Designing Gravelpit-style atriums

hadlock

L1: Registered
Jul 5, 2009
19
31
My 10 minute email to my buddy turned into an hour long email. I thought I'd share with everyone. I am by no means an expert or even a "published" mapper but I thought maybe this could spark some discussion.

Fishbus had mentioned earlier today that a lot of the maps aren't as creative as they could be and a lot of people are taking conventional designs and applying them over and over to make decent, but unoriginal maps. I'm not sure that's entirely a bad thing though; it's a good way to understand the game mechanics and make a fun and balanced map. There's a lot of work that goes into a map and it's frustrating to sink a lot of time into a design only to realize it's fundamentally flawed. For a first time mapper, "stealing from the greats" might be the way to cut your teeth. In theory, you should be able to model an atrium after one of these points below and plug it into the A or B point of gravelpit.

I'm sort of the opinion that gravelpit style maps, due to how they flow, as long as the connecting tunnels are done well, are modular in design. A&B have some general guidelines to follow, but you can sort of model them after your favorite points from almost any map, and plug . That's what I'm focusing on here today.but for the most part C is the "free play" you can use to try and tie the map theme together, since when you're defending it, it has only entrances and no exits.

ok so gravelpit style maps have three main features

1) only one entrance to the atrium for red
2) one main entrance for blue that is easily defensible by red, a primary flank used to take pressure off the main entrance, and a secondary flank to clean up any deadlock flaws in the atrium
3) three levels of vertical play, red and blu main entrances on middle level. vertical play is what makes maps interesting and allow for many more strategies (see also: badlands)
3a) indirect fire to key locations, typically from secondary flank



Badwater is intriguing since it's done at a right angle. There's a lot of options for doing a point on a right angle but nobody's really ran with it. Badwater has a contestable battlement on the attacker's side that starts off in the defender's possession but is key to winning the point; there's a much smaller battlement on the defender's side that is of no use to blue and hard to utilize for red.


in no particular order

cb_badwater.jpg


cp badwater

1) red comes around the corner, lots of cover
2) blues main entrance is easily defensible by a single sniper, primary flank goes up to battlements, secondary flank provides (3a) indirect fire to clean up red's defensive edge of that battlement
3) three levels: mid, battlements, and defense has a lower level offscreen behind that ramp


cp_beerbowl.jpg


cp beerbowl

i really like this one. it's more suited to replace gpit B however

1) red has three entrances, primary, an entrance to the battlement (left side of pic, longish distance to get there), but all the good vantage points are behind bulletproof chikenwire, and also a loooong flank that takes 3x as long to get to the point as primary (bottom of pic)
2) blue primary comes up that ramp, primary flank is through that door just to the right of the point, and secondary is a long trek to that balcony that everyone seems to forget about, but is easily countered by red's long flank and a sniper
3) three levels, the bottom of the ramp, the point, and red's battlements and blue's secondary flank


cp_downpour-pass.jpg


cp downpour pass

this is similar (modeled after?) to swiftwater B

1) red enters through the building on the left, but also has space for a semi-undefended base in the right building
2) blu comes down the canyon as primary, there is cover on either side of the canyon for blu, but both teams can spam indirect fire over the bridge, left flank is a quick trip to the battlements (bridge) or also to mess up red's base in the left house, secondary (right) flank is a long walk but makes it easy to clear out the right building and has some cover provided by the building
3) only two levels. maybe two and a half, it's an uphill battle through the canyon, but total rise of the ramp is less than 10 ft



cp_swiftwater.jpg


cp swiftwater

1) red enters close to, but from behind and downhill of the point. they have a superior defensive position, but stairs to get to the battlements (bridge overtop point) are long and on the other side of the canyon so once their defense collapses it's difficult to recover (much like gpit B). red does have some cover on the left in the form of the tower, that rock, and the shield to the upper right of the rock.
2) primary comes down the shoot into a canyon, pretty much a deathtrap. the cave makes an excellent pooling area however. left flank is designed to keep red from regaining their defensive position, right flank is to take out their defenses. lots of indirect fire opportunities from right flank to red's cover behind the rock, on the point
3) three obvious levels, bridge/right flank, mid/left flank/blue primary, lower/point/red spawn



cp_steel.jpg


cp steel

there is a lot going on here

1) red enters from top mid, also defensive platform with cover. both entrances take a long time to reach the point
2) blue has lots of entrances. primary arrives at a battlement with lots of cover. primary flank is a long walk and has a faster route than secondary flank, which is a loooooong walk to the atrium, and enters at the drainhole for demo spam.
2a) of note, there's a "two way door" on the bottom right, both teams can use this to their advantage
3) three obvious levels; red and blue flank enter at mid, blue has a higher battlement with sightlines to both the point and red's defensive platform, and the secondary flank enters at the bottom
3a) blue can lob indirect fire at the point all day long from their battlements, with plenty of ammo and health on the other side of that door



cp_coldfront.jpg


cp coldfront

just something neat. indirect fire really opens up a lot of possibilities for breaking stalemates

1/2/3) n/a
3a) this is unusual, but freight has it as well (pushing into 4th from the left ramp); it allows indirect blind fire, but it's also playerclipped off. it's a rarely used gameplay mechanic, but a really nice touch and very elegant imo.
 
Nov 14, 2009
1,257
378
Wow, this is actually a really nice read. I came in expecting trash, and this actually brought up some quite interesting points. Thanks!

PS: You should get this moved to the tutorials section: This is the questions section now.
 

lucky

¯\_(ツ)_/¯
May 25, 2009
583
145
PS: You should get this moved to the tutorials section: This is the questions section now.

This is the mapping discussion forum still, tutorials are just a sub-forum of it.
Not quite a forum exclusively for questions?
 

Jimmy

L420: High Member
Jul 6, 2009
421
228
While I've only skimmed this and will read later, what do you actually mean by "atrium"? I'm completely lost on what you're trying to show in some parts.

However, excellent analysis of the maps!

area, section, point, ect. atrium is basically a large space, usually indoors, but he is using it to say "a segment" or "a cap point + surrounding areas". Not a word we use here stick.
 

StickZer0

💙💙💃💙💙
aa
Nov 25, 2008
664
647
area, section, point, ect. atrium is basically a large space, usually indoors, but he is using it to say "a segment" or "a cap point + surrounding areas". Not a word we use here stick.

I understand in essence what an atrium is, but "1) only one entrance to the atrium for red" what atrium specifically are you referring to each time you say it?

Also, could you explain "I'm sort of the opinion that gravelpit style maps, due to how they flow, as long as the connecting tunnels are done well, are modular in design." what you mean by modular? thanks
 

hadlock

L1: Registered
Jul 5, 2009
19
31
Jimmy nailed it on the head.

I think my usage of the term "Atrium" comes from my Quake 1 background :) Atrium was a standard term for large open room designed for deathmatching. Custom Quake 1 DM maps were typically 2-4 atriums with 2-3 levels of vertical play, connected by short passageways and teleport doorways, except larger artiums in Q1 were only about the size of cp_Follower 2nd, and always inside. Gpit atriums are about 10x the size of Q1 DM atriums and have no roof, but they're not as open as say, the first points on Goldrush and Badwater.

Re: Modular, if you get out a sheet of paper and block out Gravelpit, it's a very simple design, four roughly equal sized boxes making up a square shape, each connected with short Z connector passageways; Blue spawn in one corner, connecting to A & B, and C/Red Spawn in the far corner. Since the map style isn't linear, the type of layout for A doesn't really effect whether or not B is capped any faster or slower. As a result, you can play around with the layout, testing out different geometries without it effecting the flow of the other atrium, which is what makes it modular. You could replace gpit C with cp_Furnace C and atriums A and B would play like gpit A and B, and atrium C would still play just like Furnace C.

As a counterpoint, if you were to try this point replacement on a 5 pt cp/push map; If you were to replace Granary mid with a badlands 2nd point spire, and then make the last point like Goldrush 3.3, you'd fundamentally change how the map works and have to make substantial changes to the map, respawn timing and what not in a (failed) attempt to balance the map back out (not to mention a lot of complaints about how badly you broke the map).

One of the hardest parts of building a playable map (especially if you've just picked up Hammer for the first time) is getting the layout, spacing and scale down, so if you use gpit as a template, you've already got those three ready to go. You just have to break out your paintbrush and get to the fun part, build some playable geometry in Atriums A and B, which you can model after any number of good capture points from other maps.
 

A Boojum Snark

Toraipoddodezain Mazahabado
aa
Nov 2, 2007
4,775
7,669
I'd like to note that what you have marked on Steel C as being a flank is totally not a flank. Those two paths from E put you in front of the defending team the same as the path from B. The only real flank here is coming through from D as a soldier/demoman.

Further, I think it is ill-advised to model other maps after any of Steel's points, because they are very uniquely designed for that game mode. Any of the first four points would be quite hard to take in a map with linear design where the defense wasn't being distracted by having to go defend E.
 

lana

Currently On: ?????
aa
Sep 28, 2009
3,075
2,778
You've forgotten to include Yukon's mid. I think it's one of the best designed battlefields in game today.
 

hadlock

L1: Registered
Jul 5, 2009
19
31
While Yukon mid is a lot of fun and well balanced, I'm not sure Yukon mid would make a very good atrium for a gpit style map. Mid points are designed to be contested back and forth; once the defense on Yukon mid collapses, it's not a catastrophic event for your team and one or two players can can still prevent the other team from capping the point while reinforcements arrive. The mid fight is basically team deathmatch and so the point is designed for that.

If you are going to model an atrium after a certain map you should probably look at PL or A/D CP points; typically as soon as blue breaks the defense (requiring teamwork), unless blue isn't paying attention, it's a catastrophic loss for red unless red was clever has built some teleporters in an odd location. PL or A/D cp points are designed differently, they have better defenses, but that's because once the territory is lost you can't gain it back, which is how gpit (and Steel) functions as well.
 

Ezekel

L11: Posh Member
Dec 16, 2008
818
245
Fishbus had mentioned earlier today that a lot of the maps aren't as creative as they could be and a lot of people are taking conventional designs and applying them over and over to make decent, but unoriginal maps.

- just outta curiousity, where did he mention this? cos it sounds like it may have been a blog entry, which might be fun to read.
 

hadlock

L1: Registered
Jul 5, 2009
19
31
Fishbus actually has a blog (http://skeebz.com/), but it hasn't been updated since before the first final of Freight :)

He brought this up in the SA Source SDK thread over in the games section

relevant part:

I think a lot of it is simply down to Valve buying maps off the community, it brings a lot of interest towards generating content, but it can also bring out those looking for a quick buck, pretty much copying tried and tested, enjoyable 5pt CP maps that are quite solid but feels like there's no love behind it. I think the amount - and quality - of the maps would be much different today if valve said "no monies, but we'll put awesome maps in the game" as it would bring out the real passionate people. I actually expected nothing in return for steel, I was super happy that it was just going IN TF2, nevermind getting buxx for it.

I don't disagree with Fishbus, but when you're first learning to use hammer/build maps, it's nice to design something that's playable and doesn't have any crushing flaws the first time. I think copying styles is a good way to study and learn TF2 map planning while working on your first map. No one's first novel turns out to be the Great American Classic, but it builds a firm foundation for what you can achieve later with your own creativity. Definitely branch out and try something new after your first map. Icarus and Fishbus are constantly innovating with their maps.