Copdahl Continues His Campaign Against the Humans

Copdahl Continues His Campaign Against the Humans

You might recall a long Q&A with Chris Opdahl a week or so back. Here's the rest.
bluerunner asks, "How big of an impact were armor abilities in planning out the missions?" and "How critical are using armor abilities to completing the campaign levels?"

Chris: Armor Abilities range from critical to making a mission to the spice that is added later to make it more varied. Some moments in the game were specifically designed around an Armor Ability, and those moment would not work without them. Jetpack is a good example of an Armor Ability that falls into that category. We need to specifically design spaces around the Jetpack (in the same way we need to design them around Banshees and Warthogs), and in some cases the player could not even navigate the space without it.

Other Armor Abilities were much more designed to be sandbox spice. Armor Lock, Active Camo, Hologram, and Drop Shield were all thought of as pieces that we would spread throughout the missions to let people chose what kind of Armor Ability they wanted to play with on that specific play through. And other Armor Abilities (like Sprint) were the ones we saw as ubiquitous to the player experience and would be liberally placed throughout the campaign.

And, for what it is worth, there were some Herculean efforts by the Character Engineering team to make the AI react to Armor Abilities in a very fun fashion. We were coming up on ZBR (where the Engineers should not be adding new features to the game), and the AI still needed some help in their reaction to some of the Armor Abilities (not trivially because the design specs for them were a little too complex - who the heck wrote those docs... oh yeah, I did... oops). The Engineers came in and made it work, and now all of the Armor Abilities have really fun mechanics in the game. So when you trick an Elite with the Hologram, give a little shout out to the Character Engineering team. Vandle Valsher asks, "How did the process of creating Reach's campaign experience compare to Halo 1, 2, and 3? For instance: was this project harder or easier? Did everything go relatively smoothly or were the obstacles much larger this time around?" and "How has the communication among your team improved over the other games? What lessons did you learn this time around? Were there any lessons from the other games that helped you avoid trouble this time?" and "Coke or Pepsi?"
Chris: So comparing the campaign experiences to Halo 1, 2 and 3 is pretty difficult for me. I only worked on Halo 3, and I came in for the last nine months or so. The mission I took over for Halo 3 was already moving towards finishing so I was only really there to set up the encounters and write the final mission script. So I don't have any insight into Reach development vs. Halo 1, 2 or 3.

I found that Reach went extremely smoothly. That is not to say there were no wrinkles, nor any fights (there were!). But in general we had a solid plan for everything and we course corrected properly when we needed to.

Communication was pretty strong on Reach (although I can't speak to it in relation to the previous Halo games). For me, the most important parts of communication are: 1) people know which decisions are being made 2) people know why those decisions are made and 3) people know who is responsible for making those decisions. If those basic criteria are met then people have enough ability to track down what is going on and can have a voice. Not meeting those basic criteria mean that people get blindsided by decisions and/or don't feel like they can affect change because they don't know what is going on or who to talk to.

The biggest lesson I learned (at least process wise) was to work on the core elements of the game first and then add on the rest later. It is really easy to get caught up solving the new and exciting problems now, and waiting to solve the known (old) problems later. Unfortunately the known problems are often the core elements of the game, and getting that right as soon as possible is really important. Even if that means you cut a lot of the new flashy parts of the game. That said, all the important parts of the game came together, but it was a little hot there for a while!

macgyver10 asks, "what 80's TV action hero inspired you most growing up?"

Jake Cutter (although I would pick your namesake as a close second).

FyreWulff asks, "It has been said that in Reach, players can get much farther from each other than in Halo 3. Now the 360 still has the same amount of RAM as it did when Halo 3 came out. What kind of optimizations were done to pull this off, or is the Reach engine just that much better? Or is it just better level design to work around the 360 hardware?

An example I would make is how in ODST, you can go all the way from the start of the level to just halfway over the "Space elevator explodes" bridge before you'll teleport, which is a relatively far distance for a Halo game. How far would the Reach engine let us go before a teleport on the same level with all of Reach's optimizations?"

And asks, "Have the campaign editing tools improved dramatically since Halo 3? Do designers still have to manually mark areas for AI, vis portals, etc, or have some of this stuff been automated (or the automation improved) ."

And asks, "We know now that the enemy encounters will adapt themselves to the player count in campaign. As a result of this, I imagine you can let the engine help add variety into encounters instead of manually placing the bipeds for each encounter (like Halo 1), or is there still a lot of manual setup for this? Example: instead of placing ELITE_MAJOR at 322,244,93 you can instruct the engine ELITE_BATTLEGROUP CENTER AROUND BOX 344,314,393/321,392,291....."

And asks, "In the Halo 3, the missions seemed to be designed around 'don't go backwards,' which was possible to easily do in Halo 1 and Halo 2. This was obviously a result of having to run without a harddrive. Sometimes it was done in a subtle way, like the drop-downs in Sierra 117. Sometimes it was done more blatantly, like the cave become a wall behind you when you get to the dam in Sierra 117 or the invisible wall/one way walls when you get into the Hornets on The Covenant. Does Reach deal with this in a more transparent fashion?"

And asks, "Are you amazed each time people figure out how to skip a level load or do a load early and the Xbox doesn't catch on fire?"

And asks, "From a technical perspective, which level in Halo 3 or ODST was the hardest to get working in complete order and to 'stable' status?"

Chris:

1) The engine is definitely designed around what we wanted to do with Reach, which is to say "Yeah, it is just that much better." It is always easier to build something when you have a solid starting point, and we were able to start making missions really early in the process.
The teleports entirely depend on the circumstances. We have several parts of the game where two players can get quite far away from each other, because we have all of that space loaded into memory at once. But in other locations we end up needed to be very aggressive in teleporting players (because we need to drop out sections in order to load other sections).

2) We improved many of the tools for Reach over Halo 3, but we also added a lot of new responsibility on the Mission Designers. So we simplified some elements while adding new ones. A great example is the new Thespian Tool. This is a tool that allows a Mission Designer to craft a performance in the game that keeps the AI largely in the Sandbox, but allows them to play dialog and animations. We can craft moments with this tool that would have been impossible (or unfeasible) in Halo 3. The upside to this is that we have many more of these moments in the game to keep the player in the world. The downside is that it was another element to building a mission that the Mission Designers needed to manage. All in all we kept the total amount of work pretty similar to Halo 3, in that each Mission Designer was responsible for two missions apiece.

3) Yeah, we have a new tool (that came to us from ODST) called a Squad Formation. In the past we would place a specific location for each AI to spawn. Now we can place a specific location for a group of AI to spawn. We also define Squad Templates now. Squad Template allow us to define a squad once and have multiple missions use that Squad Template. Then we can change that Squad Template in one location and it will change for all of the missions.

4) We will definitely have some of those in Reach as well. This is done for a couple of reasons. 1) It helps keep players from unintentionally backtracking. This makes it easier for the player to keep from getting lost. 2) It allows us to drop all the areas behind the invisible wall, which helps us save memory. That memory is used to make the next section that much more action packed. That said, we did make an effort to use those as rarely as possible, and many sections in the game allow the player a ton of freedom in where they can go at any given moment.

8) My guess is that the Hub in ODST would be the biggest technical hurdle we have had to deal with. "Many Bothans died to bring you that hub." But I did not work on ODST, so I am basing that on rumors and innuendo.

ElzarTheBam asks, "How have you guys updated the behaviour of the enemies, will they still surprise us? Over the years we have grown used to how certain ai's act in the halo games through natural progression over course of the halo games. Is it a challenge to keep things fresh yet identifiable?"

Chris: The thing I love most about Halo AI and the philosophy of AI at Bungie, is that we make them predictable enough that you can create plans and strategies to defeat them, but just when you think you have them figured out they do something you did not expect. I remember a couple of play throughs back, Jaime Griesemer was playing one of the missions. Jaime (knowing what he does about Halo AI - which is to say a LOT) noticed that there only seemed to be one firing point (which is what an AI uses to determine where he can move to) that the AI could reach from a specific vantage point. So Jaime moved to that vantage point and shot and Elite in the head with a Sniper Rifle. Jaime then sat there and waited for the next Elite to come into view, but the Elite did not immediately move to that firing point. Jaime sat there for like 15 seconds and was about to give up on his dirty, dirty strategy when the Elite came up behind him and killed him with a burst from his Plasma Rifle and then a melee strike. So yeah, I expect people will get surprised from time to time.

The BS Police asks, "Are any of the Halo: Reach campaign stages as large if not larger than Halo 1's biggest campaign stages while only counting the overall playable space?"

Chris: Yep. We have one mission that is probably 10,000x times larger than the largest Halo 1 mission (only counting the playable space). And if you watched the E3 news, you probably know which one I am talking about. :)

Of course using that mission is probably cheating on my part, and not really in the spirit of what I think you are asking. So here goes a real answer:

We Have occasionally brought in the Halo 1 environments into Reach (we did it for Halo 3 as well), to get a sense of scale. And I am always surprised as how much bigger everything is in the more recent Halo games than in Halo 1. Silent Cartographer in Halo 1 only takes about 1:15 second to do a lap around the island in a Warthog. I just watched a speed run of Tsavo Highway from Halo 3 and it took them 4 minutes to drive from the start of that mission to the end of it (mostly in a Warthog as well). Granted Silent Cartographer had all the Forerunner structures inside the island, so that mission is a lot longer than 1:15. But if you are just talking playable space, I would gather that Tsavo Highway is more than 2x the playable space of Silent Cartographer. Reach missions are also quite a bit longer than Halo 3 missions.

I have seen multiple Engineers shake their fists at the sky when they talk about how much bigger the spaces are in Reach, and how this problem was not a problem in previous Halo games because the increase in scale. When Engineers shake their fists at the sky, I am either doing my job or getting really close to being fired. I still have a job, so that is something.

ArteenEsben asks, 'How has the workflow and planning of the campaign changed from Halo 3? Since multiplayer and firefight levels come straight from the campaign, how do the campaign and multiplayer teams coordinate with each other? How do you go about making a map work in both campaign and mutiplayer? Halo CE's campaign had a few levels, especially Silent Cartographer the second half of Halo, which were somewhat nonlinear in structure. Have these, or other levels in the previous campaigns, influenced the design of Reach's campaign? What sort of thing does a campaign design lead do? Who do you boss around and who bosses around you? What Halo 3 missions did you work on, and what were your goals for each mission?"

Chris: One of the decisions we made early on, was that if we wanted to share spaces between campaign and multiplayer, one of the two should take point and create the space to meet their needs. The other group should work with whatever was created. We went in with the assumption that multiplayer is much less forgiving to spaces than campaign. A multiplayer space that has exploits is very difficult for the designers to fix, while campaign can set up the encounters and AI in such a way to bring out the best parts of a space and diminish any negatives of a space. Through development we found that the circular style of many multiplayer maps was difficult to design classic Halo encounters, and the more we tried to make a classic Halo encounter in those spaces the less compelling those encounter became. Eventually we embraced the non-linear elements of the multiplayer spaces and made new types of encounters in them. In the end, many of the multiplayer spaces have encounters that are a great change of pace from what we do in the campaign only spaces, and ended up making the campaign just a little more varied.

We definitely embraced both the nonlinear aspects of some of the Halo 1 levels, and the 'wide pipe' aspects of all Halo encounters. Non-linear being a mission where the player is given a variety of objectives and can tackle them in any order and/or a mission that has multiple routes through it. Wide-pipe being an encounter that has multiple lines of attack the player can approach it from and can deal with the encounter in different fashions each time they play. Non-linear is important to make the world feel larger and more varied than it is, and often leads to more exploration on the part of the player. The down side to non-linear is that it is much more difficult to set up and script, and that the player has a much higher likelihood of getting lost. One of the concepts we debated and dealt with a lot for Reach was how much we would allow a player to feel lost. Wide pipe is more about making fun encounters and is something Bungie embraces pretty strongly.

Not surprisingly, there is a very strong correlation between a player feeling lost and them not having 'fun' in a mission. The tricky part is that every player has a different threshold for when exploring becomes lost. Ideally we solve lost issues by the way the mission is designed. Routes that lead to strong landmarks, roads in the terrain map, and loops that lead the player through the spaces are all good ways to keep a mission open while still giving strong directionality to it. We would start there and see how well we could pull the player through the spaces. Then towards the end of the project, if we found that those elements were not enough to keep the player from getting lost, we would add Navpoints as needed. Another element (that came on towards the end of development) was when we added Waypoints over the heads of other Noble Team members. The Mission Designers generally would use the other Spartans to pull the player through the spaces, but they were often not easy to find if there was some obstacle between them and the player. Once we added the Waypoints the player would go to their Spartan buddies first, and then (ideally) that would get them back on track and fix any lost issues.

As Campaign Design Lead, my responsibilities were to work with the Character Engineers, 3D Artists, and Animators on the characters in the game. We defined the behaviors of the enemies, how we wanted characters to look and move through the spaces, what we needed to make encounters work, and what we wanted the player to do during the missions. I also worked with the Mission Designers to determine what the player did in the missions, what the goals were, how the encounters were laid out, and such. A huge part of this job is just setting people up with a direction, do everything you can to set them up for success and then get out of the way and watch the magic happen.

I worked on Tsavo Highway for the main encounter setup and scripting (the last nine months or so), and then on Floodgate for some final encounter design and bug fixing (last couple of months or so). My main goal for Tsavo Highway (aside from making a fun mission) was to give the player a series of encounters that dealt with vehicles in different ways. First one was the player in a vehicle and the Covenant on foot and moving to a new location, then we had the player in a vehicle against entrenched Covenant infantry, then the Covenant had some vehicles. After the bridge jump we added the infantry vs. infantry battle that was meant to be the anchor of the mission (since vehicle encounters are so much more difficult to make challenging and still keep them fun). Then it was vehicle smorgasbord, with player having access to a variety of vehicles and the Covenant coming at him with a wide variety of different looks. Floodgate was much more about doing the final passes on the encounters and trying to find what was fun about fighting the Flood and how to make that mission as fun as possible.

Silent Cartographer always sticks out in my memory as the most memorable mission in the Halo series. But there are a ton of other missions that are right up there with it.

Ragashingo asks, "We haven't heard very much about the Covenant speaking their own languages. How much did that affect level design? For instance Halo 3 had Truth deliver quite a bit of plot info on Tsavo Highway. That's not really an option anymore so did y'all have to find new ways to clue the player into things the Covenant is doing?"

Chris: We went into the game with the assumption that we wanted the Covenant to be more menacing and scary than we had done in the past. We believed the Covenant only speaking alien languages was going to make them more menacing and scary, since the player would have less chance to relate to them. A price for that choice is that we don't have the ability to get into the mind of the Covenant. When the player can overhear conversations (or listen to the Truth holograms), the player is given the chance to understand the motivations of the Covenant. The more you know something, the more it becomes understandable. And the more we understand things the less we find them scary.

All of that said, we definitely had to adapt our story telling in order to give the player a solid understanding about what is happening and still make the Covenant seem like they have a purpose (and still have the player understand that purpose). Some amount of that is done via other characters in the game having insight into what the Covenant are up to, and part of that is done by keeping the Covenant motivations as straight forward as possible (and therefore easier to grasp without dialog).

Kitekiller asks, "I don't know if this has been answered before, but now that there are 5 computer controlled players on top of any human controlled players. During 4 player co-op, will there be 9 Spartans running around during battles, or do extra players replace the usual members of noble team?"

Chris: It depends on the mission and the situation. In some cases the additional co-op players will replace the other members of Noble Team, and in other cases they will be in addition to them. We did a pass very early on to identify the locations where we wanted the player to run around with a full team of Spartans, when we wanted the player to run around with just one of the other Spartans (to give that character more face time with the player), and when we wanted the player to run around solo. So yes, there are moments in the game where there can be a lot of Spartans fighting in the same space. You almost feel sorry for the Grunts that get caught up in that cross fire! Almost...

KalamariKidd asks, "How do you know when you have designed a "good" campaign level? What are some factors that contribute to a good campaign level?"

Chris: I think missions go through three important stages to making them fun. The first one is finding the right hooks to make a mission interesting. These can be gameplay hooks like: fight in this new vehicle, or a sniper mission. They can also be fictional hooks: save Cortana, see the Flood for the first time. These end up defining the entire mission and influencing all of the decisions that follow. The hard part about that stage is that you don't know if you made good choices until you get much further along in the process (and well past the point that you can make any major course corrections).

The next stage is architecting a mission that is fun to play in. This is when the Artists come in, and they work with Mission Designers to build spaces that are: fun to fight in, the player has a clear sense of direction and all of the visual magic gets blocked in (skyboxes, set pieces, etc). You know when this is working and it is pretty obvious when it is right and when it is wrong. We also start having focus tests around this point, which allows us to have fresh eyes on the game to determine if we are off our rockers or not.

Then the last stage comes together really close to the end. Only after the sandbox is locked down, the AI are working properly and the Mission Designers have all the tools to set up the encounters the way they want to. You know this is working when you heard people around the studio having a blast playing the mission. You generally have a pretty good idea yourself, but you can also get too close to it and lose sight of what is working or what is not. So getting fresh eyes is vital.

I think good Halo missions have good hooks, good layout and good combat.

JamesD asks, "What so far has been the biggest setback you've encountered during the production, and how did you overcome it? How does it feel to get bumped up to Mission Design Lead after Halo 3? Do you feel you have more freedom in this position, or do you feel there is more pressure on you to deliver? - When it all gets too much in the studio, and you need to get away from all the Haloz? What do you guys do to let off some steam?"

Chris: I have been thinking about the 'setback' question for the last 10 minutes and I don't know what I would say to it. That is not to say that everything went smooth as silk (because it did not), but there were not any real setbacks that I can come up with. We definitely had some challenges that we overcame, like:

1) We changed many of the ways our tool set worked (in order to do a lot of the things we can do now but could not with the Halo 3 engine), and anytime you do that it takes a while to get back to a point where you are as productive as you were before (let alone becoming more productive).

2) Our spaces are much more detailed than they have been before, and that fidelity made maintaining pathfinding very challenging for the Mission Designer and Mission Architects. The designers would find the issues and then the architects would have to fix them, and every time we made even minor changes to spaces seemed to cause a series of pathfinding issues to crop up.

3) We added the new Thespian tool that had many elements in the game rely on, and it took us much longer to get it fully functional than we hoped.

We overcame all of these types of issues by devoting the resources, hiring the geniuses that scamper through Bungie to find the right solutions, course correcting to get the best use out of what you have (and not always what you want it to be), and just continuing to hammer on everything until it is done.

I was Campaign Design Lead for about 2.5 years on Reach, and I would say the first 2 years was not that much different than Halo 3. Sure the things I did day to day were different, but I still viewed the job as getting the right people together to make awesome (and when you are surrounded by the talent at Bungie, this part is kind of easy). On Halo 3 I did that for one mission, on Reach I did it for the campaign. The scope is different but the actual work is largely the same. Then the last six months I figured out what makes this job what it is (scary, crazy, hard, but awesome!)

The last six months was all about looking at the Campaign and making decisions about what to keep and what to cut, how to keep people motivated to make everything amazing, sheltering people from everything that keeps them from doing their job, making sure people don't overwork the elements that are working well (since it is all too easy to keep churning on something when you really should just step away and wrap it up with a nice bow), and (probably most importantly) walking people away from the ledge when they are concerned/panicking/freaking out about the current fire that someone else is in the process of fighting. I had access to information that gave me a perspective on the project that not many other people at the studio had. That allowed me to see what was coming down the line in ways that they could not. So a huge part of my job was calming people's nerves and trying to keep everyone focused on their respective job. Now the hard part for them in this situation is that I am new to this job, and they have not seen if I know what the heck I am doing. So it took a while to get everyone's confidence.

Gravemind asks, "How prevalent are invisible walls in Campaign? I can understand invisible walls in multiplayer to keep players in-bounds, but I don't want them to be unnecessarily obtrusive in Campaign. Halo 3 in particular was very restrictive due to heavy use of invisible walls, blocking access to places not just out of the way, but even those easily within reach, which annoys players like myself who're keen on exploration. So, can we expect invisible walls to be less obtrusive than in Halo 3, perhaps being scaled back to what we saw in Halo 1 & 2?"

Chris: We use invisible walls (soft ceilings) to keep the player in the right play space. Often that is in order to keep a player from unintentionally getting lost because they got someplace that they should not have. We also use them so that we don't have to spend the finishing art time on the less important locations in the game. All of that said, we made a much more concerted effort to design large spaces that would allow players to explore more than they were able to recently. And hopefully that pays off for people like you who enjoy exploring!

BuzzJuice asks, "When you get four players, you get ton more enemies, AI becomes stronger and more cunning. Is there a skull or an option to add this difficulty w/o having 4 players in for Mythic/SLASO (Solo Legendary All Skulls On) fans like me to get a kick out of the campaign? Any funny moments or shenanigans you would like to share? Have you ever considered music when designing levels? Like let's make this awesome encounter even longer so that Marty can make the perfect score for this? Or don't worry if this battle seems bland, the sound guys will put in some orchestra and make this epic. Has the music in the campaign influenced your design? Any tips on what it takes to be a campaign design lead? You need creativity, leadership, decision making? What do you think is most important out of all three (or besides these three)?"

Chris: We actually do not add a ton of additional enemies when you play coop, but we do make them more dangerous (length of bursts, shorten the delay between bursts, higher evasion chances, AI recharge health and shields faster). We do not have a skull to make a single player game emulate a 4 player co-op game. That said, I think you will have a pretty serious challenge on single player Legendary with all skulls on.

My most recent funny moment had to do with the Audio Team for the Reach. Jay Weinland picked up this mixer system for all the Designers. It allows you to have multiple inputs that play through one set of head phones. So I have my 360 devkit and my PC sound system go through the same headset. So anyway, I got a recent build of Reach and fired up the main menu. Right as the menu popped on screen a song from the Lord of the Rings soundtrack started playing. Without even thinking about what might cause that, I assumed it was some temp music for the main menu. Well we were getting a little late in the project to have temp music around, so I figured I would go talk to Jay (the Audio Lead) and see what was going on. Well, Jay is in his sound proof office with Cyril (our Audio Engineer and C Paul), and I ask why we have Lord of the Rings music in the main menu. Jay gets a funny look on his face and calls me crazy. We have a few minute conversations and then Jay starts playing the tracks that are in the main menu sound folder, and they are all the sound files you would expect. No Lord of the Rings anywhere. Everyone started getting curious about what could cause this, when my brain finally turned on. I figured out that I likely was listening to my PC mp3 collection, and that was my sound file, and not the main menu I heard. So I went back to my desk and double checked, and it turned out I was simply sleep deprived, and there were no unlicensed sound files on the main menu. Embarrassed as I was, I still feel like I dodged a bullet. A different sound track could have been playing and I would have stormed into his office asking why the heck Britney Spears is on the main menu!

We definitely think about the mood of music when we design levels, and we attempt to give Marty and idea of what we are thinking. But at the end of the day, Marty is the master behind all of that and we just sit around and wait for his magic to get into the missions. Marty ends up playing through the missions (when they get close to being completed) and makes all of the decisions how he wants to pace out the music to what we have in the game.

Creativity, leadership and decision making are all incredibly important for any design lead. I always view leads as people who fill in the blanks of the team as needed. If the team needs creative inspiration then the lead needs to fill that void. If the team has an abundance of creativity then the lead needs to channel that and get the most out of people. For Reach we have more than enough talent to fill the technical and creative needs of the project, so I viewed my role as one to channel that talent and make sure it was all pointing in the right directions. The best leads I have worked with know when to charge ahead and get everyone to follow and when to pull back and let people devour all of the issues that come up.

Time Glitch asks, "How will Armor Abilities play into the campaign? Will we pick them up? Will we select them before a mission? Will we GET to select them? Will high-ranked Elites still rage and pull out their swords from time to time? Will the AI in higher difficulty levels not be fooled by simple tricks like Holograms and Active Camo? Can you sword clang with a sword-wielding Elite?"

Chris: The player begins each mission with a specific Armor Ability (which the Mission Designer determines), and then the Mission Designer places more Armor Abilities around the mission for the player to pick up as desired.

Most Elites will rage out (go berserk) and charge the player for a melee strike with their current weapon. However, there is one variant of Elites that will pull out their Energy Sword when they go berserk...

Higher difficulties just mean that the AI are fooled less easily by Active Camo and Holograms, but they are still fooled. We wanted the player to always be able to fool the AI in the right situations; otherwise the player will feel like the Armor Abilities are not useful. But we also want the AI to still try to outsmart the player, even when they use Armor Abilities.

Sword clang is something only players can do to other players. The melee system for AI is completely different, and we did not devote the resources to making clang work for them as well.

SonofMacPhisto asks, "Can I finally trust an NPC to drive the 'Hog? Or is it still best just to keep them in the back? How's their decision making regarding targeting when you give them a power weapon and tell them to hop in the passenger seat?"

Chris: The AI do a pretty good job at driving the player around. A couple of the testers played through some of the vehicle missions on legendary as the gunner, and they were able to finish them pretty handily. It is still likely easier to play the game (especially on harder difficulties) as the driver, but the AI can drive if you have the itch.

AI determine the targets they want to shoot at based on a series of factors. 1) Proximity to the target (closer targets have higher priority), 2) what kind of weapon they are carrying or vehicle they are driving (more dangerous weapons/vehicles are higher priority), 3) if the target is currently targeting the shooting AI (enemies that are targeting the AI have a higher priority), and 4) is the player currently looking at the target (enemies that the player is looking at have a higher priority than enemies off screen). All of that said, there are definitely going to be situations where the AI gunners shoots at different target than you wanted him to, but it feels pretty good right now (so hopefully that is rare).

serpx asks, "What do you do to ensure that an area of the campaign you design is "fun?" Do you have any philosophies that you follow? Do you just play and tweak around till you find things interesting/fun to play?"

Chris: A huge part of what makes Reach fun is that we followed all the work that had been done in previous Halo games. There is definitely a winning formula in all of those games, and we just needed to find that magic, tweak around the edges, and create the new experiences that we wanted for Reach. Some over all philosophies are:

1) People like challenges, but they want to have a clear idea why they succeed or fail and have some reasonable idea what to change about their approach to turn a failure into a success
2) People like variety. Both in what they do and why they do it (gameplay and fiction)
3) People enjoy exploring but are not okay feeling lost.
4) People are also nostalgic, and like to be reminded of their favorite moments of previous Halo games (as long as those moments are not simply copied)
5) Lots of projectiles that you can dodge/avoid are more fun than just a few that you cannot

a. Speed of projectiles is directly related to how 'overwhelmed' a player feels (even more so than numbers of projectiles)

6) Players like AI that are generally predictable but occasionally surprise them

a. Elites go berserk and find cover when shields are down

i. But Elites also sometimes sneak up behind you when you least expect them

b. Grunts flee when leaders are killed near them

Another huge part of making things fun is iteration, iteration, iteration! So we just keep working on it until it feels right. Or until it get pulled from our cold dead hands.

RC Master asks,"How are Brutes differentiated from Elites from a gameplay perspective?"

Chris: Fictionally we view Elites as the main antagonists of the player in Reach, and are viewed as the primary forces for the Covenant. Brutes are the shock troops of the Covenant, and are sent in to do the jobs that the Elites view as beneath them.

Gameplay wise, Elites are definitely the more challenging of the two. Elites are faster, more likely to melee a player, avoid shots well, and are the equal to a Spartan. Elites are also solitary and don't generally make decisions in relation to what other Covenant are doing near them. Brutes are meant to fight in packs. A Brute is very dependent on the other members in his pack, and their behaviors are often triggered when members of their pack are killed or in trouble. We view three Brutes with Spike Rifles to be a similar threat level to the player as one Elite (who is given a lot of flexibility where he can go). The three Brutes can definitely put a lot more shots on the player than one Elite, and they would take a lot more damage together than the one Elite, but because the Elite can recharge his shields and is so mobile the challenge ends up being comparable.