5/15/17: Added a section on Dead-Ends and a section on Play-Styles.
I've been wanting to do this for awhile, but I haven't ever really got around to it. Here, I'll be writing a guide on how to build a classic style installation mission like you see in brood war within the starcraft 2 editor. I'm not talking about importing maps from brood war, I'm talking about adapting how you build starcraft 2 maps to mirror the same style of gameplay as you see in classic installation maps.
Where this started..
So before I go into the guide, I'd like to briefly explain how I got here. About 2 years ago, I decided to push towards researching what made installation missions in brood war so great. Even today, I still believe the mission "Patriot's Blood" from brood war is still the best installation mission ever made. In my opinion, it's the best campaign mission from starcraft 1 as a whole that was ever made. That's my opinion, of course, but that's where my drive came from. For about 2 weeks, I conducted intense research on brood war installation missions. I played the maps, watched videos of others playing the maps, and I built my own missions in both brood war and starcraft 2 to try and find what made installation missions click. Over time, I've put what I've learned to use. You can see some of my work here: Hybrid Campaign 01: Awakening. If you're okay with experimental work that never got polished up, I've also got this: Legacy of Liberty Campaign. I've also got some more installation missions in the works, but they are not released as of yet.
The Main Idea..
The overall goal of installation mission design is to build an "Event-driven, Tactical Mission." This means that you want various events throughout a mission to drive the player forward and keeps things fresh with every room you enter. The tactical part just means you need to make sure that players aren't just mindlessly a-moving through your map. It's okay to have an a-move game where your player is busy macro-ing in their base most of the game as it makes things just a little bit easier when you don't have to also worry about micro. However, this is a micro mission, most fights should require some amount of decision making. For the remainder of this guide, I will be using a demo map I built beforehand to help show what I mean.
Yes, events. When you enter a room and you get ambushed from behind, that's an event. Whenever the player uses a terminal which reveals an area of the map, that's an event. Events keep the player engaged at all times with constant feedback and allows for small decisions to be made. This works out to be a fast-paced level with many twists and turns along the way. This is what I found so appealing about "Patriot's Blood". Here's a quick video to help demonstrate a good event-driven sequence.
Speaking of "Drive"..
There's also something else that I wanted to note about "drive". This is something not immediately obvious in the demo map, but is present. In any good installation map, it's important to think about the pace that your mission will be played. In the demo, it's about an average pace, nothing drives the player forward such as a timer based objective, but there's also nothing holding the player back for very long. If we were playing zerg and they regenerated at normal speed, that would significantly slow the pace of the mission as players might feel forced to sit and wait for the life of their units to slowly recuperate. As a result, I often increase the default healing rate of zerg or recharge rate of protoss when the player takes control of one of those factions during an installation mission. Another way to handle it is to remove regeneration completely. This can sometimes make missions really brutal as taking too much damage in one fight could lead to a loss later in the level and I recommend being careful if you decide to go that route (you could also have checkpoints where your units are healed to full, some sc2 campaign missions do that). Another aspect of "Drive" is having something to rush players forward. You could go simple and just have a timer for how long you have to complete the mission. Alternatively, you could implement something more dynamic, such as a boss chasing you down like in the WOL secret mission. I went for the fastest map I could build which meant that I didn't add any additional drive, I just made sure the player had medics for healing.
Here I just wanted to go over a topic that I think some will find useful. Another word for play-style in the context that I'm using is "pace", though that might be less accurate than play-style. Each installation is likely going to fall into one of 3 categories (it's possible that I missed a category I have yet to discover). First is standard, which is the style that I've gone for in my demo mission and it is the most common way to build an installation mission, I think. In the standard style, the player commands a small squad of units that can recover between fights. I use the standard style as my "go-to" if I'm looking to do something more like the original starcraft 1 installation missions. Next is what I call the swarm style. This is where the player is given lots of expendable units and lots of expendable reinforcements, such as zerglings, and has to play around enemy units that crush them. In this zergling example, your player might face siege tanks, firebats, and medics which all give the player a target to try and rush as early as they can when the fight starts. This style lends itself to timers really well. The final style is steady style, which is literally the "slow and steady" style. In here, your goal is to try and slow down the player and force them to take the game much slower, make them think about fights before they start. Ways to do this is to either build it as a stealth mission or just remove all forms of recovery, which makes players have to cycle between healthy units in fights. When building your installation mission, I recommend figuring out what kind of style you want for the level, then focus on it using one of these categories.
Dead Ends in level design...
At first, I didn't really think much on dead ends, but it was brought up. It turned out, after thinking about it, that there's a lot more to dead ends than I originally thought and I'll elaborate on that. From what I've seen, there are two types of dead ends, there are obvious dead ends and there are hidden dead ends. First off, a dead end is considered a no-no in level design as it's mostly just wasteful, both wasting space on the map and wasting the player's time exploring the dead end. There are a few exceptions, such as rewarding the player for reaching the end or if it's for immersion/story reasons. A dead end for the sake of being a dead end is not fun or interesting. However, in an installation mission, we're obviously making a custom map with triggers, which gives us the freedom to put something valuable at the end to reward the player for exploring. Now, with obvious dead ends, the player will almost immediately spot that the path ahead does not progress the mission and will know they can turn back immediately, these tend to be the better kind of dead end. You can make a dead end obvious if you reveal that the path ahead is a dead end and perhaps even reveal if there are any rewards if the player goes that way. This works well if the dead end is too far to tell it's a dead end. The other type of dead end is a hidden dead end which is more difficult to use well. This type of dead end is too long to see if it's a dead end and it also doesn't reveal to the player that it is a dead end. Hidden dead ends are fine if your installation mission is slower and focuses on exploration and mystery, but it doesn't work well in an installation mission that's fast-paced as the player is focused on reaching the end rather than exploring. The last thing I wanted to note on dead ends is the walk back. When you have a dead end and the player has to turn back to be on the main path to progress, all of the enemies along the way will already be dead. This means you may have to spawn some kind of opposition in their path if the way back is especially far or you want to keep the game fast-paced. Otherwise, you can let the player have a chance to relax for a minute. One final thing I wanted to note on dead ends is sometimes they can be temporary. This means the dead end has nothing in it one time the player goes there, but it might do something later. For some examples, perhaps the dead end is just a locked door that you have to open later, perhaps enemies spawn from that location and move out of the dead end path to attack the player, or perhaps reinforcements or some other reward spawns for the player there later. One idea is perhaps the player does something on the map and they come back to find a reward, a devious secret.
There are some other small things you can do that would help make your installation mission feel more like the classic ones from brood war. For terrain, you can keep the cliff levels at bottom and mid cliff heights and reserve the peak cliff height for unreachable terrain, much like in brood war. Here's some screenshots to help show what I mean.
Another key to success, I think, is doors. Doors, doors, and more doors. Not sure what it is, but it feels fun to run around in an interior station and open doors throughout the level, whether it's automatic or if you have to use a terminal. Normally, I'd say to stay away from unnecessary pauses during the game, but the immersion of opening doors is just too appealing to pass up. I've likely missed something, but there are bound to be other small things that you may be able to get out of the demo if you try it out yourself.
I hope this has been an insightful guide on how to build installation missions in sc2. Let me know what you think and please point out any criticisms you may have, I'd be interested in reading it regardless. Here's a link to the demo: Installation Demo. Enjoy it and I hope you learn something.
The team is actually still working on the project, though only little bits and pieces at a time. Trying to get something done when everybody has their own side projects makes things pretty slow. For example, mission 2 is 99% done and we'll bring it to you guys asap, but we need that last 1 percent, haha.