For race verses race strategies, the "meta-game" changes with time. I can understand that. But often when watching VODs with announcers they say "oh! that is a good strategy for this map" or perhaps the opposite. This leads me to believe that some maps are good for some strategies and bad for other strategies and I would like to know why that is so that I can best choose and create my own strategies.
What I'm trying to figure out is what features make a map better for one approach over another. The only example I have is that some maps have a good area for drops while others do not.
To narrow this down let's pick the current (Dec 2013) 1v1 ladder maps, specifically Frost LE and Derelict Watcher TE. I play Terran mostly but it would be nice to also know what I am likely going to be facing from Protoss and Zerg. What are the most effective strategies for Frost LE and Derelict Watcher TE and why? Conversely, what are the least effective strategies for Frost LE and Derelict Watcher TE and why?
Best Answer
This is a complex issue, and changes constantly.
Basically this comes down to the idea of "how can I exploit the map to win?", but also (and less obviously) "how is my opponent likely to exploit this map to win, and how can I mitigate this"
The difficulty with picking "good" terran maps, is that often the things that make a map good for TvP make it worse for TvZ and vice-versa. In TvZ you want short rush distances and small chokes, but in TvP you generally want the opposite. Derelect Watcher and Frost are pretty generic maps, neither of them are too heavily exploitable, but I'll give you some ideas anyhow:
On the first note: consider Derelect Watcher:
You can see on derelect watcher the 4th base is either far in a corner, or beside the main. This opens up the opportunity for you, as a terran player, to use drops more effectively. If he expands in the far corner, when you drop there his army will not make it there before you cause considerable damage, and if he does send too many units to defend the drop, his army is out of position and weaker to an attack in the front. This is especially difficult for zerg where remote warp-ins are not an option. If he takes the base beside his main, the distance between his third and 4th is still fairly long, and if he defends a drop at his 4th, you can pick-up and go into the main, which is now miles away for his ground-army. As a terran TYPICALLY you don't choose much between various strategies. In TvT you probably have the most flexibility.
Derelect watcher has a wide choke and harder to defend third making bio aggression without ghosts/vikings more effective. On frost however, the existence of smaller chokes mean you won't get as far with tons of bio straight up the front. Frost does, however, have separate entrances to the third and nat, so you can make double-pronged attacks with more effectiveness.
(TvZ) A really fast gas-first factory build could be effective here, as he won't be able to wall out hellions with evo chambers and spines as easily. Or a blue-flame opener.
Frost LE:
Frost has two possible thirds: One is extremely exposed (expanding up/down) or has a ledge behind it with open access in the back (in the case of expanding left/right). As a terran you can exploit the ledge with widowmines or a few marines, without having to risk losing a medivac for more-expensive drop play.
Distance between the third and the natural (PvT) is fairly long, with the possibility of hopping into the main from the third base. This makes dual drop play slightly harder to defend. The protoss can't simply have his whole army at the third and keep a few stalkers in the main like on derelect watcher.
You can open with a proxy 2rax on any 2 player map.
Neither map is great for mech, so avoid using mech as the third bases are distanced from the natural. A notoriously good mech map would be something like akilon wastes, with an easily defendable third and close 4th base. Neither map is great for banshees, as you have no "safe-zone" between the natural and the third base to fly between or hide in. A map like Alterzim Stronghold TE or Polar Night would be more ideal for this style.
Second note:
Derelect watcher For reasons described above, your opponent might be more likely to all-in on 2 or 3 bases to avoid having to take a 4th base.
Any 2 player map has a much higher likelyhood of a proxy play. Right now with the state of oracles, I'd say that you should be ready for them regardless of what map you're playing on, but the oracle can be up to 20 seconds faster on a 2 player map. Any standard terran build will account for this anyways, so you don't have to do much other than be in position for it. Using the initial reaper to scout proxy locations is more important on 2 player maps, where there is a higher chance of proxies and a higher chance of finding any proxies.
Frost:
As terran, you'll likely want to take the up/down third vs P and the left/right base vs Z. This is because you want the open area for spreading out against colossus and storm, but the smaller choke against zerglings or banelings. This opens you up to a chargelot/archon 2/2 timing though, as you don't have anything close to a choke to use. Putting a few safety bunkers would be advisable here, or be ready to evacuate the third and defend by your natural.
Frost is more likely to see blink plays from a protoss than derelect watcher, simply because of the amount of extra area to blink into the main base with. Some other great blink maps are ones like Yeonsu (with the huge natural exposure) or star station TE.
2 base colossus all-ins are harder to hold on Frost where you have a small choke that is easily controlled with forcefields, so vikings will probably be more necessary than on derelect, for instance.
Last but not least ... Finally I'll mention the meta-gaming that happens between two good players. Consider PvP as a good example. If a map is good for blink play, you might take an educated guess and go for a 1base immortal build to counter it. Or if a 2 player map is likely to see proxy oracles, you could do a blind blink-stalker build to try to grab an early upper-hand, or DTs to punish his lack of detection. In ZvP we often see zergs 6/7/8 pool on maps that are good for forge-expands to try to catch a free win.
This often leads experienced players into playing unlikely "bad builds" for a given map, because they're expecting the opponent to play a certain way.