I spent a lot of time in high school working in their computer lab, and I had met a lot of older students who were in there. One of them graduated a couple of years before I did and went to work for a small computer game developer name Pegasus Software that made computer games that could be played over a cable TV network, using the cable box as a console and the remote as a controller. It was really ahead of its time for 1981.

The company was having trouble creating art assets for the game. There were specific rules about color and layout that the hardware required, and it was hard to find artists who would understand and follow these rules and still make good art. They wanted to make a tool to help the artists, something like Paint that would restrict the artists to following the rules automatically so they could just concentrate on the art. Neither the PC nor the Apple II could display the resolution or color space needed by this tool, but the Atari 800 could, and I had [been] programming that hardware for a couple of years when my friend who worked there called me and asked me to apply for the contracting position to make the tool. I interviewed and the company liked me, but I was only 15 years old so they couldn't hire me. I was surprised when, a few months later when I had my birthday, they called me and asked if I was still interested. In that time, they had not found another Atari graphics programmer, so I said yes and found myself employed in the game industry.

I worked at that company for the next five years. I graduated from high school and went to college, but I would come back and work there during the summers. They changed their name to Cybron when another company in Virginia named Pegasus claimed earlier ownership of the name, and they landed a deal with Electronic Arts to make Grand Slam Bridge. I worked on the bidding, playing and deck maintenance code for that game, but I graduated from college and moved to California for graduate school before it shipped. I found out years later that they didn't put my name in the credits. While this disappointed me, I hadn't really planned to go into the game industry as a career. I was working on my Ph.D. in Computer Science, and I planned to become a professor and work in academia.

Well, that didn't happen. After I got my Master's Degree, I was disillusioned with academic life, so I sent my resume out to several game companies, including Interplay. They responded the fastest and they were the closest, and between my previous experience, my programming skills, and my pencil and paper RPG knowledge, I landed a contractor position making Bard's Tale Construction Set.​

I played a lot of pencil and paper RPG's, but D&D was my biggest influence. I had played it since I was fourteen, both as the DM and a player, and I loved the open-ended quality it had. My friends and I could make up any fantasy adventure story and make it work in that game, and that really set it apart from the board games and card games that I had grown up with.

GURPS was another huge influence on me. With its generic system that allowed any setting to be used, I could finally make up any adventure story and it would work. Science fiction stories set in space, super hero adventures, a psionic system that was balanced... GURPS opened my eyes to the world of adventure beyond the realm of fantasy. And I liked the idea of learning one system as a the core, and then you simply added new skills and advantages based on whatever genre you were playing. I thought it was brilliant.

There were other RPG's that influenced me too. TORG introduced a cinematic feeling to play, by allowing re-rolls that added to the first roll, letting the players perform crazy, over-the-top actions which really made them feel like heroes. Amber taught me that dice were not needed to play an RPG, and Gamma World presented the first post-apocalyptic setting that I had seen.​

When I started at Interplay in 1991 as a contractor, there were less than 40 people there. In fact, when I got hired the next year, I was employee number 42. That small number meant we all fit in one building, everyone knew everyone else, and you knew about every product being made. The company felt almost like a family.

When I left Interplay in 1998, the company had over 600 employees, spread over multiple buildings and several countries. Games would ship that you never even knew were in development, and licenses were purchased and developed in secret. Lots of development was outsourced to outside development studios, so Interplay felt more like a factory than a family.

I left when I felt like I had lost control of Fallout. So many people took over after Fallout 1, when they changed their minds about the IP and suddenly it was a AAA game. Before Fallout 1 shipped, we had been overshadowed by D&D and Star Trek, so we were left alone to develop it our way. After it shipped, people were suddenly dictating content, box art, advertisement assets and everything else, so I walked away from it. I was proud of the game and happy that people were so passionate about it, but I realized it wasn't mine anymore and never would be.​

Rule-based AI runs into a lot of the problems that “expert systems” run into. They are very brittle systems, with a set of unchanging rules that may not cover every case. So you end up using default rules, which don't often lead to very intelligent actions. Fallout's AI was very much rule-based and brittle, and if I were to make it again, I would use fuzzy logic with a variety of parameters, and some of those parameters would be available to the player, so he could tell his followers to act more defensively or to wait for the player before reacting to potential targets.​

I don't remember the specific details of my plans for Fallout 2, but I do remember playing the game and seeing it was different from the storyline I had proposed for it. I think my biggest disappointment with the game is that each area was made in almost complete isolation from the others. There was no over-arching theme and no attempt to make sure the different areas were cohesive. It felt like a lot of Fallout-y areas, placed adjacently and connected with a storyline. Those areas were individually well-done, but they suffered from the lack of a strong central design.​

I did enjoy both Fallout 3 and New Vegas. I know that surprised some of my fans, who wanted me to hate the games and rail against their design choices (which I have repeatedly pointed out were different than the ones I would have made), but there is no arguing that more people enjoy the modern versions of the franchise than the older ones.

If I were to compare the two games, I would say that Fallout New Vegas felt like it captured the humor and style of the Fallout universe better than Fallout 3, but I have to hand it to the FO3 designers for developing VATS, a cool twist on called shots for a real-time game. I also loved the set decoration FO3. There was so much destruction, yet obviously everything had been meticulously hand-placed. So much story was told entirely through art. I ended up naming these little art vignettes and creating side stories in my head about what had happened. There was "The Suicide", a dead guy in a bathtub with a shotgun, and I figured he just couldn't handle life after the bombs. There was "Eternal Love", a couple of skeletons in a bed in a hotel room, forever embracing each other. There was "My Last Mistake", the corpse in the temporary one-man fallout shelter which obviously didn't do its job of keeping out the heat and radiation. My favorite was "Desperate Gamble", where I found a feral ghoul in an underground shelter filled with lab supplies and lots of drugs... except for Rad-X. I imagined that a scientist found himself irradiated and desperately tried to synthesize some Rad-X to cure himself before he succumbed, but he was too slow. I did notice that whatever was left of his mind sure did seem to enjoy toilet plungers.

If I had to pick something I didn't like about FO3, I would pick its ending. I hated the ending. There, I said it. I didn't like the sudden problem with the purifier, and I especially didn't like the lack of real, meaningful multiple endings beyond what I chose in the final few minutes (FEV or not, me or Lyons, and that was it?). But the worst thing about the ending was there was no mention of the fate of places I had visited. In my head I had already imagined slides for Megaton, the Citadel, Rivet City, Underworld, GNR, the Enclave or the mysterious Commonwealth. But I got... pretty much nothing.

I liked FONV's ending much better. It had a nice set of slides at the end of the game. They covered everything I was wondering about. I went with Mr. House at the end... and that seemed a worse choice after the slides, but still OK. It led to a law-abiding but somewhat impersonal Vegas. I wish I didn't have to kill the BoS, but I want House to control the future, so I had to do it. It was a great morally ambiguous choice, and the decision made me pause. That's a sign of good design, right there.​

I don't think criticizing Troika games for being buggy was unfair. They were buggy, and I think there were two big reason why that was so. First, we tried putting a lot of features into these games. We really needed to learn how to edit, because we would spend a lot of man-hours putting a feature into a game that hardly any of the players would ultimately care about. For example, Arcanum had newspapers that reported on major incidents that were caused by the player, but I don't remember a single review mentioning that. We spent a lot of time getting that working, and those hours could have spent balancing real-time combat, or fixing the multiplayer code.

Second, we kept our team sizes small, both for budget and for management purposes. This meant we had less total man-hours to work with, and all of the late nights and weekends couldn't make up for the fact that we only had about a dozen people working on the Arcanum and Temple projects. Looking back, I am amazed our games were as feature-rich as they were, but I am not surprised they were as buggy as they were. We should have made some serious feature cuts early in their development.

Troika got characterized as “always blaming the publisher” when something was wrong and I think this was unfair. We would always own up to the parts of the development process in which we had made mistakes, but it seemed that if we ever said “we messed up this, and our publisher messed up that”, some people just heard the latter part of the comment and would start screaming “Troika is blaming the publishers again!”. It got frustrating after a while, especially when I saw people at Troika quoted out of context. But I did gain quite an insight into the American political system, which seems to deal with the same kind of illogical, sound bite oriented system of criticism of its political candidates. People hear what they want to hear, and often make up their minds before seeing, or even in spite of, any evidence to the contrary.​

In saying “being original is risky”, I think Leonard was referring to the fact that publishers looked for proven hits, either in big mainstream titles or in sequels to popular products. I don't think Troika closed because we were trying to be too original. We closed because we were not getting contract offers for products we wanted to make, so we voluntarily shut down while we were still in the black, financially. We could lay off employees with severance packages and extend their insurance for a few months, rather than just shut down with no notice and kick everyone out.

So then the question is, why didn't we get any offers we liked? Especially since all of our games turned a profit? From the publishers point of view, our previous games had sold to a niche audience, so they were unwilling to fund us to make either a new IP or a large mainstream game based on a licensed IP because their numbers showed that the profit was too small. In other words, they could spend that money at other developers where their rate of return was much higher. It makes total sense from a business point of view, but it's still sad for Troika and its employees.​

Leonard, Jason, and I had made character systems before, but in the case of Arcanum, we wanted the system to reflect the setting. This meant that not only would there need to be magic spells and technological skills of some kind, but there would need to be a mechanic that measured the character's aptitude with each. We wanted the struggle between the magic and tech that was taking place in the world to exist within every character as well. If you are good at magic, you are not as good at tech, and vice versa. And this naturally meant that a jack-of-all-trades character could be good at both, but not great at either.

Using that as our basis, we developed our stats and skills and spells and tech disciplines. It's hard to pin down our influences, since between us, we had played a lot of games. But I think it's safe to say that GURPS was a big influence, with its ideas of no classes, advancement thru skill points, and the concept of advantages and disadvantages (which we combined into backgrounds). Other games influenced us too, and I think we picked and chose elements that we liked, such as spell categories, crafting from schematics, and weapon and armor types from several different games. It was very challenging to balance all of these characteristics, and I am sure you will find no shortage of individuals who will claim that we did not.​

We had felt constrained by Fallout's world map, with its relatively small coverage of North America and the inability to place handmade content without also placing a location on the map. We wanted to make a large, open world for the player to use as a playground. The term “sand box game” wasn't in use then, but that is a good term for what we were trying to make. We want a huge epic storyline, and we felt that we needed a lot of space in which to tell it. And of course, we would need that space later for the many sequels we had planned.​

It's been so long that I honestly do not remember. I know that Leonard, Jason and I had worked out the main story arc before hiring anyone to the project, but many of the sites at the end of the arc, especially in Vendigroth and the Void, were just vague ideas. Other people with better dialog and map design skills than me did the actual layout of those quests.

I did suggest the idea of master quests for skills, and I designed most of the ones that went into the game. I also made a list of one hundred random “places of interest” that people could add to the game when they had free time. If anyone did add one, they simply crossed it off the list so no one else would add the same place twice. Many of the non-storyline locations in the world came from that list.​

The sequel was based loosely on Jules Verne's A Journey to the Center of the Earth, where we planned to continue the adventures of the great explorer Franklin Payne. He has disappeared into the bowels of the earth, and his wife has hired you to find him. We had laid out most the storyline, and it included finding prehistoric monsters, subterranean humanoids, and most thrilling of all, a clue about how magic and tech can be reconciled in the same artifact, something that most learned people had believed to be impossible. Of course, none of this came to be, but our talks about using the Source engine led to our making Vampire: Bloodlines.​

We were trying to recreate the original Temple module, something that had required a group of players and a DM, in a computer game using the 3.5 rules. In hindsight, yes, I think I stayed too close to the original module, which contained a lot of nonsensical plot lines, characters and items. While we were making the game, I spent two hours each week playing tabletop D&D with the development crew, so they would see what the tabletop experience was like. I used a module of my own creation, and many of them liked it better than the Temple module. Perhaps it would have made better source material.

One event that I have rarely talked about is that during the 20 month development of Temple, I started feeling very sick, and it turned out I had a kidney stone that was failing to pass. I spent four months on pain killers while the stone worked its way out, and they certainly affected my work output. I am not putting it forth as an excuse for the problems with the game, but as more of an explanation that while Temple's schedule was very aggressive. I was not up to the task. I should have asked for more time to complete the game.​

My goal was to recreate the Temple of Elemental Evil module in a 3.5 game engine, and at that, I think we succeeded. But like I said, I wish we had created our own source material. The engine was so good, and I think we re-created the 3.5 rules and tabletop experience very faithfully. That was not the problem with that game. The storyline, characters and dialog that I wrote were simply not up to the level that Troika had set with Arcanum.​

Hommlet was intended to be a place to rest, buy and sell, and pick up followers. For players that did not want to rush to the moat house, Hommlet was also intended to provide enough side quests to grant a couple levels of experience to a new group. My biggest complaint with Hommlet is that I wish we had made it smaller. It had some interesting sub-stories going on, but it could have been a quarter of its size.​

Yes, we had great plans for that engine. For the sequel to The Temple of Elemental Evil, Troika proposed using the super-module GDQ: Queen of the Spiders, which consists of seven modules from the popular Giants and Drow series, plus the special Q-series module that completed the adventure. In fact, we were going to let the players bring their characters over from ToEE directly into the QoS, so they could simply continue playing with the same group of characters. Alternatively, we had suggested using the engine to create the long-awaited Baldur's Gate 3, and Obsidian had also expressed interest in licensing the engine to make D&D licensed games. Unfortunately, Atari never followed up on any of these proposals.​

I still enjoy MMO's, but it's obvious the entire genre is shifting to free-to-play. There are so many enjoyable MMO's that cost nothing to download and play, and that only require you to pay if you want to buy cosmetic items or speed along some process or gain access to restricted content, that I feel that the subscription model is dead. Of course, there are always exceptions, and World of Warcraft still requires a monthly fee, but even that game allows you to play up to level 20 for free.​

I am not in a designing role on the South Park RPG, and as a programmer, its casual nature doesn't influence me in any way. And as a Mature-rated RPG, it feels more like my old Troika games than a Teen-rated MMO.

As for casual RPG's, if I were to make one, I would lean towards making it simpler, both in character systems and in the UI, so that players who are not interested in stat juggling and character design could enjoy the game too. I'd like to think that someone who played and liked a casual RPG would use it as a stepping stone to richer and more complex RPG's, but for many people, you have to make that leap as easy as possible.​

The full quote was “I think if you’re not trying to make a work of art that is fun, and accessible, I don’t think you should try. You should worry less about the art, and more about how enjoyable it is." I was speaking directly to game developers who care more about form over function, who would rather have their game look good than be fun to play. These are the people who would remove a compelling game feature because they could not make the art look perfect. That's so opposite to my way of thinking that it took me a long time to even realize there were developers like that.

I care more about a game being fun than being beautiful, because no matter how good you look, people will move on to the next pretty thing and forget about you. If you make a fun game, people will remember that. And a fun game needs to be accessible, by which I mean that game had to present its rules clearly and then follow them. Don't give me a gun and then force me into dialogs. Don't let me stealth and then put in a boss who can see hiding creatures 100% of the time. Don't give me the power of flight and then force me into an underground dungeon. To me, a fun and accessible game teaches the player how to play it, lets them choose their own way through it, and then reward them for it.

In that sense, Temple and South Park are very similar. They both have classes with clearly defined abilities, a variety of weapons and armor and consumables, and a world that lets the player choose how to go through it. South Park has a more linear story line than Temple, but it's trying to tell a stronger story, with a definite start, middle and end, just like the TV show. Temple had less of a story and more of a goal, to deal with the evil inside this dungeon, and you could do it in a variety of ways, including joining the side of evil. That was more in keeping with its nature as a D&D module.​

I am fine with adding political or social statements into my games, especially if such statements reinforce the game's setting. Fallout was a post-nuclear world, a product of governments lying to each other and their constituency, so of course political comments are going to be made by the characters. And in its lawless environment, people are going to act more towards their baser natures, so we have an opportunity to explore that. As the creators of the setting, you could almost say we have a duty to fully explore the nature of the setting, so ignoring those issues would make the characters seem less real and the setting more contrived.

Arcanum was interesting to us as a setting because it took a standard medieval world and added an industrial revolution. That event would certainly have effects on the different races and governments, so the exploration of the social and political upheaval in that setting again felt like something we had to do. Otherwise the world would have felt static and not dynamic, even though we were telling the player that the world of Arcanum was in the midst of drastic change. We had to show that, not just say it.​

I like both models, and it really depends on the system. For a game with classes, I like making parties because that's the only way you can get the abilities and skills that you want. For example, if I want strong healing and stealth, I would probably have to take two different classes, and barring multi-classing, that requires two different characters. But with a skill-based game, I could make one character that had healing and stealth, and then I wouldn't need any other characters.

There are other considerations too. In turn-based games, character parties with full player control allow for more tactical combat, while single-character games allows for more personalized dialog and a stronger storyline.​

Probably the most common mistake is misunderstanding the nature of a video game and how it differs in fundamental ways from tabletop games, movies and books.. For example, for the longest time, CRPG's that were based on tabletop games kept the dice-rolling mechanic of character creation. While that works great for a tabletop game where your friends are making characters together, dice rolling is not a good way to make characters for a computer game. Since nothing stops the players from re-rolling (even if it means stopping the game and restarting it), they would usually start the game with over-powered characters.

Similarly for games based on movies and books, developers often try too hard to make the player perform certain actions or play a certain way, because that's the way the characters in the movie or book would behave. I've played games that are mostly cut scenes interspersed with some activity, and I always thought to myself, if I wanted to watch a movie, I would have watched movie. Games are interactive at their core, in a way that books and movies are not.

The best question a developer can ask when making a new system or adapting an existing one is this: what is feeling I want this game to evoke. Should the player feel tense, or scared, or excited, or heroic? And then tailor your game's setting and mechanics to evoke those feelings. Harry Potter does not spend most of his time in the movies in wizard battles, so I wouldn't spend 90% of my budget on a Harry Potter game on combat assets. Instead, I would make an interactive world containing clever dialogs and puzzles that the player could explore to advance a storyline, and periodically there would be a wizard battle or a game of Quidditch to emphasize an important plot point.​

It's an exciting time for cRPG's. There has been a rise of independent cRPG development, which until recently had been pretty scarce (with Spiderweb Software being a notable exception). I think there are more single player cRPG's again, after we passed through a decade or so of multi-player emphasis. I am a little worried that cRPG's are getting too action-oriented, with the player's skill trumping that of his character's. But as long as that is just one kind of cRPG, I don't mind if people like to play those kinds of games. I just want good, single player, character skill based role-playing games to be released too.

Also, I'm really happy to see retro gaming revive interest in older games. There are some great older games out there, and it's nice to see younger players discover some of the old classics.​

Small 2 and 3 man teams may be able to produce a few PC and console games, but mostly they are making smaller games that have much less complexity or player time investment than full-sized games, and those latter games still need a team to develop them. I am glad to see crowdfunding add an alternative to the publisher model for many developers, and digital distribution creates sales channels for smaller companies that can rival the older physical distribution of large publishers. In short, I think variety and options are good things, in the game industry as well as in games.

I am concerned about the mid-tier developer being crowded out of the market by these new methods. It seems that we are increasingly seeing two types of games, ones made by small independent developers and ones made by huge, publisher-owned teams. The mid-tier developer, which have teams of 30-60 people, are shrinking, and small teams of less than 10 people and large teams of over 100 people are becoming the norm. I am worried what this means for the types of games that will be available over the next few years. Will they be either small casual games that you play for a few hours and then move on, or gigantic behemoths that you devote months of gaming time to, possibly investing in DLC to stretch the gap between sequels? It's as if books are disappearing, to be replaced with short story collections and lengthy book series, or movies are being replaced with TV shows and movie franchises. Is there no middle ground any more? I don't know, and that worries me because some of the best games have come from such development, and it would be a shame if it was lost.​