then there was more dead time before Ted's spot motion circuitry development and then search for other partners at Ampex before Ted claims it was forgotten again until Nolan turned up with the Nutting offer. This time frame is again framed to Bristow and Alcorn's times as interns as Bristow did do some work on Computer Space during an intern session at Ampex. I think where your confusion is, is in regards to the loose use of the word "Syzygy" over the '69 to '72 period across several initiatives including the original "Syzygy," Syzygy Engineering and the formal declaration of Syzygy Co.

    1. I really don't see it. Even if we take Dabney at his word that they discussed a video game a year after Bushnell started working at Ampex, that still places it in early 1970, not late 1969. Regardless, I really see no reason to take Dabney's word over Nolan's on this point. Nolan has no reason to distort the truth here -- if anything it was to his advantage to claim developments happened earlier rather than later to support the idea that he created the first video games -- and even if the early financial statements were based on Nolan's recollection of events, these are recollections coming two years (1973 financial report with January 1971 date for Syzygy) and five years (Atari prospectus with December 1970 date for Syzygy) after the events in question, not forty. I would not expect Bushnell or Dabney to have a precise memory of the timeline after four decades, so the financial statements of the time coupled with Nolan's recollections in a deposition just five years later carry more weight with me. Also, Bristow in Zap (I know what we all think of that book, but this is a direct quote), specifically states he met Bushnell during his freshman year (1969-1970), but that it was not until he returned to Ampex in 1971 that Nolan was working on a game prototype and that Bushnell joined Nutting that spring, so Bristow's timeline also supports a 1971 date. While still a recollection after the fact, twelve years is a lot fewer than forty. In this case all of the sources, be they documents or interviews, line up around a 1970-71 timeline rather than a 1969-70 timeline except for Dabney. The guy may be well intentioned, but he appears to be an outlier trying to remember the order of events that happened forty years ago.

    2. Again, the event portion of the timeline is:

      - Nolan and Ted start looking at locations for a possible pizza/animatronic restaurant.
      - Nolan plays Computer Space at SAIL.
      - Nolan takes Ted up to SAIL and proposes them doing a coin version.
      - Nolan and Ted start looking for a programmer for the project. Larry is chosen.
      - The trio have several meetings over a few months.
      - The computer idea is abandoned because of cost.
      - Time passes with nothing happening and Ted thinking the project as abandoned.
      - Nolan approaches asking about the vertical and horizontal adjustments. Ted is asked to create a breadboard circuit to move a spot.
      - Ted begins planning and then building said circuit in his spare time over the evenings after work. This includes moving daughter Terri out of her room into his and his wife's, turning it into a study/office and building an addition onto the house (which Nolan helped with) for his and his wife's new master bedroom /
      - Ted and Nolan look for partners at Ampex for building it into a full machine.
      - Ted forgets about the project again as time moves on and Nolan is shopping it around outside of Ampex.
      - Nolan gets introduced to Ralstin via the dentist.

      Now there's no way that all occurred between December 1970 and February 1971.

  3. Good to see some nice intellectual discussion going on. Just picked up Atari Inc., by the way. Really good!

    The mystery of the term "video game" is probably the one which eats at me the most. We have so many people who regularly delve into the archives, yet we can't find a definitive timeframe for when it came into use? Guess it was just so obvious that it became ubiquitous. I find it very difficult to help put people into the mind set of a time by using unfamiliar lingo without going off on a digression. "Just call them sprites!"

    Not sure if I missed this in a prior post, but do you have an opinion on the Computer Space royalties deal? The only one that I've ever seen is 150,000, half of which was used to create the original Syzygy/Atari stock.

  4. Hey guys,

    Just uploaded an updated version of my timeline page for Syzygy/Atari, incorporating and responding to things posted here.

    Regarding the Computer Space release date, I added images (linked in the timeline) of the actual ad from 11/27/71 (proclaiming "Available now at your distributor") and the brief article from 12/4/71 ("is being readied for U.S. distribution"). I would say Bushnell's suggestion that it shipped in December 71 or Jan. 72 fits very well.

    Regarding the formal establishment of Syzygy Co., I posted (linked in the timeline) the page from the 72/73 financial statement I was referring to, where it states that Syzygy Co. was organized in January 1971. It seems to me, this is quite consistent with the Atari prospectus that refers to December 1970 (a document I have not seen).

    I also put some time into pinning down announce and ship dates for the various early models of Data General Nova minicomputers. Because those are fixed dates, and because some of the other documents and interviews discussing 1969-1971 refer to specific models of Nova machines, this may help us date more of the events we're more interested in.

    I did some additional work as well, incorporating things posted here, which maybe I won't try to recount here.

    Loving this blog, especially these Atari Deposition transcripts!

    1. Thanks a ton for your work. Your site has been invaluable to my project. I have a couple questions about some of the things I've run across though. Can I email you sometime?

    2. Certainly email me! My email is all over my web pages. :)

      Also, feel free to remove my redundant post from here, sorry about that.

  5. Hey guys, (2nd time attempting to comment on this post)

    I've updated my timeline page for Syzygy/Atari, incorporating and responding to things posted here.

    Regarding the Computer Space release date, I added images (linked in the timeline) of the actual ad from 11/27/71 (proclaiming "Available now at your distributor") and the brief article from 12/4/71 ("is being readied for U.S. distribution").

    Regarding the formal establishment of Syzygy Co., I posted (linked in the timeline) the page from the 72/73 financial statement I was referring to, where it states that Syzygy Co. was organized in January 1971.

    I also put some time into pinning down announce and ship dates for the various early models of Data General Nova minicomputers. Because those are fixed dates, and because some of the other documents and interviews covering 1969-1971 refer to specific models of Nova machines, this may help us date more pre-Syzygy related activities.

  6. Excellent work as always, Michael.