This is my attempt to synthesize my ideas about getting good at Brood War into a single thread.
I think there are three key skills: 1) A vision You need a vision of the well-executed siege of your opponent towards which you are building. You want to consider how every unit or combination of units can be a tool in your tool-belt for arriving at and winning a siege. Where a tool is lacking, consider all the ways to pair it with other tools which help cover its deficiencies - even if merely by berserking harder so the opponent doesn't have time to create some monstrosity that takes out your first tool. Of course, it is important not to get run over by an enemy army that is "just stronger" and bigger. Consequently, you start with whatever you need for the early game and midgame and add in tools to build it into your lethal siege.
2) A sharp eye My advice is to look for situations that likely indicate weakness by the opponent. When Protoss shows only 1 Zealot 1 Goon vs Terran for example, the Terran needs to consider this weakness might be real and attack. In general, you want to attack any sign of underlying weakness early on in the game and early on in your training. Eventually you will fall into some traps, and may have to randomize whether or not you attack even from the same scouting information. Playing very aggressively early on will help you map out all of these early attacking options. Understanding the early game very comprehensively will allow you to avoid the opponent being able to get away with too little army / too much weakness / too much growth / too much econ or tech. This will save you from entire worlds of nightmares. You also want to play to cover a wide range of possible moves. You'd like to cheaply poke, attack, scout, etc so that even if that move fails you're likely winning over time. Often, these moves provoke a defense that is ultimately more costly than the move itself, and the attack is justified even when responded to properly. You can't leave these early-game chips on the table and play at the highest level. It's also fun because you get to win/lose games quickly and you learn a lot.
3) A mirror As you get better and better at spotting weaknesses that need to be pounced on while developing towards a full army, look at your own play by starting with Opponent POV in replay. Where are all the weak spots that could have hurt you? Where are the weak spots in the opponent that you should consider exploiting?
On October 12 2025 04:20 ajfirecracker wrote: This is my attempt to synthesize my ideas about getting good at Brood War into a single thread.
I think there are three key skills: 1) A vision You need a vision of the well-executed siege of your opponent towards which you are building. You want to consider how every unit or combination of units can be a tool in your tool-belt for arriving at and winning a siege. Where a tool is lacking, consider all the ways to pair it with other tools which help cover its deficiencies - even if merely by berserking harder so the opponent doesn't have time to create some monstrosity that takes out your first tool. Of course, it is important not to get run over by an enemy army that is "just stronger" and bigger. Consequently, you start with whatever you need for the early game and midgame and add in tools to build it into your lethal siege.
2) A sharp eye My advice is to look for situations that likely indicate weakness by the opponent. When Protoss shows only 1 Zealot 1 Goon vs Terran for example, the Terran needs to consider this weakness might be real and attack. In general, you want to attack any sign of underlying weakness early on in the game and early on in your training. Eventually you will fall into some traps, and may have to randomize whether or not you attack even from the same scouting information. Playing very aggressively early on will help you map out all of these early attacking options. Understanding the early game very comprehensively will allow you to avoid the opponent being able to get away with too little army / too much weakness / too much growth / too much econ or tech. This will save you from entire worlds of nightmares. You also want to play to cover a wide range of possible moves. You'd like to cheaply poke, attack, scout, etc so that even if that move fails you're likely winning over time. Often, these moves provoke a defense that is ultimately more costly than the move itself, and the attack is justified even when responded to properly. You can't leave these early-game chips on the table and play at the highest level. It's also fun because you get to win/lose games quickly and you learn a lot.
3) A mirror As you get better and better at spotting weaknesses that need to be pounced on while developing towards a full army, look at your own play by starting with Opponent POV in replay. Where are all the weak spots that could have hurt you? Where are the weak spots in the opponent that you should consider exploiting?
My BroodWar ID is Bushido
There are a lot of aspects that go into being a good player.
The question is, do you want to just be good ar Brood War? Do you want to be great? Do you want to be a professional? Do you want to be an all-time legendary player? Or do you want to be the best player of all time?
For us, playing competitive/professional sports and e-sports...if your goal isnt to win then why play? If it's fun losing for you...great, but that's not us.
And for us, winning means against anyone/anything.
In terms of Brood War, that means being dilligent about a lot of things...not going to get into all of them at the moment 🤷♂️🤷♀️
On October 12 2025 17:56 perfectspheres wrote: And for us, winning means against anyone/anything. In terms of Brood War, that means being dilligent about a lot of things...not going to get into all of them at the moment 🤷♂️🤷♀️
Yeah but this is a strategy & training thread. I'm trying to lay out how to train for strategic competence. Obviously you'll also have to work on the mechanics, including build orders and macro and control groups etc etc
But specifically for strategy, I think you need to use raids and attempted busts to prevent the opponent from investing too much into growth while you develop a lethal siege.
So my advice is designed to sharpen your vision of raids and busts (which is useful even if you never get to a siege) as well as how to progress to victory even if you can't find a critical weakness to hit with a raid or bust.
What else needs to be done to train strategic vision ?
On October 12 2025 17:56 perfectspheres wrote: And for us, winning means against anyone/anything. In terms of Brood War, that means being dilligent about a lot of things...not going to get into all of them at the moment 🤷♂️🤷♀️
Yeah but this is a strategy & training thread. I'm trying to lay out how to train for strategic competence. Obviously you'll also have to work on the mechanics, including build orders and macro and control groups etc etc
But specifically for strategy, I think you need to use raids and attempted busts to prevent the opponent from investing too much into growth while you develop a lethal siege.
So my advice is designed to sharpen your vision of raids and busts (which is useful even if you never get to a siege) as well as how to progress to victory even if you can't find a critical weakness to hit with a raid or bust.
What else needs to be done to train strategic vision ?
I would consider stuff like "seeing Protoss move out with 1 Dragoon and 1 Zealot" and doing stuff in reaction to that, or "attempted busts" as being more along the lines of tactics rather than strategy tbh, but either way, my experience was that the "strategy" aspects of the game change significantly once you start getting higher up in the ladder. Some of the things that work at D will stop working at B, for example most 1 base tech PvZ builds. So, in order to know what strategies are actually viable and learn how to execute them correctly, I believe that a non-trivial amount of mechanical skill is needed to even have that hands-on exposure. The best way to get to that point, IMO, is to emulate a pro build order and run it dozens of times to see what interactions it has, and to practice mechanics on the side via macro or hotkey trainer.
On October 18 2025 05:24 Jealous wrote: I would consider stuff like "seeing Protoss move out with 1 Dragoon and 1 Zealot" and doing stuff in reaction to that, or "attempted busts" as being more along the lines of tactics rather than strategy tbh, but either way, my experience was that the "strategy" aspects of the game change significantly once you start getting higher up in the ladder. Some of the things that work at D will stop working at B, for example most 1 base tech PvZ builds. So, in order to know what strategies are actually viable and learn how to execute them correctly, I believe that a non-trivial amount of mechanical skill is needed to even have that hands-on exposure. The best way to get to that point, IMO, is to emulate a pro build order and run it dozens of times to see what interactions it has, and to practice mechanics on the side via macro or hotkey trainer.
I am trying to provide a general strategy that will (among other things) help you find all the winning tactics to then learn to execute.
I apologize if I haven't explained this well, I am still thinking through how to best express it.
One major point is you can't let the opponent be weak and over-invest in some form of growth or he'll come out and clobber you. So you need to be finding all the points where your build (whatever it is) can cost-efficiently make trouble for him, because that means he is somehow under-defended. This could be a little raid with 1-2 men, this could be a force moving in to try and bust him, or (in the long run) it will be a siege that wears him down. Keep in mind raids are legitimate all game, not just in the early game. The Boxer Eraser is a raiding tactic.
You need to find every one of these attacks that your build can cost-effectively present and at least some % of the time, use them. The alternative is getting clobbered by all sorts of nightmarish things the opponent shouldn't have had time to get.
In doing this, you need to be winning against a range of the opponent's possible moves. Eg. Protoss spinning the big tech wheel of cheese. If he's doing a strategy 20% of the time you can't counter it 100% of the time unless that counter also beats a bunch of other stuff. You'd have a 20% win-rate.
So you're trying to be very cost-efficient with each attempt at disruption or scouting in order to be able to afford to counter more than 50% of what the opponent is doing and get over 50% winrate. Opponent will let you know when you've been caught overextending somehow. It's one of your top jobs to catch him. (Edited to add: It's especially great if you can find instances where even proper defense is costlier than the attack)
As you get a sharper and sharper eye for what attacks are possible, you should review your replays from Opponent POV to try and spot weaknesses in your own play.
Then in terms of long term strategy I think it's more about identifying what a cost-efficient siege looks like and the earlier steps get reoptimized in light of getting to the winning side of a fatal siege.
Some of this is going to be messing around with weird tactics because the siege constrains what the opponent can spend on. Unit selection becomes more about addresssing specific problems rather than purely "making a good army".
I recently made a TvZ suggestion to include Firebats & Tanks alongside a competent anti-air force. This seems to be something that top-level Terrans have decided is a good inclusion and has earned a share of the metagame. Partly this is just tactics. But partly it's more of a siege mentality - needing to keep grinding the opponent down so he can't ever present a real threat. Tools that don't make much sense for optimal fighting in the field can suddenly become very efficient when you're on the sieging side.
I have many similar ideas, some of which are doubtless theorycrap. This is the only one I can so clearly point to replays to justify my position, so it's the only example I'll give, but you can think of your own. Don't be confined to always "Go Bio" or "Go Mech" etc. Get whatever tools will let you do the job.
This is one reason I think it's important to be somewhat experimental in unit choice rather than just saying X is a good unit, Y is a bad unit, ABC is a good army, DEF is a bad army. Instead, we should think under what conditions a unit or group of units could be good or bad. I'm strongly attracted to units or combinations of units that seem like they could be really efficient and maintain the pressure of the siege. Once I have an idea of some units that might go well together, it's a question of whether there is any sensible plan of development that gets all the pieces needed.
On October 18 2025 05:24 Jealous wrote: I would consider stuff like "seeing Protoss move out with 1 Dragoon and 1 Zealot" and doing stuff in reaction to that, or "attempted busts" as being more along the lines of tactics rather than strategy tbh, but either way, my experience was that the "strategy" aspects of the game change significantly once you start getting higher up in the ladder. Some of the things that work at D will stop working at B, for example most 1 base tech PvZ builds. So, in order to know what strategies are actually viable and learn how to execute them correctly, I believe that a non-trivial amount of mechanical skill is needed to even have that hands-on exposure. The best way to get to that point, IMO, is to emulate a pro build order and run it dozens of times to see what interactions it has, and to practice mechanics on the side via macro or hotkey trainer.
I am trying to provide a general strategy that will (among other things) help you find all the winning tactics to then learn to execute.
I apologize if I haven't explained this well, I am still thinking through how to best express it.
One major point is you can't let the opponent be weak and over-invest in some form of growth or he'll come out and clobber you. So you need to be finding all the points where your build (whatever it is) can cost-efficiently make trouble for him, because that means he is somehow under-defended. This could be a little raid with 1-2 men, this could be a force moving in to try and bust him, or (in the long run) it will be a siege that wears him down. Keep in mind raids are legitimate all game, not just in the early game. The Boxer Eraser is a raiding tactic.
You need to find every one of these attacks that your build can cost-effectively present and at least some % of the time, use them. The alternative is getting clobbered by all sorts of nightmarish things the opponent shouldn't have had time to get.
In doing this, you need to be winning against a range of the opponent's possible moves. Eg. Protoss spinning the big tech wheel of cheese. If he's doing a strategy 20% of the time you can't counter it 100% of the time unless that counter also beats a bunch of other stuff. You'd have a 20% win-rate.
So you're trying to be very cost-efficient with each attempt at disruption or scouting in order to be able to afford to counter more than 50% of what the opponent is doing and get over 50% winrate. Opponent will let you know when you've been caught overextending somehow. It's one of your top jobs to catch him. (Edited to add: It's especially great if you can find instances where even proper defense is costlier than the attack)
As you get a sharper and sharper eye for what attacks are possible, you should review your replays from Opponent POV to try and spot weaknesses in your own play.
Then in terms of long term strategy I think it's more about identifying what a cost-efficient siege looks like and the earlier steps get reoptimized in light of getting to the winning side of a fatal siege.
Some of this is going to be messing around with weird tactics because the siege constrains what the opponent can spend on. Unit selection becomes more about addresssing specific problems rather than purely "making a good army".
I recently made a TvZ suggestion to include Firebats & Tanks alongside a competent anti-air force. This seems to be something that top-level Terrans have decided is a good inclusion and has earned a share of the metagame. Partly this is just tactics. But partly it's more of a siege mentality - needing to keep grinding the opponent down so he can't ever present a real threat. Tools that don't make much sense for optimal fighting in the field can suddenly become very efficient when you're on the sieging side.
I have many similar ideas, some of which are doubtless theorycrap. This is the only one I can so clearly point to replays to justify my position, so it's the only example I'll give, but you can think of your own. Don't be confined to always "Go Bio" or "Go Mech" etc. Get whatever tools will let you do the job.
This is one reason I think it's important to be somewhat experimental in unit choice rather than just saying X is a good unit, Y is a bad unit, ABC is a good army, DEF is a bad army. Instead, we should think under what conditions a unit or group of units could be good or bad. I'm strongly attracted to units or combinations of units that seem like they could be really efficient and maintain the pressure of the siege. Once I have an idea of some units that might go well together, it's a question of whether there is any sensible plan of development that gets all the pieces needed.
So is this all just theory, or did you use this approach to reach a higher ranking yourself? Maybe some replays of these strategic concepts in action would be useful to demonstrate the viability of the ideas presented. I'm curious what situations you are finding where two units are applying real pressure to your opponent.
This is (I believe) a strong player following my advice on many points, with a focus on constant efficient attack:
and here, a further iteration by Light:
Edited to add: In this thread I am attempting to offer a compass to strategically sound play. I am a viewer, but I know a few things. Subgame perfect nash equilibria is mathematically proven to be the least exploitable strategy for any type of game in which uncertainty plays a major role, such as Poker. I am attempting to spell out what that means for Brood War. Obviously this is of greatest importance when uncertainty looms large, as it so often does in the early and midgame. My application of the concept may be somehow lacking, but SPNE strategic stability is simply a mathematical truth.
I suppose I am permitting myself to imagine that meaningful uncertainty is mostly gone by the late game. There I am merely describing the strategy in broad but accurate terms (which also apply to the early game). My hope there is to help players see clearly the nature of the task and all the different tools they might find efficient along the way. There's not really any theory here except dreaming up that a certain tool or arrangement of tools might or might not prove useful. So to avoid that kind of theorycrafting, I have been quite vague. TvZ Firebat/Tank is one example, TvP Bunker Range is another, but these are the kinds of theories to try to craft rather than "I tested this and it definitely works great" so I'll not list any further examples. My attempted contribution to the discussion of lategame is that even coming up with ideas to try can be guided in a logical direction by understanding the framing of the game (a mutual siege) and taking note of the constrained resources on both sides. It follows the "theories" one should attempt to "craft" involve some sort of high cost efficiency.
And before you say that's ridiculously obvious or useless insight, look at so many of the top Zergs in the world trading Scourge badly into Corsairs. You need 75% efficiency for gas parity and they're not getting anything like that. Then they build 11 Muta to snipe 2 High Templar Rest of the army is all Hydra at 100 supply and get stormed to death by 130-140 supply Protoss. Friggin retards.
This is nothing to do with uncertainty, this is just playing from poor principles. But what would good principles look like?
My ZvP theory craft with zero replays to back it up: Make exactly 5 Muta like Effort in ProLeague Pick templar with lings rather than muta (I know I've seen it but can't think of a game to cite) Go hydra ling into ling lurker hydra into hive like Soulkey and Hero in KCM Make some spores if needed to defend overlords like Larva in ASL plus make scourge as needed for shuttle / obs like every late ZvP ever
but combining all those elements is technically theorycraft, I can't show you it winning yet
I doubt we can characterize Brood War here. Let us focus on one-on-one, BO1 matchups against random opponents on a reasonably sized set of known competitive maps. If we can quantify skill in this context, I wager it would look like a bell curve. At what percentile do you want to be? This is an important question because, as someone else said in this thread, your approach to getting there in a particular amount of time depends on your answer. There are first principles that would put you roughly a standard deviation or two above the mean, like mastering a handful of build orders, materializing an army advantage in the form of, say, an economic advantage, exploiting map features, etc., but this does not sound like what you are trying to accomplish here. We can dramatically increase the complexity of this problem by generalizing it to BO3+ matchups against known opponents, in which case psychology plays a larger role, but for now, it should suffice to talk about this simpler problem.
On October 12 2025 17:56 perfectspheres wrote: And for us, winning means against anyone/anything. In terms of Brood War, that means being dilligent about a lot of things...not going to get into all of them at the moment 🤷♂️🤷♀️
Yeah but this is a strategy & training thread. I'm trying to lay out how to train for strategic competence. Obviously you'll also have to work on the mechanics, including build orders and macro and control groups etc etc
But specifically for strategy, I think you need to use raids and attempted busts to prevent the opponent from investing too much into growth while you develop a lethal siege.
So my advice is designed to sharpen your vision of raids and busts (which is useful even if you never get to a siege) as well as how to progress to victory even if you can't find a critical weakness to hit with a raid or bust.
What else needs to be done to train strategic vision ?
I would consider stuff like "seeing Protoss move out with 1 Dragoon and 1 Zealot" and doing stuff in reaction to that, or "attempted busts" as being more along the lines of tactics rather than strategy tbh, but either way, my experience was that the "strategy" aspects of the game change significantly once you start getting higher up in the ladder. Some of the things that work at D will stop working at B, for example most 1 base tech PvZ builds. So, in order to know what strategies are actually viable and learn how to execute them correctly, I believe that a non-trivial amount of mechanical skill is needed to even have that hands-on exposure. The best way to get to that point, IMO, is to emulate a pro build order and run it dozens of times to see what interactions it has, and to practice mechanics on the side via macro or hotkey trainer.
I would consider Jealous absolutely correct on this matter. It is one thing to theory craft for the game, and to hone strategy, but if your mechanics as a player are inadequate, then you will be unable to execute your strategies in a game.
I would certainly consider this to be pro level advice, and that learning the builds pros use is a necessity to developing your own strategies. You have to be familiar with the meta game before you can try breaking convention.
Considering this thread is about training, I'll offer some builds which I think would be worth emulating, for good practice.
FlaSh has a 5-0-0 build he uses sometimes in the Zerg matchup, I believe there are a couple videos of the build on ArtosisCasts. This build would be very good for training your macro skills, as the macro cycle required to support 5 rax and 2 ccs, on top of tech, will make other builds seem easy to execute in comparison.
In general, for newer players, I'd reckon learning two base all-ins is ideal for training early macro, eventually you can learn how to macro past a two base once familiar with the early to mid game transition.
For micro specifics, there are custom maps which would be more useful for training and warming up, then just playing sets would. For zerg players, there is the muta and scourge map to train your muta micro, and there is a Terran version of this map to practice your Valkyrie micro.
Aside from this, the tactics which you employ are incredibly important, as even with a sound strategy, if you have poor tactics, you will be unable to win. This is best learned IMO, from viewing high level pro matches, seeing their tactics, then trying to implement them into your own games. Light probably has the finest tactics of the Terran player base, although I think he has a tendency to make strategic blunders by being overly aggressive.
I have an immense respect for the pros of this game, I've decided to simply be a theory crafter, a spectator, enjoyer, and perhaps a map maker. It'd take me 5 years minimum to get my mechanics up to a high enough level, to maybe attempt going pro. Without a doubt, very difficult game to get even decent at. Artosis would kick my *ss, and you've seen how he does against pros.
EPT
