[ES] Build Orders

This is the eighth entry of The Elementary Series.

Everyone talks about build orders, but many misunderstood what that means.

I put The Elementary Series on hold for a long time, because there are more interesting topics most of the time. I should get the ball rolling again.

What is a build order?


We mention “build order”, “opening”, and “build” a lot. What are they? Are they different? These terms are often used interchangeably, and there is no agreement in the definitions of these terms. What I write here are merely my definitions.

A build order is the notation of what you build. It states what and when is built. It can be compared to chess notation, except you only record one player since StarCraft is not a turn-based game.

An opening is the set of precise notation you use at the very start of the game (e.g., 16 Refinery Reaper expand or Command Centre first), and the opening choice is decided before the game starts (i.e., independent to what opponent does).

A build is a broad description of the strategic choice in the early game (e.g., Reactor Hellions with Battlecruisers). This could change as you adapt to what the opponent is doing. For example, you may plan to do a Widow Mine drop build, but you change it to a Cyclone heavy push when you know the opponent picks the Stargate tech path.


Why do I emphasize these terms are conceptually distinct? Differentiating opening from build allows us to discuss build orders with higher clarity. We have a tendency of over-scrutinising and over-comparing build orders when they aren’t that different. Conversely, people overlook some important differences in seemingly comparable build orders.

For instance, Command Centre first and Reaper fast expand are two different openings (not builds), but they can shape up to the same build. Both can easily get the 1-1-1 set up and go for Hellions and Banshees. Calling them different builds would imply that they have different strategic aim, but the difference is the risk and benefit trade off in regards to economy at the start but not the strategy per se.

There are times when people incorrectly group two different builds as the same build. For example, in TvZ, Terran can go for Reactor Hellion with Cloak Banshee (double Refinery) or sacrifice cloak for an earlier third Command Centre. Both converge to three Command Centres with Reactor Hellions and Banshee. However, the former has the potential to deal harassment damage with Cloak Banshees, while the latter aims to be defensive with the Banshee(s) to defend against Roaches. Even though both builds can derive from the same Reaper fast expand opening, the two differ in their builds and strategic goals.


There are a few major misconceptions about build orders. First, players mix up builds with compositions. Composition refers to the set of units that you aim to have in certain time frame of the game. For example, Zerg can have a composition of Zerglings, Banelings, and Mutalisks. There is no such thing as a Zerglings, Banelings, and Mutalisks build. I often come across players asking for mech builds on forums. While some builds transition to a mech composition smoother than others, you can build a mech army with any opening or build. The composition is generally decided by the production structures, so you can build a mech composition simply by adding more Factory.

Second, players overestimate the impact of builds on the outcome. Just to be clear, I’m not saying builds make no difference. I merely want to point out that players seek different builds more often than they should after losses. Pro players use the same build over and over again. Although there could be some minor metagame advantages behind certain popular builds, players can benefit more from making adjustments than to use a completely different build.

Third, many think build interaction as rock-paper-scissors, whereby each player shows their build choice and sees who has the upper hand. Oh, you went for Widow Mine drop, Liberator, and Tank push three-pronged attack? Then you have a build order loss to a Dark Templar build. It does not work like that, because you can always make adjustments to a build to change it to another build for the better. That is why I promote the idea of having a framework with a flexible build (for example).

Broader concepts

I have developed two widely applicable concepts based on the conceptualisation and misconceptions I have discussed thus far.

The concept of convergent points

A convergent point refers to a benchmark in terms of the number of production buildings you should have based on the number of mining bases you own in order to have an optimum production. If you compare many professional games of different players in the same match up at the say 10:00 point (a random time stamp), you will notice that they have more or less the same number and type of production structure (assuming they use the same type of composition and nothing drastic happened). This is because that production building set up (both quantity and type) is considered the most efficient to spend your resources based on the number of mining bases you have. That is, regardless of the different openings and builds you use, you should always converge to the same production set up. For example, if you are going for bio, the two key convergent points are 3-1-1 and 5-1-1.

There are several important practical applications. At the basic level, it provides a guide for how many production buildings should be added based on the number of mining bases. If you have more buildings than the convergent point benchmark, not all buildings are producing. If you have fewer buildings than the convergent point benchmark, resources will start to bank up. Therefore, by knowing what the convergent points are, you have the optimum set up to spend the resources. The other important implication is to provide a check point for which converge you converges your build order towards, and this explains the name “convergent point”.

You can read about this concept in the original article HERE.

The concept of build order blocks

The concept emphasizes on breaking a build order into multiple small blocks. Every build can be broadly broken down into three main blocks:

  1. Opening
  2. Build
  3. Composition

This concept assumes sequentiality. All builds start from the same known set of openings. That is, the opening block is always before the build block, and so on.

The central idea behind the build order blocks concept is that you can plug-and-replace by switching a build block with another. As mentioned earlier, the same opening can diverge to different builds. This means, if you have learned a build order that uses a specific opening (e.g., Reaper expand), you can transfer that knowledge to another build that uses the same opening. This can be done by comparing the build blocks of the different builds and figuring out their different structures used in their respective builds. Subsequently, you focus on these specific differences as you learn a new build, while you keep everything else the same. This eases the learning curve of adopting a new build.

You can read more about this concept in the original article HERE.

If you enjoyed this article, I’d love you to share it with one friend. You can follow me on Twitter and Facebook. If you really like my work, you can help to sustain the site by contributing via PayPal and Patreon. You can also support me and enjoy quality tea with a 15% discount at AFKTea by using the “TERRAN” code. See you in the next article!

What do you think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s