Terraining Process

Overall process: When starting with your map terrain, the first thing you'll want clearly in mind is your intended goals. Most people think they can just wing it and 'feel' their way through to having a good map. This method leaves a lot of your success up to chance, however; even if you're very experienced. And even without experience, a good plan on hand before you start can get you quite a ways. Know what you want. What tileset do you want to use? Where are all your major, distinct terrain elements going to be located? What are they and how will they look? How much space do you estimate you'll need for each? Will these fit into your map dimensions? The better you can answer these questions (and any others you can think of), the better off you'll be later down the road. Even if your initial answers turn out inaccurate or even outright wrong.

And expect to be wrong a lot. That's what making drafts is for. You're rarely going to nail your intended design on your first draft. You'll want to play around with just plain isometric terrain in your editor until the blunt topography is hammered out. Don't bother making it look pretty. You're generally going to want to get it looking right from the minimap view first before bothering with the closeup details... and definitely before you begin any custom blending. It's just like sketching, really. You draw your guide lines first, and then build your grand image around those as you go. Don't be afraid to throw away or restart your drafts if you're not satisfied, either.

Development: When composing your terrain draft(s), also like sketching, you're going to want to work in layers. I like to start with the gist of the map topography so as to get an idea of spacial requirements and visualize where things will go. Then I focus on only the most 'important' parts of the map (the stuff I can't function without having at least in basic form), and go no further than what I need for as long as I can get away with. Again, the more planning you can hammer out of the way in advance, the better. But even still, don't act on your plans until you absolutely need to.

A lot of people make the mistake of first trying to make their map's terrain as complete and as 'pretty' as possible and then fitting all the important functionality and gameplay considerations after. What you really want to do is tolerate your map looking like a pile of garbage for the first few stages of its development, and when you start feeling comfortable with the design with regards to gameplay, start solidifying and polishing it then. Often, you may only get to know whether your designs turn out to meet to your needs only when you get start going into the details. This is why it's important to approach your specifics only as you need them. You also want to do this so you'll have less to lose from tossing out unsatisfactory designs; something you should never corner yourself into being afraid of doing, if you can avoid it.

And remember: Save your blends for last. Or at least until you absolutely need them. The more you tear up the isometric structures of your map, the more you'll regret it if (and, surely, when) you have to go back and change something. You don't want to become the slave of the copy and paste tool before you're ready.

Gameplay considerations: A few specific things you'll want to consider with your terrain plans are as follows:

1. What types of units will be using your terrain? If your map involves lots of aerial battles, you probably have more room for stylistic designs and aesthetics. Ground units, of course, constrain your stylistic freedoms somewhat with their spacial dependencies. Ranged fighters can make use of cover and high ground, whereas masses of melee fighters really only need be concerned about space availability. Etc.
2. How will units be using your terrain? For an RPG, some amount of exploration value and novelty focus can be afforded. Other game types may require more strict balancing of your terrain shape for player vs player gameplay. How tightly does the tension of your gameplay depend upon the topography of your map? Plan to invest in terrain refining accordingly.
3. How many units will be using your terrain? Is there enough space? Or too much, even? Remember Starcraft has a unit limit of 1700. You shouldn't need to abuse that if you're doing things right. Same goes with other limits, such as sprites (which max out at 499). The less you abuse the game's limits, the better off you'll be. This wiki might help you out when trying to remember Starcraft's limits.
4. Can units get around? Remember also that Starcraft uses cheap, outdated pathfinding algorithms, and as such, overcomplicated or non-intuitive terrain layouts can create a lot of hassle for players. Test out your designs. And be wary of abusing terrain elevations: Your nifty custom-blended terrain might look like a museum masterpiece, but that poor, dumb Dragoon that can't make it 1/4 of the way across your map (or the player who needs said dragoon, NOW) might not think so.
5. Are you doing a lot of things with triggers? Remember that the 'Create unit' trigger action requires empty space to place units, lest you get annoying unplaceable errors. The 'Move unit' action, too, depends on adaquate space, but will fail without even a peep if it doesn't have it. Have fun debugging that shit.

Moral: Don't let it be a problem, newb.

All in all, what you really need more than anything else for good terrain design is simple: Do not to be lazy. Patience and tenacity are key. There's just no way around it. Starcraft isn't as good as it could be with regards to instant gratification with map making, sure , but if you're willing to stick with it, you'll find your rewards sooner or later.

Credits

Thanks to Tuxedo-Templar for writing this in November 2009.