Friday, August 30, 2013

Centuri/Allied Leisure Annual Report Goodies - Part 1: Release Dates and Production Numbers

Recently, I've been looking through some annual reports and 10Ks for Allied Leisure/Centuri and across some interesting info. These things seem boring at first but they sometimes contain a treasure trove of info. For some years they only had one or the other (10k or annual report) , for others they had both. They are very similar and I've seen them referred to interchangably. The annual report contains a lot of the same info as the 10k but is aimed at shareholders so it has more PR type stuff. the 10K goes to the SEC and contains just the nitty gritty.
My library has a set of annual reports for various companies. Only public companies had to file a 10k with the SEC and most coin-op companies were private. Allied Leisure/Centuri was one of the few that did. Others include Bally/Midway, Warner Communications (parent of Atari), Williams, Xcor (parent of Williams before it spun off), Seeburg (parent of Williams before Xcor), Gulf + Western (parent owner of Sega), Columbia Pictures (parent of Gottlieb), AES Technology (parent of Game Plan), Micropin (owner of Elcon), Merit, Game-A-Tron, Ramtek (though only after they stopped making video games), and Status.

Release Dates/Production Numbers

The most interesting thing I found was some info on release dates and sales for various games. I'm sure this isn't nearly as interesting to everyone else, but I find it fascinating and am always on the lookout for production numbers. They don't have actual production numbers but there's enough info for a rough ballpark estimate.

Let's start with 1982:



Here are some more precise numbers





Note thatAllied was selling games (to distribuors) for about $1700 each on average.

Finally, we have this:




Unfortunately, they don't tell us what game goes with what letter, but it shouldn't be too hard to figure out. Are they listed in the same order they are in the paragraph above?
Luckily, they give the release date and I have plenty of info on release dates.
We know from the first document above that game D was The Pit.

Release dates are a bit hard to come by. There is a document floating around the web with Atari release dates but that is an exception. I have a lot of souces I use for release dates, but here I'm going to use 4: MAME, the September 1984 DRA  price guide, the Play Meter "catalog" issue (PM), and the RePlay catalog issue" (R)
The DRA price guide was a price guide to used games sold to operators in the 1970s and 80s. It included release dates for games. The RePlay and Play Meter catalog issues were annual/semiannual issues listing recent games by company, along with release dates. Play Meter used release dates while RePlay usually listed the month the game was announced in their magazine (though the two wer often the same).
If the game appeared earlier at an industry show, I'll note the date the show started.
For the games above, here is what I have:

Challenger: 11/1981 (DRA), 12/81 (MAME, PM, R), shown at 1981 AMOA show (10/27)
Round Up: 11/81 (DRA, MAME), 12/81 (PM, R) - also at AMOA
D-Day: 3/82 (DRA), 4/82 (MAME, PM, R) - shown at AOE (3/26)
The Pit: 4/82 (DRA, MAME, PM, R) - also at AOE
Loco-motion: 3/82 (DRA), 4/82 (MAME, PM, R) - shown at AOE (3/26)
Tunnel Hunt: 7/82 (PM, R), 8/82 (DRA)
Swimmer: 11/82 (MAME, DRA), 12/82 (PM) - shown at West Virginia MVA show (9/26)

I think that Tunnel Hunt is game F and Swimmer is G.
D-Day and Loco-motion are probably C and E.
Challenger and Round-Up are games A and B.

If we assume that the games are listed in the same order in the they appear in the chart, things line up rather nicely. Challenger and Round Up could be transposed, but they both sold about the same number. D-Day and Loco-motion could be switched too, but game C sold fewer than game E.

Can we turn this into actual production numbers? Sadly, no. the percentages listed appear to be percentage of total video game sales in dollars, not units (as will be clearer below). To turn this into actual production runs, we'd have to know how much each game sold for, which we don't.
The annual reports, however, do sometimes list the minimum and maximum sales prices for Centuri's games, so we can use that to get possible ranges of sales numbers (note that a given game could sell for different prices at various points during the year - i.e. Centuri might have offered it at a discount once sales dropped off.

So here is the data for 1982:
 
1982 (fiscal year ended October 31)
Games introduced: 7 (Challenger, Round-Up, Locomotion, The Pit, D-Day, Tunnel Hunt, Swimmer)

Aggregate # of Units produced: 8,682, Gross Revenues: $14,760,922
“Production runs…varied from approximately 500 to 2,000 games.”
Sales price varied from approximately $1,400 - $2,000 (current price of $1895)


Game       Intro     % of Rev           Revenue           Units (low)        Units (high)      

A (Chall. or Round Up) 11/81    9.1%   $1,343,244        672                   959
B (Round Up or Chall.) 11/81    9.9%   $1,461,331        731                   1,044
C (Locomotion or D-Day) 2/82  6.9%   $1,018,504        509                   728
D The Pit                        3/82     20.8%  $3,070,272        1,535                2,193
E (D-Day or Locomotion) 3/82  3.9%    $575,676           288                   411
F (Tunnel Hunt)            6/82      7.8%    $1,151,352        576                   822
G (Swimmer)                9/82      10.1%  $1,490,853        745                   1,065
Other                                           31.5% $6,140,544        3,070                4,386
Total                                                                                 7,380                10,544

NOTE - the percentages listed only add up to 62.3%. The chart only lists figures for games intrduced in fiscal year 1982 (which ended 10/31). I suspect that the discrepancy is due to games from 1981 that were still selling in FY 82. In fact, I'd guess that the other 37.7% is almost entirely, if not entirely, due to Vanguard, which was released late in 1981. If we assume that Vanguard sold for $1700 (which again, is probably not true) we can ballpark that it sold around 3,300 units in 1982.

How about 1981? That one's slightly harder.

 





 The percentages for game A and B are 7.4% and 1.1%.
Again, we known that Phoenix is game C.
Here is the release info:
Eagle: 10/80 (DRA, PM), 11/80 (R) - shown at a distributor show on 9/12
Phoenix: 11/81 (DRA), 12/81 (MAME, PR, R) - shown at AMOA (10/29)
Route 16: 4/81 (DRA), 5/81 (MAME, PR, R)
Pleiades: 6/81 (DRA, R), 7/81 (MAME, PR)
Vanguard: 7/81 (MAME - Japan rls date??), 9/81 (DRA, R), 10/81 (PM)
Challenger: 11/1981 (DRA), 12/81 (MAME, PM, R), shown at 1981 AMOA show (10/27)

These don't correspond precisely to the order listed. (i.e. Phoneix is game C but is listed second). If we eliminate Challenger and game B, howeer, the others line up almost perfectly. Could B be Challenger? The dates are way off but the sales figures seem right (Challenger or Route 16 had to be the lowest selling game). Also, Challenger was the only game on the list that Centuri designed in house. Could they have released it in limited run earlier? Another possibility is that they're talking about Killer Comet, which was also designed inhouse. Centuri released it at the same time as Eagle and later licensed it to Game Plan. Maybe they got the two confused (by the time the report came out, Challenger had been released)

Let's go with the assumption that Challenger is B and the rest are in list order.

1981 (fiscal year ended October 31)

Games introduced: 6 (Eagle, Phoenix, Route 16, Pleiades, Vanguard, Challenger)
 
Aggregate # of Units produced: 31,541, Gross Revenues: $61,460,296
“Production runs…varied from approximately 500 to 15,000 games.”
Sales price varied from approximately $1,400 - $2,000


Game               `Intro    % of Rev   Revenue       Units (lo)       Units (hi)      
A (Eagle)          8/80      7.4%         $4.548,062    2,274              3,249
B (Challenger?) 9/80    1.1%         $676,063       338                 443
C Phoenix        1/81      42.4%       $26,059,166  13,030            18,614
D (Route 16)     3/81     4.9%         $3,011,555    1,506              2,151
E (Pleiades)      6/81     15.5%       $9,526,346     6,805             4,763
F (Vanguard)    8/81     23.6%       $14,504,630   7,252             10,360
Other                              6%           $3,134,475     1,567              2,239
Total                                                                       30,730           43,900


The percentages here add up to 94.9%, so it doesn't look like many 1980 games sold in 1981.

 How about 1983?







NOTE that Centuri changed their fiscal year end from 10/31 to 12/31 in 1983.

We know that Gyruss is game B and Track & Field is game E. Once again, the rest seem to correspond to the list Unfortunately, I couldn't find figures for total video game sales or total units sold in 1983 (still looking through the document).



1983 (fiscal year ended December 31)

Games introduced: 5 (Munch Mobile, Guzzler, Gyruss, Aztarac, Track & Field)

Aggregate # of Units produced: ?, Gross Revenues: $32,529,812
“Production runs…varied from approximately 200 to 14,000 games.”
Sales price varied from approximately $600 - $2,000

Game               Intro     % of Rev   Revenue           Units (lo)       Units (hi)      
A (Munch Mobile)  3/83  1.2%       $390,358           195                   651
B Gyruss          4/83        37.0%     $12,036,030      6,018                20,060
C (Guzzler)       4/83        1.7%       $553,007           922                   277
D (Aztarac)       9/83        1.3%       $422,888           211                   705
E Track & Field 10/83     35.3%     $11,483,024      5,742                19,138
Other                                23.5%     $7,644,506        3,822                12,741
Total                                                                          28,708              95,692
 
Finally, 1984
 

That's all they gave for 1984. Too bad (they had already abandoned video games for sports equipment and seafood by the time the annual report came out).
We do have numbers for Track & Field and Hyper Sports, however (note that the two accounted for 87% of video games sales in fiscal 84, with Track & Field alone accounting for 63%).


 Games introduced: 4 (Hyper Sports, Circus Charlie, Badlands, Mikie)

Sales price varied from approximately $600 - $2,000

Game               % of Rev Revenue          Units (lo)      Units (hi)      
Hyper Sports    ca 63%   ca $12,500,000 6,250             20,833
Track & Field    ca 24%   ca $4,800,000   2,400             8,000
 
So we have some very rough numbers for production runs and some solid release dates.  That was probably mind-numbingly boring for most of you, but I was glad to find it.

Wednesday, August 28, 2013

The Ultimate (So Far) History of Exidy - Part 9


Today's post is the final installment of my history of Exidy (not counting the errata and statistics). Unfortunately, it won't be nearly as complete as the other posts since a) it's after the time covered by my book and b) I don't have nearly as many sources of information for the post-1989 years.


Cheyenne and Combat

 


1985 saw the appearance of the second and third games in Exidy's alliterative shooting series. First up was the wild-west themed Cheyenne.  The game was basically Crossbow in a western setting. The player maneuvered as single character through eight different wild-west settings (a saloon, a canyon, a stagecoach attacked by natives, a graveyard, a ghost town, a forest, and a mineshaft) collecting bounties on outlaws and avoiding enemies. Initially the game was known as Buster Badshot or Busby Grits (after the main character). Pete Kaufmann, however, insisted they name it Cheyenne. Ken Nicholson jokingly suggested that the next game should start with "C" to. Exidy took him up on the idea and dubbed their series of rifle games the "alliterative shooting" series (the idea was abandoned after the fifth game).


            Keeping to the alliterative theme, the next game in the series was Combat. This time the theme was military. The player, armed with an M-16, tried to liberate a series of countries while climbing the ranks from private to General of the Army. An easier version of the game called Catch 22 was also produced. Both Cheyenne and Combat appeared on Replays software charts in 1985 and 1986 with the former peaking at #4 and the latter at #10.

 
Exidy sales manager Mereille Chevalier
            On the business front, things didn’t go as well. Gene Lipkin (formerly of Allied Leisure and Atari) signed on as company president on January 2, only to leave within a few months to head the newly reestablished Sega. Exidy, meanwhile, was having financial problems. In an effort to raise money they worked out a loan arrangement based on leasing their Sunnyvale factory to an outside party at a higher rate than they were paying in rent. The deal fell through when the property owner objected and on July 17, Exidy filed for bankruptcy.
            After 1985 Exidy's video game fortunes went into sharp decline. The alliterative shooting series continued in 1986 with Crackshot (a target shooting game), Clay Pigeon (skeet shooting) and Chiller The latter went on to become Exidy's most controversial game since Death Race (Replay opined that it made Death Race "look like a gumball machine"). It featured perhaps the goriest graphics to date in an arcade video game. Particularly gruesome was a torture chamber scene, that featured a victim chained to a wall, and another having his head crushed in a vice. The idea had not started as a game at all.

 
 

[Larry Hutcherson The scene with the guy chained to the wall was a goof that I put together while I was developing better art tools on the 440 system. One day …[Pete Kauffman] walked by and said, we have to make that game. And one thing lead to another.

Shooting the hapless victims released a veritable fountain of blood (though squeamish operators had the option of turning it green "…for a more 'monster movie' effect if that's the way people want it to appear should they get flak from customers."[1]).
            Exidy also released some non-shooting games in 1986. Top Secret  was a James-Bond style driving game in which the player fought off enemy cars with a variety of weapons, including heat-seeking missiles, oil slicks, time warps, and force fields. It was available as a conversion kit for the company's shooting games or in a sit-down cabinet. The game had been shown as 0077, part of a new series of games (Exidy planned to follow up with 0088 and 0099, but never did so). When the game was fully released, the name had been changed to Top Secret, with the addition of 50 levels, a 360-degree steering wheel, and a new control panel (though one wonders if the estate of Ian Fleming had something to do with the name change).  Perhaps the most unexpected offering was Top Gunner, a licensed first-person shooting game in which the player flew over a series of 3D surfaces fending off groups of oncoming enemy ships. What made the game unexpected was that it was (probably) the last vector graphics arcade game ever produced. While the graphics were outstanding they were also outdated as vector games had gone the way of the dodo two years before. If Top Gunner was the year's most unexpected game, Spin-A-Ball was perhaps the most unusual - a video redemption game that offered the player the choice of four ball games like shuffleboard and Skee ball. The summer of 1986 witnessed two big changes at Exidy. First the company moved its factory to Fremont. Second, veteran engineer Howell Ivy left to join Sega.




            In early 1987 the company moved its headquarters to Santa Clara. In late summer, they final emerged from Chapter 11 and began producing gun games again, starting with Hit 'N Miss (the alliterative idea had finally been dropped). Unlike the gory Chiller, Hit 'N Miss was a kiddie themed game without a trace of blood. In the December issue of Replay, Pete Kauffman explained why, noting "We've learned the hard way from our past six shooting games that blood and guts are not accepted in the long run."  On a slightly more risqué note, Exidy also released one of its most unusual products in 1987 - a condom vender called the Rainbow Machine.  (They also announced a video war game called Under Fire that was apparently never produced).
            Exidy's final gun games came in 1988. WhoDunit was a mystery themed game in which the player used a "dueling pistol" to protect a character named Max as he made his way through s house searching for a hidden key. 1988 also saw Exidy try a new tack with Showdown, a video poker game that featured comical animated, talking opponents whose tells often revealed when they were bluffing. The game was available in a dedicated cabinet, and a countertop cabinet. The most unusual configuration was a conversion kit for Exidy's gun games, perhaps the only case of a poker game that used a rifle controller. The rifle could could also be used to play a bonus gunfight round. Payout poker games were still illegal in many locales at the time and so called "amusement only" card games were suspect (often with could reason). Showdown may have been the first, if not the only, video poker game to truly merit the label. Larry Hutcherson also developed a credit version of the game called Yukon for sale in Malta. From 1984 to 1988, Exidy concentrated the majority of its efforts on its gun game line. The first two, Crossbow and Cheyenne had proven enormously popular. Perhaps too popular. While Exidy released nuermous update kits and rebate programs for the games, no one wanted to buy them. Pete Kauffman's wife Virginia, who took a job as the company's sales manager in 1986, explains why.

[Virginia Kauffman] I found out that many Crassobws, Cheyennes, and Combats didn't need a software update because they were still taking good money in. It was a thrill to hear that, but a little frustrating when you were trying sell a conversion[2].

 
 

In 1989, Exidy finally gave up and exited the video game field entirely to concentrate on redemption games. First came Twister, a bowling/roll-up game where the player rolled a ball at a moving target. More followed in the 1990s, including Hot Shot, Troll, Critter, 4x4 (a truck racing game) and Turbo Ticket (a ticket blowing machine where the player put their arms through two circular openings and grabbed swirling tickets). In 1995, Exidy was finally forced to call it quits after an employee embezzled over $300,000 from the company (though some sources indicate that they continued to produce redemption games under the Xidy label). In 2006 Mean Hamster Software acquired the rights to develop coin-op games under the Exidy name.



Bonus Pictures

 








 
 


 
 
 
 
 
 
Hardware Designer Mark Von Striver
 
 
 


[1] Pete Kauffman, Replay, April, 1986
[2] RePlay, August, 1988

          

Wednesday, August 21, 2013

Early Unknown Computer Games Mentioned in the Magnavox Suit(s)

Recently, I was going through some of the Magnavox Vs. Activision documents that are available at the University of New Hampshire School of Law website (it looks like they were donated by Ralph Baer):

http://ipmall.info/hosted_resources/Activision_Litigation_Documents/Activision_Lit_Doc_index.asp

For those who don't know, Magnavox launched a number of patent infringement lawsuits beginning around 1974 against various companies (they were also sued themselves, I believe).

Here are some of the coin-op companies named in the suits - along with the games they mentioned in the suits):
·         Atari
·         Bally
·         Midway (Playtime and Winner)
·         Seeburg (Pro Hockey, Pro Tennis, Paddle Ball [all by Williams, which Seeburg owned at the time), Olympic Tennis [released by See-Fun, another Seeburg company, though I think it was just a brand name they used),
·         Allied Leisure (Paddle Battle, Tennis Tourney)
·         Chicago Coin (TV Ping Pong, TV Tennis, Olympic TV Hockey, TV Goalee)

They also mentioned Ramtek's Hockey and Soccer and URL's Video Action III, so they may have sued them too. Consumer manufacturers named in the suits include Sears, APF Electronics, and Fairchild.

Most of the documents were rather boring but there were some gems in there (testimony by Willy Higinbotham, for instance - lists of which Activision games infringed on their patents and which didn't - Activision sales figures).

The info I found most interesting was a list of early computer games and other devices that were mentioned in the trial. One of the tactics tried by the companies in the suits was to claim that the Magnavox patents were invalid because there were earlier inventions that constituted "prior art" or invalidated the patents for other reasons.
All such attempts failed, but it appears that these games came up again and again.
Some of them are very well known to video/computer game historians.

Spacewar and Tennis For Two were brought up numerous times. They even trotted out the old MIT Bouncing Ball program (I don't think they were implying that this was a game. My guess is that they were offering as an example of a program with "coincidence detection".)

I don't think the old "Cathode Ray Tube Amusement Device" was on the list, though the patent was referenced. If not, this was probably because this patent was actually referenced in one (or more) of the Baer patents. One of the claims the various companies made was that Magnavox and/or Sanders was aware of these games at the time they applied for their patents but failed to mention it to the patent examiner.
There were a number of games, however, that I've never seen mentioned anywhere else and some sounded quite interesting.
Here is a list, with a few notes.

 Michigan Pool Game (1954)
In 1954 William George Brown created a pool simulation at the University of Michigan. The program was demonstrated at the Association for Computing Machinery meeting in Detroit that same year. It was a two-player pool game with 15 balls and a cue ball that made use of a CRT.  

Training Device For Marksmen (1960)
Fritz Spiegel's Military Trainer (1960)
 


 
 
In 1960, the US patent offices issued a pair of patents[1] for a military training similar developed for the German aircraft manufacturer Bolkow (they had been patented earlier in Germany). The device included a CRT and a pair of controls. The trainee used a joystick-like device to steer a guided missile that appeared on the screen towards a tank while avoiding flying the missile into the ground. A second controller allowed an instructor to move the tank or set up various training scenarios. The second of the two patents was issued to Fritz Spiegel, an employee of Messerschmitt (Bolkow merged with Messerschmitt in 1968). The defendants claim that the device was (or could be) connected to a standard television set. The patent was later sold to APF Electronics, who manufactured the 1977 Pong console TV Fun. Magnavox sued APF in 1977 (they learned of the Spiegel patent the same year) but the case was dismissed for lack of venue. In 1981, Magnavox acquired the Spiegel patent. In the Magnavox/Activision case, Magnavox acknowledged (or at least refused to deny) that the Spiegel patent constituted "prior art" in relation to the "507" Baer patent. While the device made use of a CRT (and apparently a raster-scan display), it was designed as a training device, so calling it the first video "game" would seem a bit ludicrous.

NOTE - Much of the above info was taken from Activision documents, so you may want to take it with a grain of salt. While it appears that Magnavox admitted that this constituted prior art, the court did not find that it invalidated their patents. I haven't really looked into the reason why, because that's not my primary interest.

Rand Corporation Handball/Jai Alai Game (ca 1963)
Unfortunately, no details on this one are provided in the available documents for the case.

GE/NASA Scene Generator (1964)
In 1964 NASA purchased a system from General Electric that simulated various docking and landing maneuvers. The program included numerous modules: a lunar landing simulation, a spaceship docking simulation, a "game" in which the player controlled and airplane (or helicopter) and fired bullets at a moving tank, an aircraft-carrier landing simulation, and an airport landing simulation.  The system used a color raster-scan display and made use of a joystick-like controller. It was designed by Jim Van Artsdalen, (possibly the same person who later worked on Origin's Ultima III and Ultima IV), James Lawrence, and James Smith. Like the Spiegel device, however, this appears to have been more of a training program than a video game.
 
IDI Pool Game (1966)

John Drumheller graduated from MIT in 1964 with a bachelor's degree in mathematics. During his undergraduate years, he developed a number of programs for the PDP-1 in building 26, working alongside the hackers as they developed Spacewar. Among the programs he created was a Mill game (the Scandinavian equivalent of checkers) and a Go algorithm. After graduating, Drumheller went to work for Adams Associates. In 1966, Information Displays, Inc. of Armonk, New York approached Adams about developing a demonstration program for their monitors. Drumheller, who just two days earlier had started experimenting with a program that made balls bounce onscreen, volunteered to write a pool game. Written on a DDP-116, the game was shown at the Joint Computer Conference in San Francisco in the fall of 1966. Two players used a light pen to maneuver a cue stick and play a (somewhat) standard game of pool. The game included 15 balls, a crude form of English, and kept score automatically. In 1967, Drumheller's brother-in-law Peter Mullarky created a version of the game for another computer that used a vector display. From Drumheller's courtroom testimony it appears that the original version may have used a raster display, but this isn't clear. Drumheller also claims that he considered making a consumer version of the game that would run on a standard TV set. He also claims that he was later told that someone at MIT was developing a pool game at the same time he was.

NOTES - Drumheller's testimony is actually available on the site and makes for good reading. At least most of it does. A lot of time is spent discussing exactly how you hit the ball. Apparently, it was possible to hit the cue ball without actually moving the cue stick. The program just sensed where the light pen was and ignored the cue stick (though if you touched the cue stick, it moved). The court went to some length to note that this (ignoring the cue stick) was not the normal method of play.

My guess is that this was done because I believe that one of the patents covered having one object on the screen impart motion to another object and they wanted to show that this program did so (if the cue stick graphic didn't actually move, you could argue that it didn't).

RCA Pool Game (1967)





What is it with pool games? In 1967, RCA demonstrated a pool game on their Spectra 70 computer during an open house event celebrating the 25th anniversary of the David Sarnoff Research Center in Princeton. The BBC filmed the game, which was also the subject of an AP news story. The game was mentioned in the Magnavox cases and was referenced in one of the Ralph Baer patents from 1977.
NOTES - I am not sure that this was created for the Spectra 70. The text doesn't say for what system it was designed. But the Spectra 70 manual and a number of other Spectra 70 materials were entered as evidence, so I'm assuming it was. 

CDC Baseball Game (ca 1967)
This game was briefly mentioned in documents for the Magnavox/Activision case but Activision doesn't seem to have done much research into the game and no other detail is given in the available case documents. They were likely talking about that various games that were developed for the CDC 6600 (which included games like baseball, Lunar Lander, and Spacewar). A baseball game called BAT was created for the system, but that may have been a different game. Announced in 1964 and released in 1966 with a price of $7 million, the CDC is considered by some the first "supercomputer". The baseball game may have been programmed by Thomas J. Spence (who is mentioned in the Magnavox/Activision documents in relation to the game).

NOTES - When I first started researching this one, I came across this:

http://www.ebay.com/itm/Vtg-COMPUTER-BASEBALL-Electronic-Data-Controls-1969-Table-Top-EARLY-VIDEO-GAME-/110602045332?pt=Electronic_Battery_Windup_Toys_US&hash=item19c0654794

and briefly thought it might be the game they were talking about.
It wasn't but that sure looks like a cool game (though not for $825).

 



[1] Patents 3,046,676 and 3,135,815.

Thursday, August 15, 2013

Who Was the First Person To Come Up With the Idea For A Coin-Op Video/Computer Game?

Recently I came across some info about some possible early attempts at a coin-op video game that I thought some might find interesting (note that here I use the term "video game" in its current common-use sense - I am fully aware of the controversy over the definition of the term and that many don't consider vector games video games, but that's a topic for another time).

Nolan Bushnell and Ted Dabney were probably the first people to actually start buildign a coin-op video game (or at least the first anyone knows about) but who was the first to think of the idea? (and I'm talking serious consideration here - like someone who was serio.us about doing it)
Let's look at a few contenders.

Nolan Bushnell

This is probably the answer most would give. Nolan claims that he first saw Spacewar at the University of Utah. Varying accounts date this incident from as early as 1962 to the "mid-60s" and claim that Bushnell thought of monetizing the game either soon after seeing it, or while working at the Lagoon Amusement Park during this same period.
But is this account true? Ted Dabney claims that Bushnell never mentioned Spacewar (or the idea of a commerial video game) to him until he saw the game at Stanford around 1969. Of course, just because he didn't mention the game to Dabney doesn't mean he hadn't seen it earlier, but there is still some question regarding the accuracy of his account of having seen the game as an undergrad. IIRC, Goldberg and Vendel wonder if Bushnell claimed to have seen it in the mid-60s in order to establish priority over Ralph Baer (who started working on his video game in 1966). They also pointedly (again, IIRC) omit any mention of Nolan seeing the game at U of U. I did a quick search to see if I could verify that University of Utah did have the game by the mid-1960s, or even if they had a PDP-1 by then (since DEC supposedly included the game on all the units they sold) but didn't really find anything. Of course, even if he did see the game in the mid-60s, that doesn't mean he saw the dollar signs at that time.

So, if Nolan didn't come up with the idea until 1969-ish (and I'm not saying that is or isn't the case - his account may be entirely truthful), who might have done so earlier?

Steve Russell et al

At least one source I read mentioned that the Spacewar design team briefly discussed the idea of trying to sell the game but quickly dropped it, figuring that DEC was their only potential customer and they wouldn't be interested. Even if this is true, however, it doesn't really count because a) it sounds like it was nothing more than a brief discussion rather than a serious proposal and b) it wasn't for a coin-op version of the game.

Hugh Tuck

According to Galaxy Game designer Bill Pitts, Hugh Tuck was the one who suggested creating a coin-op version of Spacewar. But when did this occur? Pitts said he saw the game around two years after he enrolled at Stanford in his Junior year. Pitts enrolled in 1964 so that would place the incident around 1966 or 1967. He also says that he realized Tuck's idea was feasible about three years later when he was working for Lockheed and saw that DEC had released the PDP-11. The PDP-11 was released in December, 1970 but I don't know when Pitts found out about it. The two of them formed Computer Recreations in June of 1971, however. From this info, Tuck's suggestion of monetizing the game came somewhere between 1966 and 1968.

Richard Ball

According to the Wikipedia entry for Nutting Associates, engineer Richard Ball drew up a proposal for a coin-op video game called Space Command. After Nutting passed on the idea, Ball and coworker Ransom White formed their own company called Cointronics. There are some problems with this story, however. First, the Wikipedia article is completely unsourced. In 2012, someone named Judith Guerin left a comment on an article by Benj Edwards  about Computer Space that made essentially the same claim, but again no source is given and I don't know who Judith Guerin is. I don't know if she was using Wikipedia as her source or vice versa or if they both used a different source.
Even if the story is true, however, when did it occur? A Google books search revelas that Cointronics showed two games (Zap Ball and Ball/Walk) at the 1968 MOA show so they were around by then. A search of the California Secretary of State website shows that Cointronics incorporated in early 1969 so it seems to be a good guess that the company was formed sometime in 1968 and Ball left Nutting shortly before. I was unable to track down Ball or White (though I did confirm that White usually went by "G. Ransom White"). So let's say this incident occured in 1968.

Unknown Stanford Student/Staffer

FInally, there's this video:
http://www.youtube.com/watch?v=J8zU8WQO-PY

This is from a 2006 panel discussion that was part of a celebration of the PDP-1. The panel included a number of members of MIT's famed hacker community of the 1960s, including Alan Kotok, John McCarthy, Peter Sampson, and Spacewar creator Steve Russell. During a q-and-a session after the discussion, an unidentified audience member claims that in 1966 or 1967 American Express provided Stanford with seed money for a program to come up with potential projects for future development. The person claims he drew up a proposal to put a multi-player version of Spacewar on a dedicated, timeshared PDP-1, put it in a bowling alley or shopping center, and charge people to play. Stanford considered the idea too frivolous to submit.
Unfortunately, I have no idea who this guy is. I don't think it's Hugh Tuck (plus, his story is very different from Tucks) but if his story is true, he might be a good candidate. True, his game likely wouldn't have had a coin-slot, since it seems he was talking about charging per block of time, but he was thinking about putting the game in a bowling alley.

Someone else

Of course, ideas are cheap and there could well have been any number of people who entertained the idea of a coin-op "video" game far earlier than any of those mentioned above (and some would say that it's irrelevant who thought of the idea - it's who implemented it that matters).

Odds and ends

Speaking of Computer Space, here's an early article on the game from the May, 1972 issue of Vending Times






And speaking of Nolan Bushnell, take a look at this:





So who is that and what does he have to do with Nolan Bushnell?
That's John Bushnell, Nolan's great-grandfather. Born in England in 1823, John came to America and went to Utah in 1850 in one of the early companies of Mormon settlers. At one point, he actually wanted to return home but Brigham Young sent him to the new town of Fillmore, where his family was one of the first six in the area. He opened a store and a post office before trading his property for a farm in nearby Meadow in 1862.

OK, that really doesn't have anything to do with video games, but I found it interesting.

Finally, speaking of Cointronics, here's a photo of their countertop game Ball/Walk (games like this drive me nuts since I suck at pretty much any game other than pinball that involves a steel ball - I tried to play Ice Cold Beer once and I think I got one ball in hole).