EPTSo, I wanted to present an improved mindset on build orders, because based on a lot of the threads I've read on these forums, there are many users who talk about build orders in a very limiting way, in my opinion. Shall we?
What is A Build Order?
This seems a logical place to start. In it's most simple form a build order is nothing more than a list of instructions of when to build things based on the "food" (supply, psi, or control; depending on your race) count. If you cannot read build orders, I highly recommended this article in liquipedia to catch you up.
http://wiki.teamliquid.net/starcraft2/How_to_read_Build_Orders
How Do You Discuss/Think about Build Orders
This is where most people take a wrong turn. A lot of people discuss/think about builds as that list of instructions. Why is this bad? Because all this list does is tell you when to place building or train units, it doesn't tell you why they're being built. A build order is the very last product of a long process of refinement, trial and error, success and failure.
A player will come into the forums, find a build order and start using it, but never really understand it. They're following a list, and that's about it. The real problem with this is that without really understanding the build, it is tremendously hard to adapt to changing game dynamics to effectively continue on with their build, so a player will get frustrated and either change his strategy mid-game, or just scrap that build and discard it as inferior. Even if they do end up proceeding down the build, there are so many replays I've watched where the player had no idea what to do when he got there. They're not positioning units correctly to fit the goal of the build
Given that, I would like to break down the Build Order to get to the heart of what it really is.
The Build Order: Step 1: The Idea
I'm going to go ahead and steal a definition from Day9 here. "A build order is an optimization of an idea." In essence, a build is just a way to most effectively get to some end goal. There are a wide verity of builds formed from all kinds of ideas, from something as simple as "I bet if I get a starport with a techlab really early and research cloak, I can do a lot of damage to a player's economy" to something as complex as "I want to get X unit composition by Y time at Z point on the map to deny the Protoss's third base so he won't have enough economy to survive a timing push". It's always critical to realize the idea behind a build order so you know what you're working toward.
The Build Order: Step 2: The Components
So, once an idea is established, the next step is going about bringing it to fruition. There are several steps you are going to take to get to this end goal, and you need to realize what the important parts are. These components come in the form of workers, bases, structures, upgrades, certain units, etc. Identify these key components and keep them in mind.
The Build Order: Step 3: Arrangement
Now that you've identified what parts in the build were the critical components and laid them out, now it's time to consider the arrangement of each component. Why is this upgrade or unit placed here? Is there a logical reason? Why would it be good to be placed in this order? Consider the implications of arranging the components in this way. How does this help us accomplish our idea? Here is where you would consider augmentations of your build to respond to changing game conditions. What if your opponent was really agressive in the beginning? What if he was super passive? Too many people think of builds as a big branching tree, that starts one way and splits up and changes based on what your opponent is doing. A good player tries to figure out
ways to rearrange the components in his build to end up at that same point. In other words, don't view a build as
![[image loading]](http://www.google.com/imgres?imgurl=http://cmelande.files.wordpress.com/2008/06/y_bert.gif&imgrefurl=http://cmelande.wordpress.com/2008/06/&usg=__HEz_XgjCKIH2TMRQKdBS4YMhFFE=&h=550&w=550&sz=11&hl=en&start=0&sig2=B_FajhJh7ITEJZ2az3pBQg&zoom=1&tbnid=I60NSMfyUOm4_M:&tbnh=147&tbnw=147&ei=pending&prev=/images%3Fq%3Dy%26hl%3Den%26gbv%3D2%26biw%3D1639%26bih%3D771%26tbs%3Disch:1&itbs=1&iact=rc&dur=295&oei=m-NpTYXqJ8St8AaKwNScCw&page=1&ndsp=33&ved=1t:429,r:0,s:0&tx=91&ty=33)
Veiw it as
![[image loading]](http://www.google.com/imgres?imgurl=http://faculty.clintoncc.suny.edu/faculty/michael.gregory/files/bio%2520101/bio%2520101%2520lectures/dna/1.gif&imgrefurl=http://faculty.clintoncc.suny.edu/faculty/michael.gregory/files/bio%2520101/bio%2520101%2520lectures/dna/dna.htm&usg=__lTiJEvSYGdXxAMEU-Z_Xm4uHHag=&h=290&w=400&sz=6&hl=en&start=0&sig2=YPjTVk0f4zV-yoRm3f5Eug&zoom=1&tbnid=MYn0R600u-zSFM:&tbnh=147&tbnw=169&ei=FuRpTfXvJIH88AaI3oSNCw&prev=/images%3Fq%3Dupside%2Bdown%2By%26hl%3Den%26gbv%3D2%26biw%3D1639%26bih%3D771%26tbs%3Disch:1&itbs=1&iact=hc&vpx=243&vpy=421&dur=10643&hovh=190&hovw=264&tx=110&ty=126&oei=FuRpTfXvJIH88AaI3oSNCw&page=1&ndsp=33&ved=1t:429,r:17,s:0)
Well, there you have it, the three key components to a build order, in my opinion. I'd like to take a minute to break down a pretty popular build that just about everyone has heard of/seen and break it down using these three steps to really understand the build.
+ Show Spoiler +
The 4 Warpgate Rush:
9 Pylon
12 Gateway+ Probe sent to scout
14 Assimilator
16 Pylon
18 Cybernetics Core
(19 Zealot)
23 Pylon
24 Stalker
26 Warpgate research
27 Gateway
28 Sentry
30 Two Gateways
31 Proxy Pylon
Step 1: The Idea
Let's pretend you've never seen this build before. Lets just take a look at it. You're getting 4 gateways and fast warpgate research, and a proxy pylon. Right away, we know that we're not going to be able to support building constant probes, pylons, gateway units, and expand with this build. The goal of this combination of structures/upgrades is to get a lot of units really fast at any place there is a pylon. Considering there's that proxy pylon in there, it seems that we want to warp in a lot of units near his base and kill him fairly quickly. Ok, cool. We now understand our idea.
Step 2: The components
OK. Now that we know that the goal is to kill our opponent early, lets break down the components of this build. So the key components as I see them are...
4 gateways
1 gas geyser
1 Cybernetics Core
Warpgate Research
1 zealot
1 Stalker
1 Sentry
1 Proxy Pylon
Step 3: Arrangement
OK, so lets look at the arrangement of the components (disclaimer: I know this isn't most optimized 4 gate, its just the first 4 gate that came up in liquidpedia). OK, so we start off with the totally normal 1 gate 1 gas cyber build. We'll have enough gas with this opener to get a stalker and WG at the same time. It looks like we're then getting the next gateway as soon as we have the money and then building a sentry out of that gateway to save some minerals and get those other 2 gates quickly as well. We're then shipping out with a probe to set up a proxy pylon.
Now let's figure out why. Well, with this opener, we're getting a cybernetics core as fast as possible while still getting our probes up, which we're going to need to support 4 gateways. It looks like right when that cyber completes we're making a stalker and researching WG. Then it looks like that sentry comes down when the stalker comes out, and everything from here is when we have the money for it. So, we have 2 units out, 3 gates building, and warpgate researching, and we're about to go make a proxy pylon. A good player isn't going to let you just build a pylon near his base, so it makes sense to bring your units to defend the pylon while its building. This seems ok, though, because we have a sentry popping out soon who can forcefield the ramp if some units try to get into our base, and warpgate is about to finish, so we can warp in units back there if he tries to kill us.
Imagine we played this build, and when your pylon completed, it lined up really nicely with all 4 gateways finishing, but warpgate still isn't done. It seems like warpgate should finish around the time that the gateways and pylon complete. We can now look for a solution. First things first, how can we manipulate these components to make warpgate finish around the same time as the gates and the pylon. Well, we never spend that chronoboost, so lets start with that, because that actually doesn't require me to change the order of buildings at all. Each time you add a chronoboost, it gets closer, until you figure out that if you constantly chronoboost your cybernetics core, your warpgates line up perfectly. Good. We now have everything lining up nicely. So lets put that in a simplified notation.
Build constant probes in the beginning
Place your gateway down as soon as you have the money and scout.
Build an Assimilator after you start your gateway.
As soon as the gateway completes, build a zealot and start a cybernetics core.
As soon as the cybernetcs core completes, build a stalker and research warpgate.
Build another gateway when you have 150 minerals.
Build a sentry when the stalker completes. Start chronoboosting the cybernetics core nonstop.
Build 2 more gateways when you get the money.
After You've started the gateways, take the stalker and zealot with a probe toward his base and build a proxy pylon.
Morph the gateways to warpgates and warp in units and attack his base.
Now is the time to consider what type of things might swing us from our build. This type of push comes out really early, so we only have to worry about fast timing pushes. So, we probably have to worry about 2 gate openings vs. protoss, 2 rax openings from terran, and speedlings from zerg.
In the case of the two gate, he's going to have more units than us if we're not deviating. So how can we rearrange the components to survive this early pressure? Well, lets not even consider moving around components yet, can we manipulate things we are already doing? Well, we have that chronoboost we could spend on the gateway, but our warpgate will come out later, and we'll have to build more zealots and stalkers too, so our push is going to be delayed. That's fine though, because he is sacrificing allot to do this opener. He's not going to have warpgates anywhere near the time we will, and he will probably just be getting stalkers when we hit his base. OK cool. Suppose you try this but you're still noticing that he's getting to kill some probes. Well, we can look around this wonderful site called teamliquid for awhile, and we discover that we can build a wall out of a gateway and a cybernetics core if they're placed correctly, so that we can place our zealot to block that hole and only one enemy zealot can hit him. Suddenly we can defend what we have with far fewer resources. Awesome. So we're going to flip around building placement now, and if we scout 2 gate, we're going to spend our chronoboosts on gateways instead until the attack is over to survive, but since he's delaying his cyber core for so long as well.
Now we just need to consider the other two and make similar adjustments. The key thing to note here is that we never said "he's going two gate, I guess I will no longer be 4 warpgate rushing him" we instead said "we are going to be 4 gate rushing him" and rearranged the components of our build to fit the state of the game and still converge on that same idea. We get 4 warpgates and a proxy pylon near his base to overwhelm him with allot of units.
9 Pylon
12 Gateway+ Probe sent to scout
14 Assimilator
16 Pylon
18 Cybernetics Core
(19 Zealot)
23 Pylon
24 Stalker
26 Warpgate research
27 Gateway
28 Sentry
30 Two Gateways
31 Proxy Pylon
Step 1: The Idea
Let's pretend you've never seen this build before. Lets just take a look at it. You're getting 4 gateways and fast warpgate research, and a proxy pylon. Right away, we know that we're not going to be able to support building constant probes, pylons, gateway units, and expand with this build. The goal of this combination of structures/upgrades is to get a lot of units really fast at any place there is a pylon. Considering there's that proxy pylon in there, it seems that we want to warp in a lot of units near his base and kill him fairly quickly. Ok, cool. We now understand our idea.
Step 2: The components
OK. Now that we know that the goal is to kill our opponent early, lets break down the components of this build. So the key components as I see them are...
4 gateways
1 gas geyser
1 Cybernetics Core
Warpgate Research
1 zealot
1 Stalker
1 Sentry
1 Proxy Pylon
Step 3: Arrangement
OK, so lets look at the arrangement of the components (disclaimer: I know this isn't most optimized 4 gate, its just the first 4 gate that came up in liquidpedia). OK, so we start off with the totally normal 1 gate 1 gas cyber build. We'll have enough gas with this opener to get a stalker and WG at the same time. It looks like we're then getting the next gateway as soon as we have the money and then building a sentry out of that gateway to save some minerals and get those other 2 gates quickly as well. We're then shipping out with a probe to set up a proxy pylon.
Now let's figure out why. Well, with this opener, we're getting a cybernetics core as fast as possible while still getting our probes up, which we're going to need to support 4 gateways. It looks like right when that cyber completes we're making a stalker and researching WG. Then it looks like that sentry comes down when the stalker comes out, and everything from here is when we have the money for it. So, we have 2 units out, 3 gates building, and warpgate researching, and we're about to go make a proxy pylon. A good player isn't going to let you just build a pylon near his base, so it makes sense to bring your units to defend the pylon while its building. This seems ok, though, because we have a sentry popping out soon who can forcefield the ramp if some units try to get into our base, and warpgate is about to finish, so we can warp in units back there if he tries to kill us.
Imagine we played this build, and when your pylon completed, it lined up really nicely with all 4 gateways finishing, but warpgate still isn't done. It seems like warpgate should finish around the time that the gateways and pylon complete. We can now look for a solution. First things first, how can we manipulate these components to make warpgate finish around the same time as the gates and the pylon. Well, we never spend that chronoboost, so lets start with that, because that actually doesn't require me to change the order of buildings at all. Each time you add a chronoboost, it gets closer, until you figure out that if you constantly chronoboost your cybernetics core, your warpgates line up perfectly. Good. We now have everything lining up nicely. So lets put that in a simplified notation.
Build constant probes in the beginning
Place your gateway down as soon as you have the money and scout.
Build an Assimilator after you start your gateway.
As soon as the gateway completes, build a zealot and start a cybernetics core.
As soon as the cybernetcs core completes, build a stalker and research warpgate.
Build another gateway when you have 150 minerals.
Build a sentry when the stalker completes. Start chronoboosting the cybernetics core nonstop.
Build 2 more gateways when you get the money.
After You've started the gateways, take the stalker and zealot with a probe toward his base and build a proxy pylon.
Morph the gateways to warpgates and warp in units and attack his base.
Now is the time to consider what type of things might swing us from our build. This type of push comes out really early, so we only have to worry about fast timing pushes. So, we probably have to worry about 2 gate openings vs. protoss, 2 rax openings from terran, and speedlings from zerg.
In the case of the two gate, he's going to have more units than us if we're not deviating. So how can we rearrange the components to survive this early pressure? Well, lets not even consider moving around components yet, can we manipulate things we are already doing? Well, we have that chronoboost we could spend on the gateway, but our warpgate will come out later, and we'll have to build more zealots and stalkers too, so our push is going to be delayed. That's fine though, because he is sacrificing allot to do this opener. He's not going to have warpgates anywhere near the time we will, and he will probably just be getting stalkers when we hit his base. OK cool. Suppose you try this but you're still noticing that he's getting to kill some probes. Well, we can look around this wonderful site called teamliquid for awhile, and we discover that we can build a wall out of a gateway and a cybernetics core if they're placed correctly, so that we can place our zealot to block that hole and only one enemy zealot can hit him. Suddenly we can defend what we have with far fewer resources. Awesome. So we're going to flip around building placement now, and if we scout 2 gate, we're going to spend our chronoboosts on gateways instead until the attack is over to survive, but since he's delaying his cyber core for so long as well.
Now we just need to consider the other two and make similar adjustments. The key thing to note here is that we never said "he's going two gate, I guess I will no longer be 4 warpgate rushing him" we instead said "we are going to be 4 gate rushing him" and rearranged the components of our build to fit the state of the game and still converge on that same idea. We get 4 warpgates and a proxy pylon near his base to overwhelm him with allot of units.
How to Create Your Own Build Orders Using This Thought Process
+ Show Spoiler +
So, now that we have the process, lets use the though process of reviewing a build and let's create our own! I'm going to present a larely theoretical process I'd go though of creating a build for Protoss (I feel protoss builds are far easier to understand than zerg; my main race, and I don't play Terran)
Step 1: The Idea
OK, I HATE fast colossus in PvT, and I think that upgraded gateway units kick ass. So my basic idea is to get double forges and upgrade my gateway units. I know that late game you're not going to get by on full gateway, so at some point I'm going to need to transition into something. I could transition into either colossus, templar, or carrier. It doesn't make a lot of sense to go carrier after investing a lot of resources into upgrading my gateway units so much and not get any benefit for my power unit, so I think that templar or colossus would be a better option. Along those same lines, templar only benefit from the armor upgrades as it's spell aren't effected by attack upgrades. Therefore, I think I'll be transitioning to colossus at a later point in the game.
Key Elements
At this point, I have a solid idea, and I'm begining to flesh out the possibilites. So, I know I want a robo at some point, but I think I would rather get my forges upgrading before the robotics facility comes out. These forges will allow me to get detection in the form of cannons, so I shouldn't need an early robotics facility to deal with banshees. I know I want to FE, as well. I'm going to want 2/2 after 1/1 finishes, so I'll need a Twilight Council. I'll want a support bay at some point, as well as a third expansion.
So, my key elements thus far are
Gateways
Cyber Core
Gateway Research
Two Forges
A cannon at each mineral line (at least)
Twilight council
Fast Expansion
Third Base
Robotics Facility
Support Bay
So, based on this, I think that after I start 1/1, I can lay down the robo and start pumping out immortals, because upgraded immortals are pretty good. Since I have to get the twilight council to get 2/2, I'll probably snag either blink or charge. I think I'll get more utility out of charge, so I'll get that.
I can open up forge FE, 1 gate, 2 gate, or 3 gate expo. At this point, I have a question. What is important is that I choose something, and go with it. I can experiment with the other FE timings, but it's important to just chose something and play with it and see if it works. If I die consistently when I start testing this build, then I'll have to switch things around.
I think I'll open 2 gate expand, then immediatley lay down forges and put a cannon on each mineral line so I can deal with banshees. I'll start 1/1 upgrades as soon as I can. I'll Add on a twilight council as late as possible, so ideally when it will finish as 1/1 finishes. Since my big idea was to get upgraded gateway units, I think I'll spend my chronoboosts on the forges constantly. If I have any leftover chronoboosts, I think I'll spend those on the robo so that I can have more immortals.
Now I have a good army and it seems a waste to get these fast upgrades for nothing, so I'm either looking for a timing when 2/2 finishes or 3/3 finishes. I think a push with 2/2 will be awesome, because it should secure my colossus and third base get up safely.
At this point I have a general picture in my head of when I should be getting each thing. I now need to play by myself single player and flesh out some of these timings. I will discover exactly when I need to start that Twilight Council, I'll find out exactly when I can expand, and I'll figure out if I need quick 2 gas and sentries, or stay on one gas and expand, etc. Review the replays and figure out where you could be doing something different to improve your efficiency. If you're having trouble, check out Plexa's guide on analyzing replays which is sticked on the strategy forum, support your kick ass mods
So In my head, my order looks something like this
First Gate when I can afford it
Gas after gate
Cyber Core when gate is done and start a zealot
Second Gas after Core
Second gate when I have money
Start producing sentries
Expand when I have the minerals
Secure fast 3rd and 4th gas
Build 2 forges
Build a cannon at each mineral line
Build Research 1/1 at forges and use chronoboosts on forges
Add on a robo
Add on extra gates when I have the minerals
Start pumping immortals (use leftover chronoboosts)
Get a TC at such a time that it finishes when 1/1 finishes
Research 2/2 and charge.
Lay down the support bay when you're close to having 2/2 and charge finish.
Push out to the middle when 2/2 and charge are about to finish, and expand to your third
Poke at the terran and choose to either commit to an attack or do a little damage and draw back when 2/2 and charge finish
Start 3/3
Start pumping colossus and chrono out the colossus.
Get your colossus death ball, and push out into the pile while taking your third.
Now I'll start playing single player a bunch to flesh this build out, then take it to the ladder.
Step 3: Rearrangement
This is the time when I'll consider rearranging my key elements. I'll find out if the original arrangement worked or not. I'll find out quickly that I had to few units when I pushed out, or I had too many gateways and too few immortals, or vica versa. I'll find out that I probably should have arranged something differently. Then I'll arrange the elements until I have a good solid arrangement. Then I'll start thinking, "what if he banshee rushes me? What if he 3 raxes? What if he does X" and I'll plan out my responses ahead of time so I have an idea of what I'll do if he does something that forces a rearrangement. Lets say he's banshee rushing, I might need to get my forge earlier. If he's doing a heavy marauder timing I might need a few cannons at my front. I'll think about everything I'm worried about and think how my build can deal with them, and rearrange from there.
Step 1: The Idea
OK, I HATE fast colossus in PvT, and I think that upgraded gateway units kick ass. So my basic idea is to get double forges and upgrade my gateway units. I know that late game you're not going to get by on full gateway, so at some point I'm going to need to transition into something. I could transition into either colossus, templar, or carrier. It doesn't make a lot of sense to go carrier after investing a lot of resources into upgrading my gateway units so much and not get any benefit for my power unit, so I think that templar or colossus would be a better option. Along those same lines, templar only benefit from the armor upgrades as it's spell aren't effected by attack upgrades. Therefore, I think I'll be transitioning to colossus at a later point in the game.
Key Elements
At this point, I have a solid idea, and I'm begining to flesh out the possibilites. So, I know I want a robo at some point, but I think I would rather get my forges upgrading before the robotics facility comes out. These forges will allow me to get detection in the form of cannons, so I shouldn't need an early robotics facility to deal with banshees. I know I want to FE, as well. I'm going to want 2/2 after 1/1 finishes, so I'll need a Twilight Council. I'll want a support bay at some point, as well as a third expansion.
So, my key elements thus far are
Gateways
Cyber Core
Gateway Research
Two Forges
A cannon at each mineral line (at least)
Twilight council
Fast Expansion
Third Base
Robotics Facility
Support Bay
So, based on this, I think that after I start 1/1, I can lay down the robo and start pumping out immortals, because upgraded immortals are pretty good. Since I have to get the twilight council to get 2/2, I'll probably snag either blink or charge. I think I'll get more utility out of charge, so I'll get that.
I can open up forge FE, 1 gate, 2 gate, or 3 gate expo. At this point, I have a question. What is important is that I choose something, and go with it. I can experiment with the other FE timings, but it's important to just chose something and play with it and see if it works. If I die consistently when I start testing this build, then I'll have to switch things around.
I think I'll open 2 gate expand, then immediatley lay down forges and put a cannon on each mineral line so I can deal with banshees. I'll start 1/1 upgrades as soon as I can. I'll Add on a twilight council as late as possible, so ideally when it will finish as 1/1 finishes. Since my big idea was to get upgraded gateway units, I think I'll spend my chronoboosts on the forges constantly. If I have any leftover chronoboosts, I think I'll spend those on the robo so that I can have more immortals.
Now I have a good army and it seems a waste to get these fast upgrades for nothing, so I'm either looking for a timing when 2/2 finishes or 3/3 finishes. I think a push with 2/2 will be awesome, because it should secure my colossus and third base get up safely.
At this point I have a general picture in my head of when I should be getting each thing. I now need to play by myself single player and flesh out some of these timings. I will discover exactly when I need to start that Twilight Council, I'll find out exactly when I can expand, and I'll figure out if I need quick 2 gas and sentries, or stay on one gas and expand, etc. Review the replays and figure out where you could be doing something different to improve your efficiency. If you're having trouble, check out Plexa's guide on analyzing replays which is sticked on the strategy forum, support your kick ass mods
So In my head, my order looks something like this
First Gate when I can afford it
Gas after gate
Cyber Core when gate is done and start a zealot
Second Gas after Core
Second gate when I have money
Start producing sentries
Expand when I have the minerals
Secure fast 3rd and 4th gas
Build 2 forges
Build a cannon at each mineral line
Build Research 1/1 at forges and use chronoboosts on forges
Add on a robo
Add on extra gates when I have the minerals
Start pumping immortals (use leftover chronoboosts)
Get a TC at such a time that it finishes when 1/1 finishes
Research 2/2 and charge.
Lay down the support bay when you're close to having 2/2 and charge finish.
Push out to the middle when 2/2 and charge are about to finish, and expand to your third
Poke at the terran and choose to either commit to an attack or do a little damage and draw back when 2/2 and charge finish
Start 3/3
Start pumping colossus and chrono out the colossus.
Get your colossus death ball, and push out into the pile while taking your third.
Now I'll start playing single player a bunch to flesh this build out, then take it to the ladder.
Step 3: Rearrangement
This is the time when I'll consider rearranging my key elements. I'll find out if the original arrangement worked or not. I'll find out quickly that I had to few units when I pushed out, or I had too many gateways and too few immortals, or vica versa. I'll find out that I probably should have arranged something differently. Then I'll arrange the elements until I have a good solid arrangement. Then I'll start thinking, "what if he banshee rushes me? What if he 3 raxes? What if he does X" and I'll plan out my responses ahead of time so I have an idea of what I'll do if he does something that forces a rearrangement. Lets say he's banshee rushing, I might need to get my forge earlier. If he's doing a heavy marauder timing I might need a few cannons at my front. I'll think about everything I'm worried about and think how my build can deal with them, and rearrange from there.
Well, I hope this little guide was of some help to some people out there. I'd encourage anyone to check out day9's podcast "A New Look at Build Orders", where I first learned to think of builds in this way.
Any Feedback on how to improve the guide is always welcome and appreciated.
Thanks everyone. Peace.
