Interview: Kevin McGrath (2025-04-25) by Alexander Rojas

This is an unaltered copy of an interview of Kevin McGrath, for use as a primary source on Sega Retro. Please do not edit the contents below. Language: English Original source: Alexander Rojas at Sega Retro

AR: Could you briefly describe your background in technology? What got you into the game industry?
 
KM: My first programming job came about while I was in High School in 1982, a small Minnesota State venture which turned into a private company called the Minnesota Educational Computer Consortium / Corporation or MECC. They publish educational software (games?) and distributed that software around the country's various school districts. The school district could buy a license and make as many copies of the software collect as they wanted. Probably the most famous title being The Oregon Trail, the "You've died from dysentery" classic! I worked on the more mundane titles, to teach mathematics or how to use an automated teller machine (ATM, gotta learn somehow), written in Applesoft BASIC for the Apple ][. I left MECC in 1984 since I was no longer going to be a student, so "student programmer" didn't fit. MECC gave me a great foundation for mass market software development and even introduced me to Steve Jobs when he came for a keynote at a MECC conference. Most importantly, the other programmers at MECC introduced me to "egoless" programming, which I tried to strive toward throughout my career.
 
After MECC, I did software development for various small businesses. This included a short lived company a friend and I started with the intent of making a multiuser community message board (something in-between a BBS and Reddit, called "Topix"), until 1988 when I joined Dynamix, a small game development studio in Eugene Oregon. At Dynamix I mostly worked on porting the Apple ][ game Wings of Fury to the IBM PC and introduced a "virtual machine" library to help make game development and porting easier.
 
But then, they fired my girlfriend, who at the time was an artist at Dynamix. She wanted to move to the Bay Area in California, it had a much bigger marketplace for game artists than Eugene did. So we moved to California, she got a job quickly at Electronic Arts and I didn't think I could because of my non-compete agreement with Dynamix...
 
AR: Can you tell us about your time at EA and what you worked on there?
 
KM: ...But EA convinced me that they would handle any legal fallout and I interviewed with them in 1989 and started on their super secret reverse engineering of the Nintendo Entertainment System, which Jim Nitchals and I worked on.
 
This made Jim and I "dirty" on the NES (because we had reverse engineered code that had a copyright), so they moved us to another building, away from the other in-house programmers. This was a clean-room style of reverse engineering, like Compaq did with the PC BIOS, so we were "dirty" on the NES, but "clean" on the SEGA Genesis. So Jim and I became the team to port games to the Genesis from EA's library, the first one being Populous. We worked very long hours getting Populous to run on the Genesis from the Amiga. Both 68k CPU machines, but very different in everything else (very little RAM, no mouse, controller only, character driven as opposed to the nice fast bitmapped graphics the Amiga had). It was very challenging, but as I recall, we got it done before the 1990 CES in January, where EA struck a self-publishing deal with SEGA.
 
After EA struck the deal with SEGA, we were no longer "dirty" so Jim and I moved back with the other in-house programmers and I started working on porting The Immortal to the Genesis as well as other small projects, like the animated EA introduction logo used on all of the early EA Genesis games. The Immortal port went as well as can be expected, in my opinion... I was given 768KB of cartridge ROM to work a 2MB game into, and once I got it to fit, they increased it to 1MB of ROM. This meant I had a "huge" amount of ROM left over, so we decided to fill it with very gory animated death cut scenes. The industry rumor mill has it that Genesis version of The Immortal influenced Mortal Kombat to make gory animated death scenes, but I was influenced by the Apple ][ Swashbuckler game I used to play at Dynamix, I got the keystrokes down so I could always decapitate my opponent.
 
Not long after The Immortal port, my career ambition got the better of me and I switched from being a programmer to being a technical director. That role included resolving technical problems for games in development and doing technical appraisals for external developers who wanted to work with EA.
 
AR: What compelled you and Kenneth Hurley to leave EA and found Futurescape? What were your goals in starting this new company?
 
KM: Once again, career ambition got the better of me, and I figured the time to start a Genesis game development house was 1992. It seemed to me that SEGA was winning that segment of the console wars and would be around for decades (whoops). I had known Ken from my days in Eugene Oregon, and he also had entrepreneurial ambition like myself, so we both thought that now is the time to start our own game studio.
 
AR: Do you recall what led Futurescape to working with Sega of America?
 
KM: Ken and I didn't have a project when we left EA, but we knew that we could develop Genesis titles and we had some ins at SEGA of America, so we hoped we could get a development contract for a couple games with either SEGA or EA. We got the SEGA of America deals first worked out of a garage converted to office space. I think we flipped a coin to determine who would be President, which I "lost" and ended up titling myself "VP of Development", which sounds impressive until you learn it's a two person company.
 
AR: Futurescape worked with a number of Sega of America's Genesis peripherals, like the Menacer light gun and Sega VR headset. How much bespoke hardware had to be created on your end to begin developing for these peripherals? Could you describe the company's hardware development capabilities in this regard?
 
KM: After signing development contracts with SEGA, we started proper development in earnest, leased some real office space between EA HQ and SEGA of America HQ and started hiring artists, programmers and some business management help. I lead development for a game we named "Nuclear Rush" for the yet to be released VR headset for the Genesis, and Ken worked on a light-gun game. As I recall, the light-gun hardware was pretty solid, but the VR headset proto-type was non-existent. I had to make my own little video switching circuit to switch between the left and right eyes, so 30 Hz per eye, because they just didn't have the hardware to a point where we could develop using it.
 
Ken and I both agreed that we couldn't afford the standard development kits from SEGA, I think they were around $5K each and the artist stations were not much cheaper. Instead, mostly Ken (I did a little, but mostly Ken did the hardware and I did the firmware work) developed a really slick cartridge emulator with a 9 volt battery backup. It was faster and supported larger cartridges than the SEGA kit, and you could carry it to an office with your game on it to show progress, plus the paired down artists version was super cheap for us to make, so it was no issue to give one to each artist.
 
AR: What were some of the biggest challenges in working with the Sega VR headset? Accordingly, can you tell us about Nuclear Rush?
 
KM: I don't recall ever getting a fully workable prototype for the VR headset. I had an early prototype with two eye displays, but no video switching, and later got the head tracking prototype, but never a complete unit and the hardware I used never looked like the advertised picture. Over all we were excited at the thought of having one of the games that shipped with the hardware, but this was really toy VR. The headset itself had pretty poor displays with very limited field of view and the head tracking was plagued with tracking delays, which would get you motion sickness after using it for a little while, and the Genesis is just not a real powerful computer. 3D graphics hardware for the consumer just wasn't around yet, so the games were limited to using pre-scaled sprites and fixed point 3D to 2D conversions. You basically got a fake horizon with sprites sprinkled around to give the impression of a 3D VR world.
 
It was frustrating to work with all these limitations, but that seems to be the main thing I enjoyed the most from my whole career, getting a limited machine to do things that others didn't think was possible. I think I excelled with the technical aspects of my career, but not with game design. We went through the whole development process, got the game done in time and on budget, even got a completion bonus, and although it was a technical achievement, it just wasn't very compelling to play in the headset. You can give me a technical challenge, like shadows for a 3D character animation on the Nintendo DS, or proportionally spaced fonts on an 8x8 pixel character machine, and I'll give it my best effort, but making a game fun to play doesn't appear to be in my wheelhouse.
 
After the VR Headset never shipped, and thus we never got any royalties from Nuclear Rush, we only managed to pick up a few short term projects and failed to land any long term original title development contracts. We had done very well financially, in my opinion, and never missed a payroll, but it was looking likely that we would have to shut down or start missing payroll. So Ken and I decided to part ways, we helped the artists find new jobs and I went off to Microsoft to be the second programmer hired for their new game development group (called the Kids and Games Business Unit or KGB, run by Melinda French, or soon to be Melinda Gates) in 1994.
 
I guess Microsoft would be the start of the second half of my career, which ended up being 5 years at Microsoft and the rest mostly at a little game studio owned by Take-Two called Cat Daddy Games (first as a contract programmer, and then as a full-time programmer).

AR: That 128k cartridge emulator board… I know that was largely Ken's thing, but could you tell us more about it? Were there plans to do anything with it outside of internal development? We have a lot of technically-minded folks who'd be very interested in any details you can recall.

KM: 128K? You can't develop a 1MB game with a 128KB cart emulator. Naw, our "Romulus" cartridge emulator had up to 4MB of capacity, or no RAM at all for an artists board (just needed the video RAM). Artists needed to be able to see their graphics on a TV screen, and old TVs back then used NTSC (or Never The Same Color) so it was critical that they could see exactly what was presented to the player. It wasn't just an issue of color representation / artifacts, the pixels aren't square on these machines, so when you draw a perfect circle on the PC, you'll end up with an oval on the TV.
 
Romulus had some great features. It was very fast, using the PC's bidirectional parallel port you could download your game in seconds, flip the write protect switch, turn off the console and pull out Romulus and carry it to another Genesis. It had a battery backup for all of that 4MB RAM and was only about twice as tall as a normal cartridge. It also supported EEPROM and SRAM backup hardware needed for some games. If I recall correctly, we exchanged a Romulus for the use of Ray Tobey's 68000 source level debugger, which really turned it into a full fledged development platform.
 
Development cycle time is really important, and having to burn cartridges for every test gets old instantly. Small game companies couldn't afford to buy a bunch of real Genesis development systems, so the developers would end up sharing a system, and maybe the art team would get one. I don't think QA testers ever got a full development system, and you really should have fast turnaround times between your testers and developers. We didn't want that kind of slow paced development cycle, we wanted fast development, so the Romulus was born.
 
We did try and sell them, we went to at least one of the Game Developer Conferences (8th I think) as a vendor with a little booth and all, and tried selling Romulus to other developers. We didn't get many buyers, probably because who can trust a tiny game development house that has yet to get something published? I think we sold a few to SOA and EA, but it wasn't enough revenue to keep us afloat.

AR: Could you describe the process of developing games with the Romulus? One of our researchers, KatKuriN, was curious what it was like to write development builds to the cartridge, and what was it like to work with on the PC side. Given how poor Genesis documentation was at the time, did you run into any problems you weren't expecting? Like the Genesis trying to access some unexpected address, or the debugger picking up false positives?
 
KM: The "host" machine would've been something like a 386DX or 486 based MS-DOS machine, this was the machine on which the developer edited, compiled and debugged the code the project. The "target" machine would be the Genesis in this case, with a Romulus plugged into it an a parallel port cable linking the two machines. You could only run your game on the Genesis itself, so build turn-around time was important. You certainly didn't want to have to go get some coffee while your game compiled, especially at the end of the development cycle when you're fixing bugs and polishing. The host machine would compile all of the code (we used mostly C and some 68k assembly) and package all of the art and audio resources into a ROM image, which would then be uploaded into the Genesis through the Romulus. For artists, I think I recall we had a way to display their Deluxe Paint Animate screen through the Genesis, which helps them see what actual colors look like on the TV and critically the objects shape, your perfectly round circles are going to look squished or elongated with the Genesis' aspect ratio.
 
It wasn't too different doing development for the Genesis as what we did at Cat Daddy for the mobile phone games, but Cat Daddy's engine ran in Windows too, so the game developers could spend most of their time in Windows with fast code/debug cycles, and the game would also be compiled an run on the same engine but a different platform. Or in the case of Android, many many different platforms, it's worse than developing for PCs (driver bugs tend to remain on an Android device, with no way for the player to update/fix them).
 
That's probably what I miss most about Genesis development, the fact that it's all bare metal from your game ROM down, no middle-ware or OS to trip you up. That graphics glitch or timing bug you're seeing? That's all your responsibility on the Genesis, no one else's code is causing it. I love having complete control over the whole system. The Genesis was the last platform I worked on that was bare metal, everything other platform I worked on afterwards had either an OS or at least a thin hardware abstraction layer, and there's nothing I dislike more than having to figure out a workaround for some obscure OS or driver bug.
 
AR: How realized were Futurescape's ambitions to expand to CD technology? Rick Lucey's Carnage (something we know little about) was reportedly planned for Sega CD but never made it out of production. Further, Ken's Genesis port of Sid & Al's Incredible Toons was once rumored to be coming to Sega CD.
 
KM: One of my last tasks at EA as a technical director was to attend the technical release of the SEGA CD in Japan and learn all about it (sorry EA, bad timing I know). So I already knew a good deal about the system before its North American release, and in my opinion it looked hacked together. Doesn't it make more sense to release a whole new console with backward compatibility to the Genesis? Same with the 32X. It's like they went through two generations of gaming systems by hacking on hardware to their older gen system.
 
I think a lot of the issues with these "add-ons", the SEGA CD, 32X, SEGA VR and even the light gun is it splinters the market. Your game doesn't just need the original console, the player has to have this additional hardware and the console. That greatly limits the size of your market, or worse if that additional hardware is never released like in the SEGA VR case.
 
CD game development is a completely different beast to cartridge development. Going from creating content for a 1MB cartridge to a 650MB CD left a lot of us wondering what we could possibly fill it with. Plus cartridge access was incredible fast, instant access in the Genesis case, and loading data from a CD took ages comparatively. A game like Myst, in my opinion, is perfect for that first generation of CD games, with huge amounts of content and level loading time isn't as critical. Futurescape was just not large enough to be able to quickly develop games with that much content.
 
AR: Monster Hunter looks like it was full of potential. I know that was Ken's project, but do you recall much from it, particularly why it was cancelled?
 
KM: I don't think the light gun sold as well as SOA wanted, so I don't think they wanted to risk manufacturing a bunch of cartridges for an accessory that didn't sell well. Plus there seemed to be a really odd dynamic between SEGA of America and SEGA of Japan. SOA was very good about marketing the Genesis, but I think they really wanted to have more of a hand in the next generation console development, but SOJ only allowed them to do accessory development. Meanwhile, SOJ was really late with the SEGA Saturn, and it seemed like SOA was filling in the gaps with accessories, trying to extend the life of the Genesis.
 
I think things would've been very different had SEGA released something like a Genesis 2 or "Genesis Pro", with a cartridge and a CD drive built-in and backwards compatibly, more RAM and a sprite scaler / 3D accelerator. The Genesis could *almost* do flat shaded 3D graphics (I wrote the polygon rendering routines for EA's "LHX Attack Chopper" for the Genesis), a little more built-in horsepower would've really extended the shelf-life of that platform.
 
We should've proved Futurescape was a solid game developer on a popular vanilla gaming platform, like just for the Genesis, but our first two development contracts tied us to hardware that either never shipped or didn't sell well. That just took all of the wind out of my sails.
 
AR: On a personal level, Lucey's Relentless is fascinating, and I'm sorry to hear EA passed on the contract. What can you recall from Relentless' development? Is that a digitized player sprite I see?
 
KM: I was also very sorry EA passed on extending the Relentless development contract. They did pay for the design development, they just didn't extend the contract to the actual development process. My concept for Relentless was to essentially be a platformer that "pushed" the player through each level, relentlessly. I was very into film noir at the time, so the premise was that you played a 50's private investigator, being chased from clue to clue. We wanted to use rotoscoping to create the rendered characters, similar to the original Prince of Persia, as a more efficient way of creating the large amounts of animations. Hand drawn animations are great, but getting fluid animated movement was incredibly time consuming to render by hand in something like Deluxe Paint/Animate. So we digitized video of me doing things like walking up stairs and attempting to run (I'm a programmer, not an athlete). The artists then took this video, cut me out of every frame and stylized me into what I hoped looked like a film noir private detective. Our way to creating graphical content quickly with only a few artists.
 
Relentless is a great example of me being too focused on the technical aspects of game development, such as how we plan on getting so much content developed with so few artists, instead of what would make the game fun. Sure, we needed to explain how we could make the game, but I really should've focused more what makes the game compelling to play. Nice smooth pretty animations are one thing, a story that draws you into the game is something else. What makes the game fun to play?
 
During my career I was fortunate enough to work alongside some truly game development greats, like Ray Tobey (Skyfox and Budokan) and Dani Bunten Berry (M.U.L.E. and The Seven Cities of Gold), and I really wanted my own legacy to include a popular or even just highly praised original game. Instead it's just littered with technical achievements; "oh, I wrote that graphics engine" or "I ported that game to another platform that it was never designed for". Working for MECC, Dynamix, EA, Microsoft and finally Take-Two, I've certainly had my hand in a lot of great high selling and praised games, I've just never designed one from scratch. My career was more like being a character actor than a movie star. There's nothing wrong with that, unless you always just wanted to be a star.

AR: Regarding that 128k thing, I seemed to have conflated Romulus with something Ken mentioned in his LinkedIn: "Designed hardware for a 128k Save ROM for Sega Genesis games." Which is also why I attributed the hardware in question to him. Sorry! I get very easily confused with anything technical. Anyway, it sounds like you did everything right with Romulus. I'd kill for a photo of your GDC booth!
 
KM: I'd love a photo of the GDC booth too! But smart camera phones were a good decade after, and I don't recall the "booth" being more than a folding table with paper fliers describing what Romulus did, and prices I believe. I wish I could find that flier!
 
AR: Do you recall what price Romulus sold for? And perhaps who at SoA and EA it was sold to? We'll be creating a Romulus page following this interview, and that'll contain a "List of developed games" section listing all the known software developed on or with the hardware. Is it safe to assume all Futurescape titles fall in that category? And do you know anything about this 128k Save ROM (RAM?) thing I confused Romulus with?
 
KM: I don't recall exactly, I think an artist Romulus was less than $1,000, and the full blown all of the RAM, backup EEPROM for programmers was around $1,750. I'll keep looking for that price sheet, it must be here somewhere.
 
AR: I 100% agree about Sega CD. I get it, it's alluring, it's sexy, it's the new thing. But like you said, it just limits the market, and then you end up with no one developing for it. Among other issues. So it sounds like Futurescape never seriously considered developing for CD… like it was discussed internally at one point, with the result of those discussions to not pursue CD development? Is that about right?
 
KM: I think it was more about the scale of the CD projects that was the problem for us. For the Atari 2600 or Apple ][, all you needed was one programmer who could do the coding, "art" and music all alone, which is fine for a 4KB ROM or a 140KB floppy disc. But by the time of the fourth generation of game consoles, you needed a couple artists per project just to be able to complete all of the art asset work (sprites, animations, backgrounds, fonts, etc.) that fit in a 1MB ROM cartridge.
 
So now look at the fifth generation consoles, all of a sudden you go from 1MB ROM to ~650MB CD and the console has a primitive 3D accelerator in it so all of those assets are expected to be rendered in 3D. 3D modeling and animation is a very different skill set than what 2D sprite/background artists trained for, so you probably can't even use the same artists you already have without retraining.
 
Another technical consideration to remember is that first generation CD-ROM drives were horribly slow. Like ~150KB/s transfer rate slow, so like 7 seconds to load your previous 1MB ROM cart. Nobody is going to want to wait through 75 minutes of loading screens, right?
 
AR: Curiously, Rick Lucey says that Carnage "was being created in conjunction with Crystal Dynamics for the Sega CD". Do you know what was going on with this arrangement? The only information we know of Carnage is from Lucey's website, anything you can remember of the project would be stellar.
 
KM: I do recall that we were in discussions with Crystal Dynamics, but I don't recall if we ever came to any real business agreement. Ken may have taken the project with Crystal Dynamics further after I left, but I don't think he did.
 
AR: And I didn't know you were Relentless' designer! Wow very cool! Sorry for attributing that to Lucey :p Again, a lot of this research was done in the dark. Anyway. You used the phrase "compelling to play". That's the best way to describe the core fun of it all, I think. Those are good words. I'm legit loving all your postmortem analysis here, a lot of this is like super relevant to gaming today more than ever.
 
What other games did you design at Futurescape? (Other than Nuclear Rush, for which we have in-game credits with a designer tag for you.) Actually, if you can remember anyone else on any other game, no matter the role, just rap it off here and we'll make sure they get a credit.
 
KM: The only game designs I worked on at Futurescape was for "Nuclear Rush" and "Relentless". My main role was to be lead programmer on Nuclear Rush, but also business / employee management / scheduling / hiring / accounting / leasing office space and all of the other jobs trying to run a small business gives you. Ken and I split the project lead work for the two games between us, and we shared business tasks. I believe at our peak we had eight or nine salaried employees (including Ken and I), but this list is from memory which might not be all inclusive:
 
Kevin McGrath - Lead Programming (Nuclear Rush) / Management
Ken Hurley - Lead Programming (Monster Hunter) / Management
Tom Collie - Lead Artist (Nuclear Rush) / Animator
Jeff Sturgeon - Lead Artist (Monster Hunter) / Animator
Tom Debry - Programming
Matt Hubbard - Programming
Rick Lucey - Art / Animation
Tara Packard - Art / Animation
Bill ? - Business Management
 
Sorry Bill, I've completely blanked on his last name. We also had external contractors for things like music and sound effects development.
 
AR: Following Futurescape, you worked at Microsoft for the latter half of the 90s, and then at Cat Daddy Games. I know this is a large period of time to summarize, plus you covered this period quite succinctly in your 2005 Sega-16 interview. But we did touch on your background as a segue into Futurescape, and I feel a segue out of Futurescape would be appropriate. In your own words, can you describe leaving Futurescape and your following game career to your retirement?
 
KM: Leaving Futurescape was very difficult, as you can imagine. No one wants to give up on their dreams, and for a couple years it seemed like we could grow into a successful game development studio. The plan was to grow with royalty revenues, but we never got any because neither Nuclear Rush or Monster Hunter were released. The projects we tried to line up after it became obvious that neither game would get royalties, didn't provide enough income to cover salaries for the long term. We did have some contracts in the works, but they were mostly for porting existing games or for technical help, nothing that resulted in longer-term contracts for a team of programmers and artists. The future of Futurescape did not look very bright, and I didn't want to cut staff to just Ken and I and continue to try getting larger development contracts while we did small piecemeal contracts to keep afloat. So we helped everyone find new employment and closed shop.
 
Microsoft was just forming its internal game development group, and I was lucky enough to be the second programmer they hired for the group (called the Kids & Games Business Unit). I was hired to get up to speed on the SEGA Saturn, which was not yet released, but we never got a development system for it! Other groups at Microsoft must've had one, but for some reason the Games group couldn't get our hands on one. So the first game I worked on at Microsoft was meant as a technological and group demo, to show how you could actually play real-time full-screen games in Windows and that we could develop games in-house. They gave us an impossibly tight time-line for that project, but "Hover!" got made and included on the Windows 95 CD. It even ran on the lowest end machines that barely ran Windows 95. I wrote the ray-casting rendering engine (like what Wolfenstein 3D used) and level editor, but we also had three other programmers on the project, John Seghers, Patrick Wilkinson and Larry Shatos. Pat, also who I shared my first office with at Microsoft, later started Cat Daddy Games with Harley Howe (who was also an artist on "Hover!"), even after listening to my endless warnings about start-ups and going out on your own. Good thing they ignored me, since Pat convinced me to come do some development contracts with Cat Daddy Games after I left Microsoft, and I spent the last 15+ years of my career working for Cat Daddy Games / Take-Two.
 
At Cat Daddy, I probably worked on twice as many games as I had worked on during the previous 15 years. From getting their Windows based 3D game engine to work on the PS2 for the game "Food Fight" (which thankfully got canceled), to several shipped Nintendo DS titles and finally a pile of Android mobile games.
 
AR: Looking back at your time with Futurescape, what stands out the most? What are you most proud of? Do you use any of the insight learned through your experience with Futurescape in your current work? And what's your biggest postmortem takeaway from it all?
 
KM: I'm most proud of not missing a single payroll. How many Silicon Valley startups can say that? We gave it our best effort, and I'd like to think we would've still been around today had both games sold well. That was the risk we took though, and we just didn't have enough capital income to continue at the same scale we had been running on for two years.
 
AR: If you could say one thing to the gaming world as a whole, to the programmers and developers out there, the aspiring and up-and-coming, all those people who want to become game makers… what would it be?
 
KM: A common question in game development interviews is "why do you want to work on games?" or "what excites you about working on games?" You should know the answer of that question for yourself before any interviews. The red-flag answer to that question is something like "I really enjoy playing games". My personal answer to that question is that I enjoy the various programming challenges presented while making games, it touches on graphics, UI, networking, AI, simulation, audio, etc. No other programming role gives you as wide of access to every kind of programming challenge as game development does.
 
As far as coding interviews go, it's vital that you explicitly say what you're thinking, what steps you want to take to solve the problem. The interviewer is trying to figure out if you can break a problem down and solve it, and if you just stare at the whiteboard and not share what you're thinking, the interviewer isn't going to know if you can solve the problem. Even just asking questions shows off your analytical thought process better than the deer in the headlights look. I would much rather like seeing pseudo-code of how you might solve the problem with nested loops or recursion, than a couple of lines of syntactically perfect code that doesn't solve the problem.

Please let me know if you have any more questions.  I'm very slow to respond (sorry about that), but it's a lot of fun reflecting on past projects.