Exclusive Interview with Chris Park about AI War

I thought I'd pose a few questions to Indie Developer Chris Park regarding his recently released co-operative 2d space RTS AI War: Fleet Command.

UKGamer: What was your original inspiration to make AI War a co-operative game against an asymmetrical AI, rather than pursue a more traditional single player campaign with player versus player multiplayer providing much of the challenge?

Chris Park: Well, to be perfectly honest, my motivations were quite selfish. The short answer is that its simply the sort of game I want to play. The full story is that, since around 1998 or so, my dad, my uncle, and myself have had a weekly RTS session together -- and in recent years, one of my uncle's colleagues has also become a regular. At any rate, we started out with the original Age of Empires (I had already exhausted Warcraft I and II for myself), and then moved on through the AoE series, the Empire Earth series, Empires Dawn of the Modern World, Rise of Nations and its offshoots, Supreme Commander, and also a few outside the genre like CivIV and Neverwinter Nights, both of which were much more shortlived.

For me, that was the most fun I had in RTS gaming, was playing with that group, co-operatively against the AI in skirmishes in those games. There were a lot of other RTS games that I played and enjoyed, such as Star Wars: Empire at War and a few entries from the Total War series, but they didn't support the sort of multiplayer I was interested in and so they got a lot less play.

The problem, even with the other games that we did play, was ultimately the AI. In the earlier entries such as AoE, AoE2, and Empire Earth, we were all less skilled and so we got a lot more use out of those games. In later games such as AoE3 and Supreme Commander, we reached the point of competence much more quickly -- once you're competent at most RTS games, the AI doesn't hold a candle unless it really outnumbers you, which isn't possible when you are trying to play 4 humans against the AI players.

When we were winding down on Supreme Commander, in the middle of 2008, I was looking around at the other possible candidates to become our next RTS mainstay, and I was really frustrated by the fact that nothing matched exactly what I wanted. I had been programming and designing games as a hobby off and on for years, but it never occurred to me to attempt an RTS until then. When it came time to decide on modes of play -- campaigns, skirmish etc -- I chose to focus only on the modes of play that I actually personally play, since those are the modes that I have enough experience with to do a good job with the design. Any scripted campaign I was likely to create, or pvp offering I might try to put out there, is likely to be sub-par for the genre simply because I don't really play those modes enough to know what's already been done, what works, what's been found to be annoying in the past, etc.


UKGamer: Where did you derive the idea of having an escalating AI Progress mechanism (incorporating AI aggressiveness and attack consequences) into the game?

Chris Park: This basically arose out of my desire to have the game be able to support long seiges. The most fun I ever had with an RTS prior to AI War was some 12-hour marathon player-vs-AI skirmishes in Empire Earth. Playing on team islands map types, their AI didn't know how to effectively use transports and so they would get this huge mass of guys on their territory, while they were completely unable to attack us except with nuclear bombers and the like. This made for a very interesting situation as we tried to take the smaller islands, get a foothold on the bigger islands, and ultimately win. We did win, eventually, but it took many hours.

I wanted to have the game design of AI War support that sort of extended seige (with all the feelings of satisfaction that come after winning that), and I think you can see that throughout the design of the game. The AI Progress gives the game a bit of the feel of a turn-based game, even though it is entirely realtime, and I think that not only allows for the seiges (it's original purpose), but also allows for a lot more strategic thinking. The turn-based level of strategy that the AI Progress provides was also a really motivating factor for its inclusion -- I was really impressed by the level of strategy in CivIV (and quite addicted to it in single player), but multiplayer with my group usually devolved into a lot of waiting around for the next turn if we weren't all attacking at once. So I basically tried to duplicate that level of strategy in a realtime context where there isn't any waiting.


UKGamer: What do you think is the biggest selling point or hook of AI War, that would act as the magnetic pull for RTS and strategy gamers?
Chris Park: For me, it's right in the name: the AI. If you're playing solo or cooperatively, you either need great AI or really well-designed scripted campaigns. So far we've seen a lot of the latter in the genre, but almost never in a cooperative context (very recent games like Red Alert 3 and Dawn of War II being the exceptions to that general rule). A lot of people have been really excited by the high unit counts, too, even though at this stage I am so used to them that I hardly remember it's a unique point any longer (that's always a side effect of spending so much time with any one product, and I've heard of it happening to a lot of other software developers).

But mainly, I guess, it all boils down to the fact that the game is actually fun, and a bit different from your standard fare while still feeling reasonably familiar in its overall control scheme. With any RTS game that I've really enjoyed, it simply "felt right," and I'm really happy that AI War is finding a growing pool of players for which that holds true here.


UKGamer: How did you come up with the variety of AI types that are currently in the game?

Chris Park: Some of them, like the turtle and the raider, are simply genre standbys. Others, like the Attritioner or the Special Forces Captain, are really built around highlighting a specific game mechanic or new ship type. Generally I wanted each AI type to feel different enough so that players could recognize over time which one they were playing against without being told. That, and I wanted to maximize the number of possible scenarios people might see; the effect of adding more AI types is multiplicative in that regard.


UKGamer: What's the most outlandish idea for a new Ship design you've ever had, whether it be your own idea or a member of the Arcen Games community?

Chris Park: Hmm, that's a tough one. A lot of the designs that made it into the game are pretty esoteric by RTS standards. Electric shuttles, slow-moving ships that emit giant bolts of lightning that lightly damage all nearby ships jump immediately to mind. So do vampires, which steal health from the enemy ships they attack. The ship that most people have commented on as being outlandish is the Astro Train, however. That ship is held only by enemies, and is completely indestructible (an odd choice for a wargame), but it's not primarily a combat ship. Mainly it just causes disruptions-over time in your defenses, most notably your turret emplacements and minefields, but it can also cause problems for your scouts.

The players have a variety of ways to deal with the Astro Trains, most notably some counter-turrets that protect against the train attacks, and the ability to destroy the train stations, thus causing trains to re-route. These sorts of decisions are not something you really see in most RTS games, where the usual response to a troublesome enemy is just to kill it. When that is no longer an option, and you have to employ some other methods, you get some very interesting and unique gameplay scenarios. There are a variety of other ships that fall under that category in the game, but Astro Trains are the most commonly seen.

UKGamer: Are there any plans for truly massive starships with their own communities of support ships, fighters, bombers and transports?

Chris Park: Yes, very much so! Players have asked for some ships specifically along these lines, and I already had some plans for this, too. The main limitation that caused these not to be in the original version of the game was art. All of the art from the game was either from free sources (mostly Daniel Cook's work), or stuff that I cobbled together myself. I'm a competent artist, but not a great one, and large ships like that are beyond me. Now that the game is out and is actually selling, I'm interviewing artists to contract for some general graphics upgrades, as well as new ships that I simply couldn't do before because of a lack of art. Most of the really massive new ships will probably be in expansions, but there will be a couple in free DLC.


UKGamer: Any chance of seeing colonization of planets, planet resource management and planet defense, come into the game or future incarnations?

Chris Park: My original design for this game actually had planned for a planet-level view (with ground/air/sea combat) in addition to the space-based level. However, through further testing and design, I eventually decided that was simply out of scope and would not only take too much work to develop and test (and largely be redundant with the existing space combat), but I was also worried that would just be too much for players to manage. I've always liked that idea, though, so perhaps in a later expansion a limited form of that sort of thing might make a resurgence. I doubt that will be in the first expansion, though, since its focus is on bringing even more scope and variety to the space-based aspects of the game.


UKGamer: We have seen that you are committed to providing regular free downloadable content (DLC), often incorporating many ideas from the Arcen Games community. Do you think you'll be able to sustain such a regular stream of new content for very long, and how are you deciding which new features to bundle in DLC and which ones to hold back for future expansions to the game?

Chris Park: I think I'll be able to maintain a fairly regular stream of free DLC for as long as there is player interest in the game. Some of the interface additions and other gameplay shifts can be quite time consuming to implement, but usually the additions to the AI logic or the additions of new ships are actually quick and easy by comparison (when I have the needed artwork for them). This game was designed to be modular and easily extensible (from within -- not by outside modders). I still play this game every week with my RTS/alpha group, and we've been basically looking for a game that we could stop and really sink our teeth into for longer than just a year or so (our average time with most other RTS titles). So far we've been playing it for around 8 months, and no one is really getting tired of it yet, so that's a good sign. I'll keep adding to it just for us, at the very least.

However, that said, I think that the nature of the free DLC is going to shift over time. So far there have been a lot of features requested that basically made the game more playable for people coming from other specific RTS titles with conventions different from those I'm used to, or for people with different play styles or preferences in general. A lot of those really come down to polish and usability. I feel like the product we delivered at launch was quite polished and extremely usable, but at the same time it has made some pretty amazing leaps forward since that time. Take that as you will, I guess. Having a dedicated fanbase giving constant, detailed, quality feedback is a real asset for any game. One reviewer called an earlier version of the interface "almost completely fantastic," so hopefully we're getting closer to "completely fantastic" with the free DLC.

Those sort of changes presumably won't keep coming in forever, though -- already I've seen a noticable drop in those sort of "I could play this a lot better if only" requests. A lot of the requests now are more "nice to have" features, or game-expanding features. The latter is where I see the focus ultimately staying in the long term, and as I've said that sort of thing is comparably quicker for me to implement. My original plan with the game was to do weekly DLC for a month after release, and then switch to monthly, but the weekly DLC has been so valuable and such a hit that it's two months since release and I'm still doing weekly releases with no immediate plans to cut back (vacation/holiday weeks excepted). A year from now, I expect the weekly (or possibly bi-weekly) free DLC releases will have quite short release notes compared to now, with just a key new ship or feature or two added in.

As far as how I decide what to hold back for expansions instead of free DLC, that's fairly straightforward: if it affects usability, the interface, playability, or AI quality, it's free DLC. If it's a large new game mechanic or ship that is going to take a lot of my time to implement, then that's most likely for an expansion. Those sorts of features need a longer testing cycle than one week, anyway. Most of the content for the free DLC releases is focused on augmenting the existing game, or offering minor extensions and new options to it, rather than huge game-changing new content. But, that said, sometimes just making significant improvements or shifts to the AI logic, as has happened a few times in the past, can make a huge shift in the game even just from the free DLC.


UKGamer: AI War is currently available on Stardock's Impulse digital distribution client, are you looking to pursue some of the other digital distributors (such as Steam, Gamersgate, Metaboli, Direct2Drive etc)?

Chris Park: Yes, definitely. We're one of the more popular games on Impulse at the moment, and I couldn't be happier with the folks at Stardock, but to really make it long-term Arcen Games needs to have more distribution partners. It's never good to have all your eggs in one basket, no matter how awesome that one basket is -- I've learned that the hard way in past business ventures (unrelated to gaming). We're currently under review at five other distribution channels, but these things just take time. One of the great things about Stardock is how responsive and quick they are, but our experience with them was even quicker than average simply because a staff member saw our inquiry, tried the demo and loved it, and accelerated the entire submission process. So the combination of those two factors has put us out on Impulse far before most of the other platforms even responded to our initial inquiry emails.

I'm very hopeful about getting the game on the other platforms, and given our success at Impulse hopefully that is something of a no-brainer for them. But it's always a challenge for indie titles from little-known developers, so we'll see. A large part of Arcen's future rests on whether AI War really finds a self-sustaining audience or not, and getting the word out through print/web publications, as well as every distribution channel that will take us, is the only way for the entirety of that audience to find the game. There are a few distribution channels out there that have rather unfair terms or royalty rates, and we're avoiding those, but anyone who is on the level is someone we are interested in partnering with.

UKGamer: Can you ever see any future iteration of AI War supporting a 3d graphics engine and perhaps incorporating the 3rd dimension of space into the gameplay somehow?

Chris Park: Actually, probably not. The game wasn't built in 2D as a poor-man's substitute for 3D. I played Homeworld back in the day, but for a lot of reasons just never connected with it. Obviously there are millions of players who did, but it just wasn't my style. When it comes to 3D space games I am more a fan of action-oriented games like the Rogue Squadron or Descent: Freespace games. But, more than that, pretty much every Arcen Games title is going to be in 2D, at least for the forseeable future. Part of me would really like to make a 3D first-person shooter game at some point, but in general I feel like there are more quality 3D games than 2D games coming out these days, and I want to help fill in the gap with quality 2D. I'm a sucker for good pixel art, and I'd really like to develop the sort of games that might have been common in a parallel universe where consumer-level 3D never became common. There are lots of other companies out there that are better than me at making 3D games, and in general I prefer to just play their offerings while making my own games to fill the 2D void.


UKGamer: Have you writing aspirations to include a backstory and AI War Lore in either expansions or future versions of the game?

Chris Park: For me, RTS games aren't too much about the story. I enjoyed the story very much in Warcraft II and its expansions, but I can't think of another RTS title where the story mattered much to me. This is a funny thing for a guy who has written several novels to say -- I still hope to be a published novelist some day, even though my main focus is going to remain software development. But, plain and simple, I just don't feel like every game is made better by having a story. Chess doesn't have a story. Neither does the skirish mode of any other RTS game, which is what I mostly play. I have aspirations to do some very story-centric RPG games in the future, but I don't want to try to graft that sort of thing into an RTS context. I like the fact that players have the freedom to imagine all sorts of scenarios with AI War. They can picture themselves in any of their favorite sci-fi settings, then, instead of any lesser knock-off I could conjure up (there's not a lot of room for quality storytelling in this sort of genre without scripted campaigns).

UKGamer: Do you plan to support the modding of AI War? If so, what areas would you open up to the community? Graphics? Ship behaviour? Star system build and details? etc.

Chris Park: I go back and forth on this. If I ever abandon development on the game, I might just open-source the whole project. Sometime down the line, ten years or so later if sales have slackened off, I'd probably do that anyway. In the meantime, right now I'm basically keeping it closed because I want to act as a gatekeeper for quality and balance. Right now, I'm the content aggregator that takes all of the community ideas, implements them in a consistent and quality way, and makes them available to everybody. Too often with other RTS titles you see the community fragment around a series of mods, often many of which just contain a unit or two that are not cross-compatible with one another.

The way I'm doing it, the overall quality of the game never drops, the community doesn't fragment, and instead of having a lot of little optional offshoots that the majority of players of the game will never see anyway, the player suggestions go right into the core experience that everyone sees. I think that's a win for everyone, but it's only possible while I remain engaged and active with the project. If ever there comes a time where I am not able to fulfill that role, then I'll do something to make sure that someone else can come along and take up the responsibility (or several someones).


UKGamer: As a new indie game developer what do you think you can do to maximise your chances to make the gaming press (both print and web) sit up and take notice of AI War?

Chris Park: With this, I am still learning. I submit a lot of press releases, offer a lot of review opportunities, but of course the majority of those go unanswered because gaming professionals have enough to do with their jobs without looking at a complex game they've never heard of. So mostly it's a matter of getting through the first few tiers of filters first, as far as I can tell -- get as many reviews from smaller sources as possible, get as large and healthy a player community as possible, get on as many of the reputable distribution channels as possible, and then keep sending out those press releases and review offers every so often.

Over the past two months, I've seen a marked improvement in response since we've been on Impulse, and then since we've been doing so well on Impulse and in the few reviews that the game has received. Sites that once didn't respond at all now run the occasional little news blurb about a press release, and occasionally an opportunity just comes out of nowhere like the recent podcast with techZing. Right now it's summer, and so everything is slower even than it otherwise would be, so that doesn't help, but overall things are trending in the right direction.

What major action can I take to really maximize the game's chances of success? I don't think there is any one thing. I don't have any special connections, or inside tracks, or easy paths, or marketing budget. The only thing I know to do is to keep pressing onward with every possibility, to never turn down an opportunity no matter how humble it appears (any publicity can lead to surprising other opportunities, I've learned), to keep writing (hopefully) interesting articles that people see fit to link to, and to keep improving the game through free DLC and player suggestions. Good word of mouth has been invaluable so far. Beyond that, it's just a matter of whether people like the game and therefore choose to give it some coverage. I wish there were a silver bullet, but so far I have not seen one.


UKGamer: Have you learned any lessons so far, whether good or bad, that would help other Indie developers hit warp speed a little faster?

Chris Park: In many respects I don't have enough accumulated data to really give a solid answer to that -- two months isn't a long time for an indie game. Also, I'm really coming to believe that every indie game is going to take a hugely different trajectory. Writing interesting articles works for me because I'm also a writer and there's a lot of unique tech and design elements in play here. On another game, without those talking points, such an approach would be worse than futile. It also really depends on who picks up your game for distribution, where you are able to squeak out some coverage, and where/how big your player base is.

Contests are helpful, obviously -- we entered the PAX 10 but sadly did not win a spot. We'll see how IGF and similar go later on. Having great production values is also clearly important -- I think our slightly-underwhelming art has been a sticking point for some players, but there was nothing I could do about that when all the money was coming out of my pocket. I am not independently wealthy, and taking out a loan for something so speculative as a cooperative 2D RTS seems the height of foolishness. So now that AI War is selling well enough, I'm looking at bringing on an artist to spruce up the visuals and hopefully make the game accessible to a larger crowd. Music in the game is already terriffic thanks to my lucking out in meeting Pablo at just the right time in our respective careers, and I think that has helped win some people over and show that we are serious about the game, not just hobbyists. Professionalism really does count.

Having excellent tutorials and also putting a lot of effort into usability in general has also been a large part of our success, I think. The more complex the game, the more players need to be eased into things. They also need to feel like they are really able to play the game without constantly stumbling or fighting the interface, which can be a challenge in a genre with many varied (often conflicting) expectations. Flexibility in the interface thus becomes key, as does listening to player feedback.

So, thinking about it, I guess that's the number one tip I could give other upcoming indie developers: listen to your players. I really mean it. Don't just nod along and then do your own thing, and for God's sake don't get angry at them for deigning to give you advice. Often their advice is really solid, and even when it's not there's often an underlying need that they are expressing that you should take note of. Big AAA developers have the advantage of giant QA teams, but we indies have the advantage of flexibility. Make the best game you possibly can before release, make sure you have zero open issues on your list when you go gold with any product, but then don't expect there to not be any issues or requests after release. Plan on supporting and expanding your product through at least a honeymoon period with your playerbase, and take their needs seriously. If you treat this as a "take it or leave it" situation, a good number of them will just opt to leave it.


UKGamer: What does the future hold for Arcen Games and Chris Park in terms of games development?

Chris Park: Well, money permitting, I intend to keep doing this for a long, long time. I've been doing game programming and design to some degree as a hobby since I was around nine years old, and it's what I'd like to do with my life if the fates allow it. I'd like to hire a small team of full-time staff with a composer, an artist or two, and another game designer or two. Possibly one more programmer, also. Beyond that, I think we'd be getting too bloated for the sort of lean, highly-creative company I want to run.

I see us making a lot more 2D games, with ever-increasing production values, across a lot of different genres. I really want to make some action platformers, but I also have some very large and involved quasi-Japanese-style RPGs planned. It's always been a personal dream of mine to make a game that could stand in the ranks of Chrono Trigger, Final Fantasy VI, or the other greats of that time period. Those are my two favorite games ever right there, incidentally. Those sorts of titles will require a lot of up-front capital, however, so sales of prior games will have to reach a certain threshhold before I can be comfortable taking on such an ambitious project like that.

When it comes to RTS, I think that AI War is probably going to be it for Arcen Games for the next five years or so, if not longer. I plan to keep on adding to the game with free DLC and expansions for as long as there is interest, but I'm not sure I ever see myself making an AI War 2. To go back to the drawing board and lose all the existing content seems foolish to me with a game like this. Instead, in a few years I expect to basically have so much content out there that it is like three or four games all rolled into one, with a solid and ever-refined interface and engine underlying all of them.

I'm all about iterative development and constant improvement, rather than single releases that you just drop out into the wild and then forget. That approach works very well with business software, and there are a few other games (mostly free or indie) that take this approach, but it will be very interesting to see how my particular brand of iterative development fares with the gaming community. Indications thus far are excellent, but I'm also quite conscious of the fact that I'm standing near the start of a very long road. Good thing I love what I do, or I'd never be able to do it.


A very big thanks for answering our questions Chris, and we here at UKGamer wish you all the best for the evolution of AI War and the success of Arcen Games in the future.