Obsidian Entertainments Dungeon Siege 3 is a polished and highly addictive button-mashing action-RPG that succeeds in creating a story-oriented co-op fantasy dungeon crawler for mainstream console audiences in the tradition of Secret of Mana or Baldurs Gate: Dark Alliance. Though the Dungeon Siege franchise was originally developed by Gas Powered Games, Obsidian is known for taking on sequel projects from other developers, most notably the sequels to BioWares Star Wars: Knights of the Old Republic and Bethesdas Fallout 3. SuicideGirls spoke recently to Dungeon Siege 3s Lead Designer, Nathaniel Chapman, also a veteran of Neverwinter Nights 2 and Fallout: New Vegas, about where this sequel fits in with its predecessors, his design philosophy, and how Diablo is Diablo.
Keith Daniels: So how did Obsidian end up working on a new Dungeon Siege game?
Nathaniel Chapman: Its an interesting, weird triangle of circumstances. Square[-Enix] for a while has been wanting to develop an RPG with a Western developer and just in general work with more Western developers in their games. At the same time, they were working with Gas Powered [Games] to do Supreme Commander. Gas Powered had the Dungeon Siege license, but they had decided to focus more on real-time strategy games and werent really doing much with the Dungeon Siege license. At the same time, Square was also working with us, wanting us to start working on something. So this happy coincidence came together where Square was able to buy the Dungeon Siege license from Gas Powered and we were able to develop the game with it. Thats how it all went down.
KD: I understand that Chris Taylor [Gas Powered Games founder and designer of the original Dungeon Siege] still had some input on this game. Are there some specific places where Dungeon Siege fans can spot his influence?
NC: Theres the broad direction that we took from the earlier games, the lore stuff. We really wanted to bring a lot of touchstones from the earlier games lore into Dungeon Siege 3. We also took the general feel of the game: its a loot-driven action-RPG; its non-stop action. One of the things we do at Obsidian is story, so there is more story and more dialogue, but still the pace is pretty hectic -- which is influenced by the previous games. The loot system is very in-depth; theres lots and lots of gear you can get, which is also influenced by the earlier games.
In terms of specific stuff that he contributed to this game it was more kind of... feedback. When we were in initial development there were some ideas that our Creative Designer George Ziets would throw out and [Chris] would say, Yeah, that fits well with Dungeon Siege, or, Heres how you can modify that to fit better with the franchise. He also gave us some specific game feedback on things like character control and issues like that throughout development. So its kind of hard to point to it and say, This one thing is Chris Taylors influence on the game. Its more the general attitude of the gameplay and the way that the game feels.
KD: What were the essential elements of the earlier games that you wanted to preserve?
NC: I think the critical thing that people remember from the first Dungeon Siege games is tons of loot. When fans started talking to us on forums the thing they brought up most often was having a really in-depth loot system and having a ton of loot that the player can collect and optimize. That was something we definitely wanted to keep. Also, Dungeon Siege is kind of a no-drama game; its not really complex. None of the games were too complex to get for a new gamer or someone who doesnt want to get into a ton of complexity. Within that, we found ways to provide the RPG customization that you find in our other games while trying to keep the barrier-to-entry low for people who arent used to a ton of stats and a ton of abilities in other RPGs.
KD: The original Dungeon Siege almost played itself it was so automated, but Dungeon Siege 3 is almost arcade-y, very twitchy, action-oriented. Were there some non-RPG games that inspired the gameplay in this?
NC: Oh definitely. I play a ton of fighting and action games like Devil May Cry and God of War and games like that. Our [inaudible] director was also a big fighting game player. So there are definitely influences from that. Obviously, we still wanted it to play like a Dungeon Siege game. The biggest change for the franchise is this is the first one thats coming out on [home] consoles. So we really felt like having a game with a whole bunch of classes wasnt going to be something that console gamers really wanted to play. It also wasnt something we really wanted to do. We wanted to have you take much more control of a single character. So really there was an influence from those kinds of action-y games, but it was also part of bringing it forward for console players and people who want more of an intense action experience in the game.
KD: Where do you think many developers have gone wrong in trying to bring hardcore PC RPG franchises to consoles, and how did you solve that challenge?
NC: I think one mistake that a lot of developers make is that they try to take the mechanics that work on the PC and just directly map them onto a console controller. Then when you inevitably hit the fact that the console controller just doesnt have enough buttons to support the variety of input that you have on a computer game you just start cutting. Like, We have to simplify the number of spells, we have to simplify the mechanics, we have to cut down the amount of stuff that we let the player do. Then I think you end up losing part of what makes the games interesting, which is the ability to choose between those different options. You want to have all those different options available because if you do there are situations where one option is best or another one is best. So you end up, I think, watering down the experience if you take that route. That was one thing that motivated our development of the stance system in Dungeon Siege 3, which is basically that you hit a button and it swaps you to a different stance and the face buttons get mapped to new abilities, because we wanted to have a lot of abilities available for the player to use at any one time without having to go through any menus to remap them or anything like that. So with one hit of a button you now have three new abilities available. Its sort of akin to hitting Shift on the keyboard and now all your keys do something slightly different. That was one of the examples of how we tried to keep the complexity of a PC RPG while molding it onto the console controller instead of just cutting stuff from the PC RPGs to make it work on the consoles.
KD: I really enjoyed the stance mechanic. I felt like the core skill to learn in the game was juggling those stances, and in the higher difficulties you just had to do it faster.
NC: [Laughs] Yeah. Thats definitely the main thing for players to learn.
KD: I liked your blog post discussing your philosophy on adding difficulty to games. How did you go about it with this game?
NC: Increasing damage and increasing monster health has value but only in as much as it requires... Maybe the best way to think of it is that usually when you lose in a game its because you failed to do something, or when it feels good to lose, so to speak -- obviously it never feels great to lose -- but when you feel like you can improve and get better at it you see clearly that you failed at something. How you can make a game harder in a way thats fun is by increasing the number of fail points, meaning that there are new challenges you have to master to beat a harder difficulty level, or you increase the severity of earlier fail points. A good example in Dungeon Siege 3 is that if youre playing on Casual you dont have to block and dodge as much -- you still have to do it some but its not as much of a requirement. By increasing the difficulty, instead of it being a linear scale and increasing the number of times you have to hit a guy, it actually increases the importance of the mechanics in the game relative to just hitting the A button and attacking over and over again.
If our game was more tactical, I actually think increasing monster health and damage [would be] less interesting. It doesnt force you to use a mechanic more or less in a lot of cases, it just makes fights last longer, which does consume more resources but is not an interesting kind of difficulty. I think for an action-type game it actually is a huge difference when monsters do more damage because thats the core action challenge: avoiding getting hit. Thats what we wanted to do. Thats the goal of the Hardcore difficulty. Hardcore plays a lot like Standard except that you need to execute your dodges and your blocks very well, and you need to know how to optimize your stances so you can get guys down [and] do as much damage as possible in as little time as possible. Obviously there were things that we would love to do in a future game. I would love to add new abilities to monsters and do things that introduce new fail points for players on harder difficulty levels, but a lot of that is constrained by time. The challenge that you get is that only a small number of players play on the hardest difficulties. When youre developing a game you have to choose where to spend your resources. You dont want to introduce a crazy new awesome ability, but only on Hardcore, so 80% of people arent going to see it when they play through the game. That can be fun but its something that you have to be really careful about when youre developing a game because its often a bad idea to spend resources on something that only a small number of players are going to see.
KD: In the old Dungeon Siege games, you would pause and give everyone orders, but in this one the AI does its own thing. Did leaving the control of the companion characters up to the AI pose any design challenges? For example, did some encounters have to be planned around the limitations of AI?
NC: One of the things that we decided pretty early on, because we were going with a more action focus, is that we felt it would just be too stressful for the player to manage their companions directly, to directly order around their companions. We were originally going to go half-and-half; we were going to give you a vague control system where you could order a stance, put on Offensive stance, take off Defensive stance, whatever. We actually found that generally that wasnt helpful, that it made more frustrating than not. Weve found that from other games weve worked on, too, where you feel like you almost have control of companions but you dont really. What we thought would work better would be to have our companions work as satellites to the player. They help your damage a little, but when having a companion in Dungeon Siege 3 feels really good is when you get knocked out and they come help you up. Thats the kind of thing we focused on: having them try to be aware of whats going on in the combat and try to help you out when they can. So theyre almost like an extra bonus to the combat, not something extra that you have to manage to win fights.
KD: The example I was thinking of in particular was my favorite boss fight: the [spoiler] in Stonebridge -- [a fight involving waves of mobs along with environmental targets that have to be destroyed and environmental hazards that have to be avoided simultaneously]. That was the only fight where I really noticed the AI struggling, and I was wondering if there werent more fights like that because of some limitation.
NC: It was a challenge. Have you finished the game yet?KD: Oh yeah.
NC: The [spoiler] fight. The [spoiler] will do an attack which, unless youre standing behind a barrier, will nuke everyone in the area. We actually got the AI to run behind those [barriers] and avoid them. The trick really became how much specific AI scripting we did on a specific fight. We didnt want to make it feel like the AI was cheating. Its a tricky line. If the AI dodges the enemys attacks perfectly every single time, but you were not dodging them, it feels unfair. Like..
KD: Hes better than me!
NC: [Laughs] We didnt want to make the player feel bad that he couldnt perform to the level of the AI. It is a tricky balancing act, and its not perfect all the time. But we didnt feel like we had to really pull back on our boss mechanics because of that.
KD: What attracted you to the story of the 10th Legion and the Kingdom of Ehb?
NC: We really wanted to have the player feel like they were part of something bigger because for a lot of players this is their first introduction to the Dungeon Siege franchise. We didnt want to have to introduce new players to this world where theres a bunch of stuff thats already happened and they dont necessarily have a direct investment in it. Thats another reason why we had characters with their own backstories. We wanted the character to have a personal investment in this organization that will restore the Kingdom of Ehb. Its a theme that a lot of people can identify with. Where you live sucks, you want to make it better; it used to be better, so youre going to try to get people working together again and rebuild what used to be great. That kind of theme was very interesting to us, and it worked very well with the fact that this was an established world but not a lot of players would necessarily know a lot about it. We didnt want to have you dealing with the intricacies of the lore from the first two games or anything like that. [The previous games lore] was more like the icing on the cake of the core story of you bringing back the 10th Legion.
KD: Do you see this game as a reboot of the franchise?
NC: Yeah. Certainly in terms of core gameplay I think it is a reboot. Its much more action and a little less automated and tactical. Its a lot like Fallout 3, in a lot of ways, where it is a reboot but it doesnt feel like a reboot. Bethesda stayed really close to the feel of the first two games even though the gameplay is totally different. It still feels like a Fallout game. The core gameplay [in Dungeon Siege 3] is very different. Youre directly controlling a single character, you have combos and all that stuff, but it still feels like a loot-driven action-RPG set in the same world.
KD: I really enjoyed the way the motivations for Odo, Jeyne Kassynder, and the Queen were all complex and understandable in some way. Is it more interesting to you if a villain is relatable?
NC: Absolutely. Something that I joke with Josh Sawyer [Project Director on Fallout: New Vegas] about is the mustache twirling, scheming villain who is doing everything that theyre doing just because theyre evil. You dont encounter someone walking down the street and go, That guys evil! People generally feel like theyre doing the right thing even when they do something that everyone else thinks is terrible. Because our games place a lot of emphasis on choice and consequence, we intentionally try to make the villain a bit more relatable because we often like to let the player choose, Do you think that what the villain did was acceptable? We set up the story to put the villain in the place of someone that you have to fight, someone that is constantly working against you and whom youre working against, but at the end of the game we question the assumption that the player makes throughout the game. Were you doing the right thing? Was this person that youre fighting against really hunting you for no reason? Is there something more deep behind it? I think it makes the characters feel a lot more real and the players choices a lot more meaningful.
KD: I also liked that you could choose to spare the lives of many of the major enemies and their stories continued later in the game. Where did that idea come from?
NC: Its something that weve tried to do in the past in a lot of cases. It is very tricky. We did it recently with Alpha Protocol. Its something that for Obsidian and the previous company that a lot of people came here from, Black Isle Studios, is kind of a hallmark of those [classic PC RPG] games. Its another part of making the enemies more relatable, because when we present players with a character with these complex motivations, we dont want to force the player to kill them. Its not satisfying. We generally like to let the player define their role in the story because it gives them more investment in the story and the world. Its also fun to play around with the idea: what happens when you choose to let someone go? In certain cases it can be bad, in certain cases it can be good. Part of what makes RPGs really interesting is when you have complex choices, and when you make one of those choices its fun to see what the consequences are from them -- especially if theyre not something that you necessarily expect.
KD: One thing that concerned me... The first time I play through a game I usually play the good guy, the paragon. The second time through I usually play like an asshole. [Laughs] In this game when I was playing like an asshole there were times when I felt like I was missing out on certain things. For example, not destroying the [spoiler] in [spoiler]. Are those missed opportunities paid off later if you play that way?
NC: There are some. One example is if you choose to spare [spoiler] or kill [spoiler]. If you choose to kill [spoiler] you get her spear as a weapon that you can equip, and its a really cool, very powerful spear. Whereas, if you dont kill her you dont get it, but you get to talk to her and she takes a role in the story later -- which Im sure if you spared her on your first play-through you saw. In a lot of games you play the good guy and get a great reward, and if you play the bad guy you dont get anything; its very easy to be good. I think its a lot more interesting where if youre good its actually harder. Being a jerk is the easy way to get what you want, and being good is a lot harder. So if you choose to spare someone, sure you might get more story, but you might not get as much loot; there might be some other consequence.
KD: The environments in this game are gorgeous. Is there one that youre particularly proud of?
NC: The art team did an awesome job overall. The thing that Im most impressed with is that they were able to hit so many different architectural and stylistic points really strongly throughout the game. You get the really creepy forest vibe in Rukkenvahl, then the Mournweald is this mystic, messed up forest, then Stonebridge is a huge transition to this steampunk, Renaissance German city. Then Glitterdelve which is the snowy area. I feel like they did a really great job of defining what the style was of each area, visually, and then sticking to it and executing it well. Its hard to pick a specific area because Ive been so close to it for so long, but to me whats really impressive is how they were able to make them all so different but all look very close to the vision of what they wanted them to be.
KD: You mentioned Fallout 3 and Fallout: New Vegas earlier. When New Vegas came out it got a little heat for some technical issues. What did you learn from that, and how did that affect your process going into this game?
NC: Dungeon Siege 3 isnt as huge as Fallout: New Vegas. We feel, and we hope, that when people get it that it doesnt have a lot of those technical issues and its very polished. Weve learned that we can pull back on the riskiest elements. As weve become more experience weve learned, These elements of development are more risky, these ones arent. Even to focus on something that may be very creatively risky, something like choosing whether you kill guys or not and then having the consequences of that, we actually are pretty good at that and its something that weve learned to do fairly well. So we pulled back on things that were not as good at. It takes a lot of self-knowledge [to say,] Yeah, were not going to do that for this game. Instead, were going to focus on the things we do well, doing those really well and making sure its a polished technical experience. Thats part of it.
The other part of it is developing our own technology. This is our first game that uses our own engine, Onyx, which is completely internally developed. Its not necessarily that any of the other engines weve used are bad, theyre all good engines, but having the people who developed the engine in-house has been a huge help. So when we have technical problems its just as easy as going down the hall and getting the programmer who wrote every line of code that were dealing with, bringing them in and having them look at it. So its much easier to diagnose those kinds of problems.
KD: Playing through it about one and a half times so far on the PC version I didnt encounter a single bug or crash.
NC: Thats great, Im glad to hear it.
KD: But what you were saying reminded me of a quote I heard recently: creativity is subtraction.
NC: The designer of Shadow of the Colossus [Fumito Ueda] has a great philosophy. He said something like he knows a game is done not when theres nothing left to add but when theres nothing left to take away. I think thats a really great way to look at developing games. Especially those games like ICO, Shadow of the Colossus, and whatever theyre going to release soon -- which Im excited for but no need to give it too many props. Those games are very focused. Theyre very beautiful in the way they execute on a very core set of ideas and they dont spend a lot of time dealing with the fluff. Every game has to have X, Y, Z, and their approach is, What if we take away X and Y, and Z is some weird version youve never done before? Yeah, were going to have boss battles, but only boss battles, and theyll be the craziest boss battles youve ever seen.
I think RPGs are very different. Part of being an RPG is defining how you want to play the game. So by definition RPGs have a lot of different ways you can play the game, different mechanics, but I think focusing in on what exactly youre trying to say with a game and what experience you want the player to have and then pushing ancillary things aside can really make a game communicate a lot more clearly and make it a lot more focused.
KD: The only thing I found myself wishing for in this game was some sort of Zelda-style lock-on. Im assuming you guys thought of that, too, so I was wondering what the thought process was for not using that, or why it wasnt needed.
NC: Ill be honest. I think our melee feels really, really good. Our ranged combat... if there was a point of the controls that I was less happy with it would be the ranged combat. Actually, I think that where our games control scheme works better than something like Zeldas lock-on system is when youre dealing with a large number of enemies around you and you need to quickly swap between targets -- which is very helpful in our game, especially on harder difficulties where a lot of what youre trying to do is interrupt enemies before they attack you. Youre trying to bounce around between all of the different targets around you. So [in] something like a Ninja Gaiden style we do have a kind of soft-lock where if you hit a button it will try to pick the correct guy to aim at. It doesnt just blind fire. Where thats a lot trickier though is in ranged combat. I think thats where a lock-on system might be better. Characters like Katarina still feel really satisfying because her pistol/shotgun stance feels kind of like melee in a way. I think that works really well. I still think ranged combat is really fun in DS3, especially with Katarina because you can kite guys. That character is super fun. But if I were to do everything over again, given infinite resources and infinite time to polish, I would probably have some sort of ranged lock-on system.
I was talking with another designer and one of the things that occurred to me was that when you do two things in a game, the benefit isnt additive, its multiplicative. Like if you do two things at 50% you dont get a full game at 100% that people love. You get a quarter of a game thats infuriating to people. So if you have half of a good combat system and half of a good dialogue system its not like, Oh, you guys made a great game! It all adds up to one full game! You have to make sure that each aspect is completely solid on its own. So part of the question is, If wed focused on doing lock-on for ranged, what would the cost have been in our melee combat, and would that have hurt the game rather than helped it? Thats one of the things where having a sequel is nice because you can build on top of the stuff that youve already done. So, like Fallout 3, they had Oblivian, and so they had all the stuff that theyd done in Oblivian and they could make ranged combat and guns feel really good in Fallout.
KD: What do you anticipate will be the major areas of praise and criticism for this game?
NC: Im pretty close to it, so I guess Ill say more what I hope. I hope people really like the combat because I think its a big departure from the previous games combat and is, I think, pretty different from most RPG combat systems because it has that heavy action influence. I hope that people enjoy that and get into having more of a straight up action experience mixed with the RPG stats and item gathering. As a developer you can pretend to be confident about anything but you dont really know until it hits a lot of players hands. Obviously we do play-testing, but especially as the gameplay guy my biggest concern is always, Do people enjoy and get the core mechanics? I do think that people who play it, especially in multiplayer, like Baldurs Gate: Dark Alliance or the old Secret of Mana games -- I played Secret of Mana beginning to end with my best friend when I was growing up -- people who play the game like that, co-op, I think are going to have an awesome time.
But we intentionally did not set out to make Diablo from the very beginning. I think people who only want to play Diablo are probably not going to be satisfied with the multiplayer in the game. Thats OK, because Diablo is Diablo and Im sure it will be great at being Diablo, but we wanted to do something different. We wanted to make that experience of going through the game from beginning to end with your buddy, even if your buddy is online. Its hard to know exactly what peoples reaction will be. My hope is that people want what it is the game does well.
KD: What do you have planned for the future of the Dungeon Siege franchise?
NC: Wow. Well, we dont have any super-specific plans. Obviously, Square still owns the franchise. We would love to continue to work on it but its kind of up to what Square wants to do. Its been great working with them, so I think that hopefully well make a Dungeon Siege 4, but it all depends on what we can do and what Square wants to do.
But, from a gameplay perspective, we have those choices in the game where you see the consequences of them later on. I would love to see more of that layer throughout the game even more. We probably could do a more fleshed-out multiplayer system in a future game. When youre developing a game, people will always bring up ideas and you know theres no way youre going to get to them in time, so the joke is always, Oh, thats Dungeon Siege 4, or Thats Alpha Protocol 2. So every time you finish a game you always have this long list of things that you thought of that would be really awesome but you just had no way to get to during development. Part of it is also looking at peoples different reactions to the game. So if we find that people love the combat system [you say,] Awesome! Lets give them more.
Keith Daniels: So how did Obsidian end up working on a new Dungeon Siege game?
Nathaniel Chapman: Its an interesting, weird triangle of circumstances. Square[-Enix] for a while has been wanting to develop an RPG with a Western developer and just in general work with more Western developers in their games. At the same time, they were working with Gas Powered [Games] to do Supreme Commander. Gas Powered had the Dungeon Siege license, but they had decided to focus more on real-time strategy games and werent really doing much with the Dungeon Siege license. At the same time, Square was also working with us, wanting us to start working on something. So this happy coincidence came together where Square was able to buy the Dungeon Siege license from Gas Powered and we were able to develop the game with it. Thats how it all went down.
KD: I understand that Chris Taylor [Gas Powered Games founder and designer of the original Dungeon Siege] still had some input on this game. Are there some specific places where Dungeon Siege fans can spot his influence?
NC: Theres the broad direction that we took from the earlier games, the lore stuff. We really wanted to bring a lot of touchstones from the earlier games lore into Dungeon Siege 3. We also took the general feel of the game: its a loot-driven action-RPG; its non-stop action. One of the things we do at Obsidian is story, so there is more story and more dialogue, but still the pace is pretty hectic -- which is influenced by the previous games. The loot system is very in-depth; theres lots and lots of gear you can get, which is also influenced by the earlier games.
In terms of specific stuff that he contributed to this game it was more kind of... feedback. When we were in initial development there were some ideas that our Creative Designer George Ziets would throw out and [Chris] would say, Yeah, that fits well with Dungeon Siege, or, Heres how you can modify that to fit better with the franchise. He also gave us some specific game feedback on things like character control and issues like that throughout development. So its kind of hard to point to it and say, This one thing is Chris Taylors influence on the game. Its more the general attitude of the gameplay and the way that the game feels.
KD: What were the essential elements of the earlier games that you wanted to preserve?
NC: I think the critical thing that people remember from the first Dungeon Siege games is tons of loot. When fans started talking to us on forums the thing they brought up most often was having a really in-depth loot system and having a ton of loot that the player can collect and optimize. That was something we definitely wanted to keep. Also, Dungeon Siege is kind of a no-drama game; its not really complex. None of the games were too complex to get for a new gamer or someone who doesnt want to get into a ton of complexity. Within that, we found ways to provide the RPG customization that you find in our other games while trying to keep the barrier-to-entry low for people who arent used to a ton of stats and a ton of abilities in other RPGs.
KD: The original Dungeon Siege almost played itself it was so automated, but Dungeon Siege 3 is almost arcade-y, very twitchy, action-oriented. Were there some non-RPG games that inspired the gameplay in this?
NC: Oh definitely. I play a ton of fighting and action games like Devil May Cry and God of War and games like that. Our [inaudible] director was also a big fighting game player. So there are definitely influences from that. Obviously, we still wanted it to play like a Dungeon Siege game. The biggest change for the franchise is this is the first one thats coming out on [home] consoles. So we really felt like having a game with a whole bunch of classes wasnt going to be something that console gamers really wanted to play. It also wasnt something we really wanted to do. We wanted to have you take much more control of a single character. So really there was an influence from those kinds of action-y games, but it was also part of bringing it forward for console players and people who want more of an intense action experience in the game.
KD: Where do you think many developers have gone wrong in trying to bring hardcore PC RPG franchises to consoles, and how did you solve that challenge?
NC: I think one mistake that a lot of developers make is that they try to take the mechanics that work on the PC and just directly map them onto a console controller. Then when you inevitably hit the fact that the console controller just doesnt have enough buttons to support the variety of input that you have on a computer game you just start cutting. Like, We have to simplify the number of spells, we have to simplify the mechanics, we have to cut down the amount of stuff that we let the player do. Then I think you end up losing part of what makes the games interesting, which is the ability to choose between those different options. You want to have all those different options available because if you do there are situations where one option is best or another one is best. So you end up, I think, watering down the experience if you take that route. That was one thing that motivated our development of the stance system in Dungeon Siege 3, which is basically that you hit a button and it swaps you to a different stance and the face buttons get mapped to new abilities, because we wanted to have a lot of abilities available for the player to use at any one time without having to go through any menus to remap them or anything like that. So with one hit of a button you now have three new abilities available. Its sort of akin to hitting Shift on the keyboard and now all your keys do something slightly different. That was one of the examples of how we tried to keep the complexity of a PC RPG while molding it onto the console controller instead of just cutting stuff from the PC RPGs to make it work on the consoles.
KD: I really enjoyed the stance mechanic. I felt like the core skill to learn in the game was juggling those stances, and in the higher difficulties you just had to do it faster.
NC: [Laughs] Yeah. Thats definitely the main thing for players to learn.
KD: I liked your blog post discussing your philosophy on adding difficulty to games. How did you go about it with this game?
NC: Increasing damage and increasing monster health has value but only in as much as it requires... Maybe the best way to think of it is that usually when you lose in a game its because you failed to do something, or when it feels good to lose, so to speak -- obviously it never feels great to lose -- but when you feel like you can improve and get better at it you see clearly that you failed at something. How you can make a game harder in a way thats fun is by increasing the number of fail points, meaning that there are new challenges you have to master to beat a harder difficulty level, or you increase the severity of earlier fail points. A good example in Dungeon Siege 3 is that if youre playing on Casual you dont have to block and dodge as much -- you still have to do it some but its not as much of a requirement. By increasing the difficulty, instead of it being a linear scale and increasing the number of times you have to hit a guy, it actually increases the importance of the mechanics in the game relative to just hitting the A button and attacking over and over again.
If our game was more tactical, I actually think increasing monster health and damage [would be] less interesting. It doesnt force you to use a mechanic more or less in a lot of cases, it just makes fights last longer, which does consume more resources but is not an interesting kind of difficulty. I think for an action-type game it actually is a huge difference when monsters do more damage because thats the core action challenge: avoiding getting hit. Thats what we wanted to do. Thats the goal of the Hardcore difficulty. Hardcore plays a lot like Standard except that you need to execute your dodges and your blocks very well, and you need to know how to optimize your stances so you can get guys down [and] do as much damage as possible in as little time as possible. Obviously there were things that we would love to do in a future game. I would love to add new abilities to monsters and do things that introduce new fail points for players on harder difficulty levels, but a lot of that is constrained by time. The challenge that you get is that only a small number of players play on the hardest difficulties. When youre developing a game you have to choose where to spend your resources. You dont want to introduce a crazy new awesome ability, but only on Hardcore, so 80% of people arent going to see it when they play through the game. That can be fun but its something that you have to be really careful about when youre developing a game because its often a bad idea to spend resources on something that only a small number of players are going to see.
KD: In the old Dungeon Siege games, you would pause and give everyone orders, but in this one the AI does its own thing. Did leaving the control of the companion characters up to the AI pose any design challenges? For example, did some encounters have to be planned around the limitations of AI?
NC: One of the things that we decided pretty early on, because we were going with a more action focus, is that we felt it would just be too stressful for the player to manage their companions directly, to directly order around their companions. We were originally going to go half-and-half; we were going to give you a vague control system where you could order a stance, put on Offensive stance, take off Defensive stance, whatever. We actually found that generally that wasnt helpful, that it made more frustrating than not. Weve found that from other games weve worked on, too, where you feel like you almost have control of companions but you dont really. What we thought would work better would be to have our companions work as satellites to the player. They help your damage a little, but when having a companion in Dungeon Siege 3 feels really good is when you get knocked out and they come help you up. Thats the kind of thing we focused on: having them try to be aware of whats going on in the combat and try to help you out when they can. So theyre almost like an extra bonus to the combat, not something extra that you have to manage to win fights.
KD: The example I was thinking of in particular was my favorite boss fight: the [spoiler] in Stonebridge -- [a fight involving waves of mobs along with environmental targets that have to be destroyed and environmental hazards that have to be avoided simultaneously]. That was the only fight where I really noticed the AI struggling, and I was wondering if there werent more fights like that because of some limitation.
NC: It was a challenge. Have you finished the game yet?KD: Oh yeah.
NC: The [spoiler] fight. The [spoiler] will do an attack which, unless youre standing behind a barrier, will nuke everyone in the area. We actually got the AI to run behind those [barriers] and avoid them. The trick really became how much specific AI scripting we did on a specific fight. We didnt want to make it feel like the AI was cheating. Its a tricky line. If the AI dodges the enemys attacks perfectly every single time, but you were not dodging them, it feels unfair. Like..
KD: Hes better than me!
NC: [Laughs] We didnt want to make the player feel bad that he couldnt perform to the level of the AI. It is a tricky balancing act, and its not perfect all the time. But we didnt feel like we had to really pull back on our boss mechanics because of that.
KD: What attracted you to the story of the 10th Legion and the Kingdom of Ehb?
NC: We really wanted to have the player feel like they were part of something bigger because for a lot of players this is their first introduction to the Dungeon Siege franchise. We didnt want to have to introduce new players to this world where theres a bunch of stuff thats already happened and they dont necessarily have a direct investment in it. Thats another reason why we had characters with their own backstories. We wanted the character to have a personal investment in this organization that will restore the Kingdom of Ehb. Its a theme that a lot of people can identify with. Where you live sucks, you want to make it better; it used to be better, so youre going to try to get people working together again and rebuild what used to be great. That kind of theme was very interesting to us, and it worked very well with the fact that this was an established world but not a lot of players would necessarily know a lot about it. We didnt want to have you dealing with the intricacies of the lore from the first two games or anything like that. [The previous games lore] was more like the icing on the cake of the core story of you bringing back the 10th Legion.
KD: Do you see this game as a reboot of the franchise?
NC: Yeah. Certainly in terms of core gameplay I think it is a reboot. Its much more action and a little less automated and tactical. Its a lot like Fallout 3, in a lot of ways, where it is a reboot but it doesnt feel like a reboot. Bethesda stayed really close to the feel of the first two games even though the gameplay is totally different. It still feels like a Fallout game. The core gameplay [in Dungeon Siege 3] is very different. Youre directly controlling a single character, you have combos and all that stuff, but it still feels like a loot-driven action-RPG set in the same world.
KD: I really enjoyed the way the motivations for Odo, Jeyne Kassynder, and the Queen were all complex and understandable in some way. Is it more interesting to you if a villain is relatable?
NC: Absolutely. Something that I joke with Josh Sawyer [Project Director on Fallout: New Vegas] about is the mustache twirling, scheming villain who is doing everything that theyre doing just because theyre evil. You dont encounter someone walking down the street and go, That guys evil! People generally feel like theyre doing the right thing even when they do something that everyone else thinks is terrible. Because our games place a lot of emphasis on choice and consequence, we intentionally try to make the villain a bit more relatable because we often like to let the player choose, Do you think that what the villain did was acceptable? We set up the story to put the villain in the place of someone that you have to fight, someone that is constantly working against you and whom youre working against, but at the end of the game we question the assumption that the player makes throughout the game. Were you doing the right thing? Was this person that youre fighting against really hunting you for no reason? Is there something more deep behind it? I think it makes the characters feel a lot more real and the players choices a lot more meaningful.
KD: I also liked that you could choose to spare the lives of many of the major enemies and their stories continued later in the game. Where did that idea come from?
NC: Its something that weve tried to do in the past in a lot of cases. It is very tricky. We did it recently with Alpha Protocol. Its something that for Obsidian and the previous company that a lot of people came here from, Black Isle Studios, is kind of a hallmark of those [classic PC RPG] games. Its another part of making the enemies more relatable, because when we present players with a character with these complex motivations, we dont want to force the player to kill them. Its not satisfying. We generally like to let the player define their role in the story because it gives them more investment in the story and the world. Its also fun to play around with the idea: what happens when you choose to let someone go? In certain cases it can be bad, in certain cases it can be good. Part of what makes RPGs really interesting is when you have complex choices, and when you make one of those choices its fun to see what the consequences are from them -- especially if theyre not something that you necessarily expect.
KD: One thing that concerned me... The first time I play through a game I usually play the good guy, the paragon. The second time through I usually play like an asshole. [Laughs] In this game when I was playing like an asshole there were times when I felt like I was missing out on certain things. For example, not destroying the [spoiler] in [spoiler]. Are those missed opportunities paid off later if you play that way?
NC: There are some. One example is if you choose to spare [spoiler] or kill [spoiler]. If you choose to kill [spoiler] you get her spear as a weapon that you can equip, and its a really cool, very powerful spear. Whereas, if you dont kill her you dont get it, but you get to talk to her and she takes a role in the story later -- which Im sure if you spared her on your first play-through you saw. In a lot of games you play the good guy and get a great reward, and if you play the bad guy you dont get anything; its very easy to be good. I think its a lot more interesting where if youre good its actually harder. Being a jerk is the easy way to get what you want, and being good is a lot harder. So if you choose to spare someone, sure you might get more story, but you might not get as much loot; there might be some other consequence.
KD: The environments in this game are gorgeous. Is there one that youre particularly proud of?
NC: The art team did an awesome job overall. The thing that Im most impressed with is that they were able to hit so many different architectural and stylistic points really strongly throughout the game. You get the really creepy forest vibe in Rukkenvahl, then the Mournweald is this mystic, messed up forest, then Stonebridge is a huge transition to this steampunk, Renaissance German city. Then Glitterdelve which is the snowy area. I feel like they did a really great job of defining what the style was of each area, visually, and then sticking to it and executing it well. Its hard to pick a specific area because Ive been so close to it for so long, but to me whats really impressive is how they were able to make them all so different but all look very close to the vision of what they wanted them to be.
KD: You mentioned Fallout 3 and Fallout: New Vegas earlier. When New Vegas came out it got a little heat for some technical issues. What did you learn from that, and how did that affect your process going into this game?
NC: Dungeon Siege 3 isnt as huge as Fallout: New Vegas. We feel, and we hope, that when people get it that it doesnt have a lot of those technical issues and its very polished. Weve learned that we can pull back on the riskiest elements. As weve become more experience weve learned, These elements of development are more risky, these ones arent. Even to focus on something that may be very creatively risky, something like choosing whether you kill guys or not and then having the consequences of that, we actually are pretty good at that and its something that weve learned to do fairly well. So we pulled back on things that were not as good at. It takes a lot of self-knowledge [to say,] Yeah, were not going to do that for this game. Instead, were going to focus on the things we do well, doing those really well and making sure its a polished technical experience. Thats part of it.
The other part of it is developing our own technology. This is our first game that uses our own engine, Onyx, which is completely internally developed. Its not necessarily that any of the other engines weve used are bad, theyre all good engines, but having the people who developed the engine in-house has been a huge help. So when we have technical problems its just as easy as going down the hall and getting the programmer who wrote every line of code that were dealing with, bringing them in and having them look at it. So its much easier to diagnose those kinds of problems.
KD: Playing through it about one and a half times so far on the PC version I didnt encounter a single bug or crash.
NC: Thats great, Im glad to hear it.
KD: But what you were saying reminded me of a quote I heard recently: creativity is subtraction.
NC: The designer of Shadow of the Colossus [Fumito Ueda] has a great philosophy. He said something like he knows a game is done not when theres nothing left to add but when theres nothing left to take away. I think thats a really great way to look at developing games. Especially those games like ICO, Shadow of the Colossus, and whatever theyre going to release soon -- which Im excited for but no need to give it too many props. Those games are very focused. Theyre very beautiful in the way they execute on a very core set of ideas and they dont spend a lot of time dealing with the fluff. Every game has to have X, Y, Z, and their approach is, What if we take away X and Y, and Z is some weird version youve never done before? Yeah, were going to have boss battles, but only boss battles, and theyll be the craziest boss battles youve ever seen.
I think RPGs are very different. Part of being an RPG is defining how you want to play the game. So by definition RPGs have a lot of different ways you can play the game, different mechanics, but I think focusing in on what exactly youre trying to say with a game and what experience you want the player to have and then pushing ancillary things aside can really make a game communicate a lot more clearly and make it a lot more focused.
KD: The only thing I found myself wishing for in this game was some sort of Zelda-style lock-on. Im assuming you guys thought of that, too, so I was wondering what the thought process was for not using that, or why it wasnt needed.
NC: Ill be honest. I think our melee feels really, really good. Our ranged combat... if there was a point of the controls that I was less happy with it would be the ranged combat. Actually, I think that where our games control scheme works better than something like Zeldas lock-on system is when youre dealing with a large number of enemies around you and you need to quickly swap between targets -- which is very helpful in our game, especially on harder difficulties where a lot of what youre trying to do is interrupt enemies before they attack you. Youre trying to bounce around between all of the different targets around you. So [in] something like a Ninja Gaiden style we do have a kind of soft-lock where if you hit a button it will try to pick the correct guy to aim at. It doesnt just blind fire. Where thats a lot trickier though is in ranged combat. I think thats where a lock-on system might be better. Characters like Katarina still feel really satisfying because her pistol/shotgun stance feels kind of like melee in a way. I think that works really well. I still think ranged combat is really fun in DS3, especially with Katarina because you can kite guys. That character is super fun. But if I were to do everything over again, given infinite resources and infinite time to polish, I would probably have some sort of ranged lock-on system.
I was talking with another designer and one of the things that occurred to me was that when you do two things in a game, the benefit isnt additive, its multiplicative. Like if you do two things at 50% you dont get a full game at 100% that people love. You get a quarter of a game thats infuriating to people. So if you have half of a good combat system and half of a good dialogue system its not like, Oh, you guys made a great game! It all adds up to one full game! You have to make sure that each aspect is completely solid on its own. So part of the question is, If wed focused on doing lock-on for ranged, what would the cost have been in our melee combat, and would that have hurt the game rather than helped it? Thats one of the things where having a sequel is nice because you can build on top of the stuff that youve already done. So, like Fallout 3, they had Oblivian, and so they had all the stuff that theyd done in Oblivian and they could make ranged combat and guns feel really good in Fallout.
KD: What do you anticipate will be the major areas of praise and criticism for this game?
NC: Im pretty close to it, so I guess Ill say more what I hope. I hope people really like the combat because I think its a big departure from the previous games combat and is, I think, pretty different from most RPG combat systems because it has that heavy action influence. I hope that people enjoy that and get into having more of a straight up action experience mixed with the RPG stats and item gathering. As a developer you can pretend to be confident about anything but you dont really know until it hits a lot of players hands. Obviously we do play-testing, but especially as the gameplay guy my biggest concern is always, Do people enjoy and get the core mechanics? I do think that people who play it, especially in multiplayer, like Baldurs Gate: Dark Alliance or the old Secret of Mana games -- I played Secret of Mana beginning to end with my best friend when I was growing up -- people who play the game like that, co-op, I think are going to have an awesome time.
But we intentionally did not set out to make Diablo from the very beginning. I think people who only want to play Diablo are probably not going to be satisfied with the multiplayer in the game. Thats OK, because Diablo is Diablo and Im sure it will be great at being Diablo, but we wanted to do something different. We wanted to make that experience of going through the game from beginning to end with your buddy, even if your buddy is online. Its hard to know exactly what peoples reaction will be. My hope is that people want what it is the game does well.
KD: What do you have planned for the future of the Dungeon Siege franchise?
NC: Wow. Well, we dont have any super-specific plans. Obviously, Square still owns the franchise. We would love to continue to work on it but its kind of up to what Square wants to do. Its been great working with them, so I think that hopefully well make a Dungeon Siege 4, but it all depends on what we can do and what Square wants to do.
But, from a gameplay perspective, we have those choices in the game where you see the consequences of them later on. I would love to see more of that layer throughout the game even more. We probably could do a more fleshed-out multiplayer system in a future game. When youre developing a game, people will always bring up ideas and you know theres no way youre going to get to them in time, so the joke is always, Oh, thats Dungeon Siege 4, or Thats Alpha Protocol 2. So every time you finish a game you always have this long list of things that you thought of that would be really awesome but you just had no way to get to during development. Part of it is also looking at peoples different reactions to the game. So if we find that people love the combat system [you say,] Awesome! Lets give them more.