Hellraiser and Ramirez
Bravely going where no TF Newsie has gone before... Poking at the gray matter of the community is definitely not a thing for the light hearted!
— The PF Staff
Before you read this interview, please take a minute and read the excellent UKFN TF2 review that HORUS did after the ECTS in London a few months ago. He did a great job and you should read it before you read the interview because several things mentioned in the interview are related to HORUS' fine write up. Hellface (Matt) and Ramirez (David) are working cooperatively to produce maps for inclusion in TF2. They have a certain chemistry and work very well together, complimenting one another's efforts. This proves that TF2 is 100% TEAM oriented from the ground up...
Now, on with the interview!@~!!
Bundy:
First of all, I want to thank you both for granting me this interview. It is the
first time you have been interviewed since you started making maps for Half Life:
Team Fortress and I appreciate you being patient with me. Ok, you knew this was coming ;) Tell us a little about yourselves and what drew you to map making and how you got started.
David:
I've been interested in this sort of work (world building, etc) ever since I started playing with Legos at age 2 or so. The first game I remember building maps for was Bard's Tale, but the possibilities were pretty limited and that wasn't very fulfilling. I experimented with mapmaking for Doom 2 and Hexen, but I didn't get really hooked until Quake. When TF came along, I became really interested in how the map code allowed you to do almost anything you liked with the scoring system. After having played TF for few months maybe, I got tired of playing the same old CTF map, and decided I wanted to create my own scoring system. The map code let me make TF into the kind of game I wanted to play even more.
Matt:
Every mapmaker alive had Legos. If you didn't own legos or an erector set, give it up. I still have my legos, too. I toyed with Doom2 a tad, but the maps were far too complicated to build something that was far too generic and simple to be worth the time. I wanted to do something radically different, and it just wasn't possible. I did some very minor tinkering when Quake came out, but honestly I just didn't much care for Quake. It didn't have the drive and excitement of Doom. Then Team Fortress came along. I thought it was just a variation on CTF with teams. All the maps made it look like that. It was great fun, and there were some great maps, but I wanted more than that. I wanted a real battlefield where you had to capture land to succeed. I didn't think that was possible with TF until I ran across a page David had called "Urban Warfare" where he was contemplating a series of maps that took place in a realistic militarized city. That got my attention. It was the setting I wanted with the rules I wanted for a game I loved. I sent him a blind email telling him I liked the page and was looking forward to seeing his map, and he wrote back thanking me for being supportive and it grew from there. He asked if I'd like to join on, but I didn't know enough about Quake mapping. So I grabbed Worldcraft and tried my hand at mapmaking.
Bundy:
What do you think is needed in a good map?
David:
That's a pretty general question… *8-) I think the very most important thing is the basic fun-factor. What it all boils down to, pretty architecture and lighting aside, is that if the players have a good time and want to come back for more, it's a good map.
Matt:
I agree with David, but I also toss in a couple of things in I learned from TF1. It needs to be simple enough so you can jump in without ever playing it before and be useful to your team in only a few minutes. But it also needs to have enough aspects and options available so that experienced players can grow and learn new tricks for some time to come. But most of all, it has to rock.
David:
Ahh, well said. Much better answer, in fact. *8-)
Bundy:
Many of our readers may know that HellFace and Ramirez teamed up to make
some improvements to Canalzone. Would you like to tell me how that project came
about and how it led to you working for Valve/TFS on HL: Team Fortress?
David:
I first met Matt through email a short time before I released CanalZone. We started exchanging ideas through email every few weeks or months, and our ideas were always pretty compatible. Finally this spring, after I'd dropped myself out of the TF world for a few months, we got in touch with each other again, and Matt rekindled my interest in TF mapmaking. He talked me into working with him on an update to CanalZone by offering his help with the Map Code, which I'd gotten extremely rusty with. But the CZ ent update was also intended as a way to get me geared up again and into a mapmaking frame of mind. We'd talked about map ideas, and I told him about a map I'd periodically fiddled with called Afghanistan. He was enthusiastic about the idea, and together we reworked the basic concept considerably. As we started to plan things out more, we realized some of the things that we thought we could do in TF1 (but which
would be bastardly difficult to pull off), would probably be very easy in TF2. Last year, in the very, very early stages of TF2, I'd been briefly slated to work on the TF2 map team, but other more immediate jobs pulled me away (start-up companies which of course fell went bust - bah).
Anyway, as our ambitions for Afghanistan started to grow, we decided to approach TFS with the map proposal. Happily, they liked the idea and signed us up.
Matt:
When it came time to do the new CanalZone Entities, I felt like Darth Vader confronting Obi Wan. "When I left you, I was the student...". I would run into a problem area and ask "Why did you do this like you did?" And he'd respond, "I dunno, that's just how it worked for me."
Didn't take him too long to get up to speed again, though.
Bundy:
It is my understanding that Valve licensed both the Quake 1 and Quake 2
engines from iD software and they used a heavily modified combination of them for Half Life. Since this is the basis for the HL: Team Fortress engine, can you tell me how it has influenced your map mapping (easier, harder, etc.)? HORUS said the map code is a great deal more versatile than TF1 map code. I've heard that the full potential of TF1 map code hasn't been exploited yet...
David:
I certainly enjoy HL mapmaking more than Quake mapmaking, partially just because the end result looks so much nicer. *8-) I think I agree with the comment that TF1 map code hasn't been fully exploited. Even now I'm still terrifically impressed with it's flexibility, but its biggest problem was that all the logic was tied up and made more complicated by all of the entities you had to mess with to set up even a simple if/then process. TF2 map code allows mapmakers to essentially create their own custom entity code to do exactly what they want to do with a minimum of entities. For example (warning: semi-technical stuff), in TF1, I had to use 4 entities to make a single Command Point work, each of those with about a dozen fields which defined the logic of the thing. With TF2 map code, a Command Point could be (and is) made to work with a single, custom entity, defined through MapC just once and then placed easily throughout the map.
Matt:
TF1 map code came pretty close to it's full potential with the CanalZone update we did. That sucker is pretty impressive, if I do say so myself. As for adjusting to the new engine, the most difficult adjustment for me was the change in tone. Quake is very dark and large in nature. Lots of extreme shadows and lava and dark ominous castles at night. Half Life is much more realistic. The lighting is brighter and cleaner and the textures are more real. I have to bring the scale down to a realistic size or it feels like all the players are small children. This is a good thing. It matches up well with the more realistic play style and weapons of HL: TF.
Bundy:
Can you tell us a little more about Afghanistan?
David:
We're wary of going into specifics because, of course, the map is still in development. So instead of specifics, we'll tell you about the basic concept of the map, and how it should play.
Afghanistan, as Horus put it, is kinda the spiritual successor to CanalZone. I think though that Afghanistan is more than simply a Command Point map with twists; to me it's much more evolutionary than that makes it sound. The basic concept of the Command Point map did of course come from CanalZone. One of the things I've never been happy about with CanalZone was how games have a tendency to devolve into near constant base-spamming runs. I wanted to make a map that put the focus back into what I'd originally intended with CZ, fighting for territory represented by the command points.
What Matt and I came up with as a solution was the idea of a command point map without bases as everyone knows them. Instead, each team spawns in their own Hercules drop plane flying over the level, and reach the map by leaping out of the plane and parachuting down into the actual playing field. This means that base-spamming simply won't exist. It also, more importantly I think, means that control of the battlefield is going to be completely dynamic. In CZ, the bases are the natural centers of power, so the front line always shifts back and forth the same way. In Afghanistan, a team's center of power is what they choose to make it. When the map starts, neither team controls any actual territory; all they've got is their drop plane. This means each and every game will play out differently depending upon which CP's each team chooses to take first.
Oh, one very important thing about Afghanistan. CanalZone can be a very difficult map to find your way around, due to the fact that so many of the areas look and feel too similar. I've learned my lesson from that, and we're trying hard to make each area in Afghanistan distinct enough that people should be able to find their way around much more easily than in CZ.
Matt:
I want to reiterate what David just said. The individual Command Points in Afghanistan are instantly recognizable by name. Having two people working on it initially helped us devise a nice variety of realistic-feeling Afghani locations. When you add that to the compass and waypoints, Afghanistan becomes a very easy map to navigate. Of course, it can be a very dangerous place when you leave your base on a raid, and return to find that it isn't actually your base anymore...
In a dynamic map, you need to be very flexible and quick to change your course of action. This is a map where listening to your Commander will be a very good idea.
Bundy:
I'd like to ask about your maps' philosophy. It seems to me that I've seen a lot
of TF1 maps that were made to either impress other map makers, or were made to
suppress particular classes while promoting others. Without naming a particular map
or map maker, I would make reference to maps that deliberately make it hard to be,
for example, a sniper. How do you determine the overall playability of a map? How
do you ensure game and team balance?
David:
Prepare yourself for a rambling answer on this one. I know I've got my own set of mental criteria in the back of my mind as I develop a level, but truthfully, I don't often think consciously on them, so what I'll write now is probably more philosophical than something concrete and useful.
From the time I start work on a level, I've got a mental image of how I want the map to play. The core of any map is the actual gameplay. The architecture and the lighting and even the scoring system all serve to shape the basic gameplay experience you want the players to have. I tend to think up all sorts of cinematic situations that I'd like to actually play, and I try to build levels that allow those cinematic moments to happen on their own. As I work on a level, I'm constantly loading it up to simply run through the level, shooting imaginary enemies and letting my mind wander and envision all the various situations that might arise with what I've made. That's where a lot of the basic balancing issues get worked out, by simply running around the levels again and again and trying to picture what players might do and how the different classes might interact, things like "This would make a great sniper perch.. now I need something to balance it out a bit.", or "The approach to this CP is too open, I need to add some obstacles to keep it from being a total killing zone." (Sidenote: One of the coolest things about making CanalZone was when players started doing things in the level that I'd never realized could be done, such as jumping on top of another player as a stepping stone into the enemy base. Though I didn't like the base-spamming that resulted, I just loved that the trick was discovered.)
Matt:
And on the other side of the coin, there are at least 2 secret doors in CanalZone that almost nobody ever uses!
Watching how other people reinvent your maps is one of the greatest things about mapmaking. Both David and I believe this, so we're trying to make Afghanistan open ended and full of options and alternatives. A good example: NONE of the Command points are mirrored. They are only balanced in that the rewards and dangers of owning one are usually about the same value of owning any other, but that's it. Other than that, they are all different areas with different attack and defense strategies for each.
Bundy:
Can you tell us anything about other maps you're making for HL:TF?
David:
Afghanistan was the first map we signed on for, but we've got several others in the works as well. One of the concerns we're aware of is that not everyone wants to play the kind of large-scale, strategy-based map that Afghanistan is. We're also at work on several smaller, faster maps that should be well suited to both quick pick-up games and small LAN party play.
Matt:
HL:TF is going to try to have a nice variety of levels from very small to very large in size. Afghanistan will have a couple of sibling maps in a similar style, and we are also doing some fun things with the Team Fortress MapC. The new Coding language was too powerful and exciting to do just one map on, and having access to coders, modelers, and texture artists lets us do almost any map we can dream of. That's saying a lot considering which two guys are doing the dreaming.
I want to thank Matt and David (or is it David and Matt) for taking the time to do this interview. I want to thank TFS and Valve for allowing us to post it. You guys are the best!
|
RSS Feeds
