Collected Works
He was the golden boy of the golden age of arcades – and even has the T-shirt to prove it
Ed Logg, Golden Boy of the early Atari era, on Asteroids, Gauntlet, Xybots and more
SUPER BREAKOUT
Developer/publisher Atari, Inc Format Arcade Release 1978
ASTEROIDS
Developer/publisher Atari, Inc Format Arcade Release 1979
CENTIPEDE
Developer/publisher Atari, Inc Format Arcade Release 1981
GAUNTLET
Developer/publisher Atari Games Format Arcade Release 1985
XYBOTS
Developer/publisher Atari Games Format Arcade Release 1987
TETRIS
Developer Atari Games Publisher Tengen Format NES Release 1989
WAYNE GRETZKY’S 3D HOCKEY
Developer Williams Entertainment Publisher Midway Format N64 Release 1996
When John Salwitz [co-creator of coin-op classic Paperboy] left Atari, he had a T-shirt made saying ‘Golden Boy’ and gave it me as a parting gift,” Ed Logg says, grinning. “I think he started that. And when I was at Tengen, I was heading to my first CES and they said I needed a business card. They asked what I wanted as my job title and I said I didn’t care. A producer, Howard Lehr, suggested ‘Super Duper Game Guy’. I said, ‘Fine.’”
Monikers to be proud of, indeed, but then when you’ve created Atari’s bestselling coin-op, Asteroids, along with such iconic titles as Gauntlet and Centipede, they are perhaps to be expected.
It didn’t start out like that, though. When Logg joined Atari in 1978, he was assigned to a project already in development, Dirt Bike, and later joined the team working on Wolf Pack, an ambitious submarine commander game featuring a hulking periscope you could rotate 360 degrees. Neither went into production. After all that, did he ever wonder whether he’d made a mistake getting into the fledging world of videogame development?
“Oh, I had no doubt I’d made the right decision,” he says today. “You make something and it doesn’t always work. Games can look good on paper but you take them to the public and they don’t always like them.”
Logg’s early games were literally drafted on paper, such was the state of game development in those frontier days, but he stayed in the videogame business through the next three decades, producing dozens of titles for both the arcade and home consoles. In 2012, he was presented with the Pioneer Award by The Academy Of Interactive Arts & Sciences. Here, he takes us on a journey that begins in Sunnyvale, CA among piles of paper tape.
SUPER BREAKOUT Developer/publisher Atari, Inc Format Arcade Release 1978
I remember playing the original Breakout at a pizza parlour in San Jose when it came out [in 1976]. I thought it was really fun and I put quite a few quarters in. A few years later, when I was at Atari, I was talking to Owen Rubin, one of the programmers, over lunch and he said Nolan [Bushnell – co-founder of Atari] suggested he do an update of Breakout, with multiple balls and so on. I said, ‘I can do that!’
Back then, we were using 6502 processors and we had a black box which we fed paper tape into. That loaded code into memory and simulated the EPROM. You wrote up your program on paper and gave your listing, with instructions to make any changes you needed, to some gals in the office who would type it up on a PDP system. They’d give you a new printout and paper tape after a few hours and you could feed that back into your box. That meant I had time between making the changes and when I got the new listing back, so I ping-ponged between Dirt Bike and Super Breakout.
The original Breakout was all done in TTL [discrete logic] so there was no
code to look at and I didn’t have any notes from [Breakout project manager] Al Alcorn or anyone else about the algorithm for the ball angle or the speed. I wrote it in 6502 Assembler from scratch and tried to mimic it as best I could.
I came up with six different versions and I remember I had versions with side collision which didn’t make it into production. They picked three to use in the final game and the most popular was Progressive. I think it was because of the anticipation of the bricks getting closer and closer and then there was that moment when you got the ball between two layers of bricks and it was pingponging around, racking up the points. That was very satisfying.
It was good to have a successful game out there. I could tell when I was playing Progressive it was going to be a winner. And when I saw Arkanoid years later, I thought, ‘My God, I should’ve done that!’”
ASTEROIDS Developer/publisher Atari, Inc Format Arcade Release 1979
Asteroids really started with a conversation I had with [Atari head of engineering] Lyle Rains. We’d both played a game being made by [Atari engineer] John Ray’s team and we’d both done the same thing. There was a large asteroid in the middle of the screen and you had your ship, so you shot at it – everyone who played the game did. But it didn’t do anything! Lyle said, ‘What if the asteroid blew up?’
I said, ‘OK, what if it split into smaller and smaller parts, because then you bring in some strategy?’ You’re OK with a few big asteroids but if you have tons of smaller ones, that’s chaos. And I knew I needed something to stop you just sitting in one place, doing nothing, so that’s why I put in the saucers, to shoot at you.
Lyle wanted me to do it in raster but those screens were 320x240. I’d already done the alphanumeric [font] for Lunar Lander so I was familiar with the X-Y monitor we had at the time and knew it was 1024x768, so it was umpteen times better. I’d also played Spacewar! back at the Stanford AI lab [when I was in college] and really wanted that higher resolution. A lot of the inspiration came from Spacewar!
When you’re working on a game, you know you’ve got something when you arrive in the morning and you can see people have been playing your game all night. I’d implemented a high-score table, which I’d taken from an Exidy game, I believe [Star Fire], and it would be filled up when I arrived, usually with the initials OR, which was Owen Rubin. I actually put in some code so that if he entered his initials, it would just blank it out. He came to me the next day saying, ‘Hey, I think there’s a bug in your program…’
In the original version of the game, the saucer would shoot based on a timer, so if you shot him just before he was going to shoot, next time he’d shoot immediately. I didn’t reset his timer so if you were next to him, you were a dead duck – and people complained. So I reset the timer each time, to give you a chance. When I started hearing about these huge scores, I realised people had got really good at picking off the saucer before it could shoot at them but, you know, it didn’t seem to hurt the cash box. I actually made a new EPROM which got rid of ‘lurking’, because operators asked for it, but I heard they pulled them out because it killed earnings.
What I do regret is not limiting the number of extra lives you can accumulate to, say, ten. People would fill the screen with, like, 200 lives and the whole game would slow down because the vector beam took time to draw everything.
As for why Asteroids was so successful, I’d have to go back to Nolan Bushnell’s [mantra]: ‘Easy to learn, difficult to master’. You look at the game and straight away you can see the objective. You know what to do and if you die, it’s not the game, it’s you that screwed up. The first field test was in an arcade up in Sacramento and I watched a guy put a quarter in and he died after 20 seconds. You’d think he would just walk off but he put in another quarter. He knew he could do better.
CENTIPEDE Developer/publisher Atari, Inc Format Arcade Release 1981
Centipede came from a brainstorming session off-site. People would present ideas, we’d vote on them, then split up into committees and massage them, and after a few days we’d have a book of game ideas. One of those was ‘Bug Shooter’. It was just a few short sentences – about shooting bugs. No details, no suggestion of how you’d implement it or what it would look like. Just the concept.
They’d promoted me to staff engineer, a position I resigned from after a year because I didn’t want to manage people – I wanted to make games. I’d just hired a new programmer, Dona Bailey, and there was a technician that worked on the cabinet, an engineer that worked on the electronics, and I was project leader and game designer. Some things I did included getting the legs moving – that ripple, like a centipede actually moves – and the selftest stuff. The colour scheme and a lot of the programming was done by Dona, though we did change some of the design after we’d got the prototype working. The mushrooms were initially a static field and Dan Van Elderen said, ‘Why can’t you shoot the mushrooms?’ I thought about it overnight and realised that meant you could create patterns. Then you would need something, the flea, to add new ones, and the spider could destroy them. It created a very dynamic playfield.
Trackball was the only way to go: you could go very fast across the screen but also have very fine micro-movements. It gives you that freedom. Marketing wanted us to use a joystick, so I had Dona do a version, but it was never going to play well. We had a problem with some trackballs because the rollers got worn out and would develop a groove in them. That’s how much play they got. We had a similar problem with Asteroids – on some machines the buttons stopped working because they were played so heavily that there was a build-up of carbon on the contacts. It was the first time we’d seen that. We had to redesign the switches.
Little-known fact: in the cocktail version, one of the cabinet engineers noticed the second player always seemed to get better scores. The spider is programmed to reduce his range [as you progress] and get closer to player one but when the screen is flipped for player two, the spider stays off the bottom and you can sit there and shoot him. Yes, there was a bug in Centipede!
GAUNTLET Developer/publisher Atari Games Format Arcade Release 1985
My son had been bugging me for a year to do a game based on Dungeons & Dragons. I couldn’t see how to do it but then [Atari engineer] Robin Ziegler brought in this game, Dandy, and I thought, ‘Oh crap, this could work!’ Robin and I wrote up the game design, then mapped out the scenarios and the characters with their attributes. We hadn’t really got the hardware to do it on so I asked an engineer to work on getting more than the 16 motion objects running [per scanline], and he created a system that could pretty much fill the screen with objects.
I created a level editor with a compression algorithm to save the level data. I went on a sabbatical in 1985 for six weeks and left my co-programmer, Bob Flanagan, working on it. We had several dozen levels when I left and when I came back there were over a hundred. People just came in and made their own – anyone [in the department] could create a level.
Having multiple players is central to the success of the game in many ways. That came from Dandy, which showed it could be done, but my problem was: how many players could I get round the cabinet? I thought maybe four was my limit. Marketing said that we couldn’t get four people who didn’t know each other playing the game together. They said the game had to have an ending. They put up a lot of resistance. Gave me a lot of shit.
I also had to have the monitor at an angle, not vertical, so everyone had an equal chance to see what was going on, and if you do that, you can’t have plexiglass over the monitor because overhead lights shining on it would create blind spots. A lot of operators said that players would break the screen without that protective glass. We dropped large ball bearings on it from a height to show that you’d need to hit the screen with a hammer to break it.
The original hardware was another problem. The board was huge. It was too big to fit in the cabinet, so I investigated the idea of doing a four-layer board, and that let you shrink the board and really reduced the RF noise, too. It was great. I even made a twoplayer Gauntlet cocktail cabinet for Dan Van Elderen while he was on holiday, and when he came back he thought that was fabulous.
We put it out on field test in a tiny little arcade out in El Paseo de Saratoga and the deal with the operator is that they’re supposed to keep it secret. After a few days, I went to check on it and there was David Rosen of Sega with some Japanese engineers photographing the game. The operator must have tipped them off. And I know Gauntlet has the record for most takings in one week – $2,400 in an arcade in Toronto. That’s paid for itself in seven days!
XYBOTS Developer/publisher Atari Games Format Arcade Release 1987
Xybots was supposed to be Gauntlet III with a firstperson perspective but marketing told us, ‘We don’t want any more Gauntlet’. I didn’t pursue it and I regret to this day not putting Gauntlet characters in the game. Instead, I took the Major Havoc character [from Atari’s 1984 coin-op] and created two buddies for him.
I know the game was before Doom but I did get the idea of a vertical drawing scheme to make that kind of shooter from some PC game on the market. No, I don’t think I started anything new. There were games out there doing this kind of stuff.
I only had stamps of 8x8 pixels for the background and trying to do rotations with just that hardware that looked right wasn’t easy. I wrote down all the angles and the stamps I would need, and drew it out. The difficulty is, you need to know what’s going on to your left and right and behind you but you can’t see it, so I had to create a radar, and that was difficult to understand for a lot of players. No one likes getting shot in the back.
People weren’t used to the controls, either. That game suffered because it was getting away from the simplicity of a joystick and a button. We had problems with that, even though I put
“GAUNTLET HAS THE RECORD FOR MOST TAKINGS IN A WEEK – $2,400 IN AN ARCADE IN TORONTO”
something on the knob that said ‘turn to turn’ [laughs]. People just didn’t realise they could turn, and that affected its appeal. Also, the radar being in 2D and top-down and then your view is in a 3D firstperson perspective – it was just more complex and needed more thought [to play]. It’s just not obvious.
TETRIS Developer Atari Games Publisher Tengen Format NES Release 1989
I saw an Atari ST version of Tetris and loved it so much that I went to our chief legal counsellor and said, ‘You need to license this game’. Dennis Wood got the licence from a British company who got it from the Hungarian who licensed it from the Russians.
I did the NES version for Tengen and took it to the January 1989 CES. The marketing people wanted me to make some updates and I worked out how to do a better version, and then we took it to the June CES later that year for our official rollout. I remember we did the worldwide launch at the Russian Tea Room in New York and a game tester and I competed for the World Championship. I won, by the way. I think that makes me the first world champion of Tetris.
I can’t remember exactly when I heard there was a problem with the licence. It was something about Nintendo having the rights to console versions. I presume Nintendo went straight to the Russians, offered them a million dollars and said, “Fuck the other guys”. We received an injunction. It was very strange because if I remember right, we were going to trial and Nintendo were asking for a summary judgement or something… We sued them, we lost and were told to get all our products off the market. Now, to show you how good our version of Tetris was, people would go to video rental stores [where the cartridge was still available], rent our game and then keep it and pay the fine as if they’d lost it.
WAYNE GRETZKY’S 3D HOCKEY Developer Williams Entertainment Publisher Midway Format N64 Release 1996
The coin-op industry had gone downhill in the ’90s and I really wanted to do some consumer stuff. That was where the big numbers were showing up. I’d gone to EA for a while and then I got a call from Dan Van Elderen, saying, “Why don’t you come back and work on Wayne Gretzky’s 3D Hockey for the N64?”
I think we started working on the hardware about ten or 11 months before the N64 was released, with a modified development system that wouldn’t check for security codes to see if the cartridge was legit. I’m sure we had to sign a bunch of NDAs. The N64 chip was done by Silicon Graphics, which was located not too far from Atari, and I knew an engineer there, so I could ask him questions. It was one of the few titles available at the release of the console, and that helped a lot too. Its sales exceeded all our expectations.
For a good sports game, you have to make it as realistic as possible. A licence with the team names, the players and the colours gives you the sense that you’re looking at the real guys in their real positions. But then it needs to be fun! You need to get all the mechanics right, the controls right… There’s a lot you can get wrong. I always like to throw something new in, that’s different or better than players have seen in the past. I was superhappy with the instant replays – the way it would zoom in and you could see the puck bouncing off the edge of the cage or ricocheting off people’s legs. I thought that was fabulous.