How to practice

From Liquipedia StarCraft Brood War Wiki

When learning StarCraft, what's more important? Mechanics? Strategical knowledge? These issues have been discussed to death, but there is one general consensus by now: In order to improve, you need to become mechanically flawless.

Developing a Practice Routine[edit]

Starting in 2008 foreigners have gone to Korea again, starting with the acquisition of (T)IdrA by first eSTRO, then CJ Entus, Tasteless as the commentator for GOM events, later Artosis as another shoutcaster and finally (P)NonY as a recruit for eSTRO.

The general doctrine about progamer practice was confirmed when foreigners made it back to Korea. Artosis writes about how he learned his TvZ BO:

The anti 3-Hatch Mutalisk build order given ... is an exact build order used by one of the best Terrans in the world. Every pro Terran knows this build and practices the hell out of it. Their training partner Zergs in the mean time do the same 3 hatchery mutalisk build over and over. By doing basically the same game OVER AND OVER AND OVER you will memorize it quite well and see the holes in your game. This basic play is the result of countless hours of progamers playing each other and finding the most robust and powerful builds and styles.

As you master this build order in TvZ you will have to learn how to adapt to various things different zergs do. That is the last thing you really need to learn. Because if you know this build inside out and can macro it control it know your timings and everything like that then you are just going to roll people over who do lesser builds.


The point of this: Mechanics are more important than any other aspect of the game currently. The game is getting more and more mapped out. You need to be able to follow that map.[1]

This view is by now the general consensus about how to improve.

Basic Steps for Improvement[edit]

This is an excerpt from "How to get better", a forum post[2] by the former professional Protoss Legionnaire. The wording has been retained for the most part, only minor edits have been made.

For D level players there is not much you need to think about. Just do the basics well. Steps 1-4 are the most important for all beginners.

  1. Load up a replay of yourself. Click on the Nexus/CC (harder with Z - but click on Hatchery and watch the Larvae / minerals) and just watch the Probe production. The Nexus should never stop building. It should always have 1 queued. Particularly early to mid game. If you need to ask 'when do I stop building them' you aren't at the stage in skill to ever stop!
  2. Hotkey buildings. more importantly, hotkey Production buildings. and learn the unit production keys.
  3. Watch your mineral count. If it gets high, utilize step 1 and 2!
  4. Scout. Scouting provides information, and information provides counters to builds. So many people build things just for the sake of doing it. Why waste lots of money building something you don't need, when you can build something you do need?

For beginners, just learn the 2-3 most common strategies that your opposition race uses. And scout for it. If you see 1 fact or 2, your basic build should switch over to a better counter. etc. The counter doesn't need to be 'huge' it just needs to take advantage of what you've learned. If you see a 2nd upgrade going at a fact, you can expect Vulture/mines soon! Get Goons! etc. etc.

  1. When you get an advantage, expand.
  2. Learn early game builds. Late game doesn't matter, early game is where it counts. This used to mean a lot more until everyone became replay whores, but the point still stands.
  3. Macro well. this is all hotkeys, mineral expenditure, and expanding.

Later on you will learn things like:

  • Scouting also involves knowing when the army is moving out. Then utilize the terrain to attack them.
  • Does it favor you to wait for them to move into open ground to attack, or move to high ground and keep them on the low ground.
  • Expand to areas that help this, an expansion in a certain location might take longer to get up and running, but for them to attack it is a lot harder.
  • Expand to other mains! Especially when they have ramps. You can build a few gates up there, then for the entire game you can do flanking attacks, while purely defending all game. (Defensive fights almost always has an advantage - of course you can do defensive attacking as well...). No one is going to attack up ramps against an army half its size, while letting the other half army flank them (or attack a now undefended base). Thus you can macro and abuse the terrain!

Maybe more advanced thinking like -

  • Is their starting position in the corner of the map (think 3 player maps like longinus with spots are in top left corner, mid right and bottom left.) Top left can only be 'attacked' from 1/4 of the total map area, while mid right can be attacked by 1/2. Thus if they are midright, harass is so much more powerful.
  • What does your opponent know? What can you make him think? If he has a bigger army but you have more of the map, you may need to buy time, keep your army outside his base to make him scared, if he moves out (think PvT) keep attacking/retreating, sieging tanks takes a long time, and you can buy 2 lots of production cycles from your bases just by not even doing anything but making them think you are attacking.
  • Buying time. One of my favorites. You may have sneaky expansions, but you have less units as a result. You know that if the game lasts just a bit longer you will have the same/more units while having more bases. But they are about to attack. So the key is to harass.

You see them get ready to move out. (i.e. they start building Vultures after massing Tanks, then you know in the next minute they will attack.) or if they start to move. Have a shuttle with a reaver waiting out of their vision, use it to attack when they move out. Or attack their expansion with some units. They will stop their attack and focus on defense, buying you time. The important thing with these attacks is NOT to lose your 'attacking force' easily - remember, you are buying time, if they are about to be killed, run them to the corner. A few more seconds of them chasing your last 2 goons to the edge of the map and then having to retrace their steps can be critical. This is why I liked Reaver drops so much, you kill a few units, run it away, then as they get ready to move out again, you drop it back in. You can hold them up for minutes. The object is not to do critical damage (unless the opportunity arises) so you shouldn't do risky 'do or die' attacks on the mineral line defended by anti air stuff etc. If you can only attack a crappy building, then attack it! He will still be just as scared as if you have gone after the mineral line. (buying time is also for waiting for an upgrade to kick in etc.)

  • Do the map starting locations favor certain builds? Some have more open chokes, forcing them to do different openings. Thus you have better openings to take advantage of it.

If you do steps 1 to 4, and keep doing it, you will become a C rank player. Even for good players. Load up a rep of yourself and do step 1, I bet you don't do it perfect, I know I didn't.

Another bit of advice. Never be afraid to lose. Try and go out of your comfort zone whenever you can. Losing if utilized correctly is a lesson to be learned. I know its a good ego trip to win 80% of the time vs your friends or to be the best player in a clan or channel. But you aren't learning anything that can make yourself better.

These are all quite general thinking tips for improvement and can be applied across other games ... sc2 comes to mind.

Resources to Improve your Understanding of StarCraft[edit]

Know where to find games[edit]

One of the most important things about practice is being able to find game in a friendly and constructive environment. In recent times, this has been offered by the iccup Server. The server has an inbuilt feature of offering LAN latency settings, therefore allowing for better control.


Overall, work on your basic mechanics. Once those are done, you can start doing fancy stuff and adding more exquisite move to your builds. Being fast is not about being able to move your hands. It is about knowing what to do in every given situation and about building good habits. One of the biggest reasons why Korean progamers are so much faster than foreigners is because they know just what they are going to do.[1]