In general[Spoiler]
- Ringing with water and/or cliffs is better than straight mapcutting, as it allows for more interesting tactics from the fringe.
- Puddles and cliffs that offer small situational advantages reward adaptation and make for interesting play
- Maps should be pretty if possible, I personally like the sedentary layer effect you get on maps like titan and titan duel. Trees and grass can be nice too, though I tend to prefer this skin in other games where the features don't get in the way.
- Map should not hide information from player. Water and pathing should be obvious. Features that give metal or energy should be obvious.
- An easily secured initial base, with increasingly difficult and varying expansion options is a pretty common formula for success.
- Areas of high contention (typically mega mex and reclaim) should not be singular, as this forces games to play out the same way every time with a 'king of the hill' style land grab. It's a nice dynamic, but it shouldn't be overdone. If there are multiple points of contention, there are options. I personally prefer 3+ if any, and prefer reclaim stacks to mega mex.
- Features should try their best not to get in the way.
- Map should either seek to best represent a specific condition and thus become 'classic' (CCR, icy run), or should cater to a variety of factories by mixing up terrain.
- Maps should be designed around a particular start box. This isn't a hard set rule, but it allows the map maker more control over what sort of gameplay they're trying to encourage.
- Map should have deliberate intentions about what sort of movement traverses where. If you make cliffs and you don't know whether a pyro can make it over, you're leaving your design up to chance.
- Symmetry is safe, and will appeal to hyper competitive types. Asymmetry is stylish, and breaks the mirror inferencing that tends to happen at the competitive level. It is however almost always inherently unfair on someone. I personally would like to see more asymmetrical maps, but understand they're sort of hit and miss.
- Reclaim features are good for high level play, but bad for low level play. It complicates the economy and rewards innovation, forethought and economic mastery. These things are highly variable at lower elos and thus make for lopsided games.
(the rest are shorter I promise)
1v1[Spoiler]
Need a lot of raiding paths to keep it interesting, as a large draw of 1v1 is the multifrontal fighting. Economy should be scaled such that it's not so low cheesing becomes standard, but not so high that there's no meaningful early game. I personally prefer mex to be in clusters, as this rewards a more structured expansion plan, area mexing is sort of afk eco with no decision making.
Small Teams:[Spoiler]
I think Onyx couldron is an example of a great 2v2 map, giving each partner a side to take advantage of, with several areas of interaction between sides (lakes, flats, and cliffs) which spice up decision making.
3v3 and 4v4 tend to be fairly similar, though they play differently. 3v3 planes/GS are not necessarily mandatory, in 4v4 they are. I'm not sure how to convert that into a principle which could be applied to create meaningfully specialised 3v3 or 4v4 maps though. Corner starts are usually good, as instaplopping fac and slamming out raiders isn't necessarily the best move on those maps.
Big Teams:[Spoiler]
Big teams need multiple terrain types and points of contention. Big teams are boring if everyone is the same few factories, and they slow down a lot of there aren't serious rewards to battle for scattered across the map.
Should be deliberate about whether you've got obvious spots for air players or not.
Maps need to be a bit more open, since porcing otherwise gets out of hand and makes for unfun games.
FFA:[Spoiler]
Depends on the players. For players that just want to build up and simcity, they want big chokes and convenient static locations. For competitive FFA players (huehuehue), there needs to be neutral territory that rewards expansion. Personally I consider the map a failure if people can win semi-consistently without expanding or attacking.
Extra points if you introduce interesting mechanics to shuffle traditional FFA diplomacy, lava rising being a great example.
Chickens: [Spoiler]
I don't really play much chickens. I assume a successful chickens map would be one that gave unusual options or challenges. I think usually people prefer the ones with the best choke points, then scale up difficulty by increasing the number of chickens they verse. I don't think this is a very high priority, since game-play is homogeneous regardless.