This section is intended for it's interest value as well as it's pure strategical value. I am including it only because I am quite impressed with the level of strategic thought that the Art of War was written with given how old it is. While I do not think that the Art of War was written in a way that newbies can really appreciate (I owned it when I was a newbie and I did not truly appreciate it until I learned it's lessons the hard way myself), but for those of you who have not read the Art of War, I have interpreted and summarized each of the relevant sections of the Art of War to Starcraft.
First, a bit of background info on the Art of War. The Art of War in it's original form was a series of scrolls, apparently written by a mysterious author known as Wu Sun-Tzu. While nothing is known for certain about Sun-Tzu, or even whether he actually existed, the best guess is that he was a general for the Wu kingdom during the Chinese spring and fall period, sometime around 2000 years ago. Credit to diabloxdj for corrections/further info.
Without further ado, let's get right to it. The section headings are given as presented in my copy of the book, The Art of War.
Section 1 (Initial Estimations), verse 6:
"Warfare is the Way (Tao) of deception. Thus although you are capable, display incapability to the enemy. When commited to employing your forces, feign inactivity. When your objective is near, make it appear distant, when it is far away, make it appear near."
Sun Tzu was perhaps the first great general to grasp the importance of deception. You already know my own opinions of it. Sun Tzu lays out in general terms the most important ways to decieve your enemy. Namely in your capabilities, and in your intentions. Those are the two pieces of information your enemy most greatly desires. So, in practical terms, if you are planning to tech up one tree, show him you are teching up another. If you plan to tech up slowly, show him you are teching quickly. If you plan to tech up quickly, show him a slow tech.
The easiest way to accomplish those things are with the opening scout. When the enemy is scouting you near the beginning of the game, always use that to your advantage. The first thing you should do is build a hidden pylon somewhere where his scout is not going to be, such as the bottom of the 6 oclock natural or the far right side of the 3 oclock natural on gamei temple. Build up your true intentions there, while building your fake intentions in plain sight. Do this just as you are about to kill the enemy scout, then cancel them right after you have killed their scout. That level of deception works to some degree about 90% of the time at most levels of play.
A practical mid game application? Hide your army or some part of it somewhere. Move out a containment force to the enemy's natural that you know to be inferiour to his army. Make him think that you are overconfident. Allow him to break your containment, then flee, then use your backup force to encircle his pursuing troops. How to ensure that his troops do pursue you? Again it's all about the deception. Allow him to scout you expanding, so that he must come out to stop the expansion. As soon as his forces are committed to the fight though, you can cancel the expansion and focus on troop production to press your advantage.
Further, mask your intentions of attacking or expanding. Particularly against zerg in the early game, pretending to attack while expanding, or pretending to tech while attacking, can be highly effective. Zerg will have constant vision of you until you get dragoons; you can use that to your advantage. Be aware of what his overlord can and cannot see, and assume that the player is aware of everything the overlord can see. Imagine, what would it be most detrimental for your opponent to think, then try to make him think that by giving his overlord clues to your false intentions.
Here is a simple example of fooling a protoss player:
  |
First, I move my early game goons to the bottom of my ramp and kill his second scouting probe before it can enter my natural. My opponent can only assume I expanded early, and is moving out to attack with some dragoons. However, I made no such expansion, and instead am moving my goons into position. When his army attacks they are hopelessly outnumbered and outpositioned. I killed a few of them and chased them back to his natural, preventing him from expanding while I do so myself with a comfortable lead in mass.
This applies to more than just early game zerg of course. If the main thrust of your plan is some kind of mineral line drop, be it a reaver drop, dt drop, ht drop, then always attack him in the front of his base. Force his attention to the battle for his natural or his ramp, force him to think that you have committed to a frontal assault build order, thus pulling away his defenses or at the very least his attention to the back of his base. Then drop your reavers or templars.
Here is another protoss vs protoss example:
   |
First I place my robotics just outside his base in order to reduce the travel time dramatically and prevent him from scouting it if he had for some reason made a super fast observer or got up my ramp unexpectedly. Once my reaver is ready, I head out to fight his containing force. They are perfectly positioned across my bridge, but I gladly fight the losing battle to hold his attention while my reaver lands in his mineral line. The two goons I lost were more than worth the numerous probe kills my reaver scored.
The inverse is true for frontal assaults. If you want to knock out his mineral only on Lost Temple for example, force him to move his troops away from it. This is surprisingly easy to do with 3 or more shuttles. Load them up and drop them very close to edge of his main cliff. The forces should mostly consist of dragoons, or reavers if you happen to have any. This is so that your units, having range, will not be induced to wander off attacking things, but will remain close to the cliff and your shuttles. He will pull back most of his frontal defenses to defeat this drop. You know this, so after a couple of seconds, load your troops back up and drop them on the low ground right beside the cliff, and proceed on foot with the rest of your army with a frontal assault. The reason this works is that his troops have to travel much further then yours. Your troops are getting picked up and carried a distance of a couple of matrices, and then have to travel only a few steps to get to his mineral only. His troops, however, have to travel around cliffs and up and down a ramp. It takes about 5 seconds less to move 6 dragoons from a main to a min only with shuttles then it takes them to move that far overland on most start positions. By then your forces are already perfectly positioned and his forces have to attack-move into your forces, which is to your advantage.
verse 7:
"Display profits to entice them. Create disorder to take them. If they are substantial, prepare for them. If they are strong, avoid them. If they are angry, perturb them, if they are arrogant, be deferential. If they are rested, force them to exert themselves. If they are united, separate them. Attack where they are not prepared. Go forth where they do not expect it. These are the ways military strategists are victorious. They cannot be spoken of in advance."
Verse 7 is connected to verse 6 in that it shows some more concrete ways to decieve the enemy. I have already gone over that in some detail. The sentence, "If they are strong, avoid them" is an important lesson that ties into the mass/power equation. Do not be forced into a direct confrontation with the enemy if they have more mass, because you should have more power, and in that case you should wait for your power to translate to mass before fighting him.
Verse 7 also comments on psychological warfare somewhat. This advice can be somewhat useful in tournament play, though manners-wise many players frown on attempting to use psychology to influence the game. Basically it says that if your enemy is someone prone to be angry, fuel the fires a little and make him angry. If your enemy is cocky and arrogant, pretend to be overrawed. I must admit I'm a little guilty of this myself, because going into a tournament I almost always attempt to portray myself as the underdog, and if I have any pre-tourney practice games with someone who I am playing against or against a friend of a person I am playing against, or if a person I am playing against is watching, I will certainly throw those games using strategies that I would never use in a real game. I call that process 'prepping the competition'. I portray myself in a way that I hope will influence them to play against me differently--and worse--then they normally would. For example against one opponent, I portrayed myself as a hopeless techer. I gassed after 1 gate against his zerg for 3 games in a row. In one game I allowed myself to be defeated by a zergling rush. In the other games I did a dt/sair attack that hopelessly failed and a reaver drop that was killed by a handful of hydras. In the actual game I showed his ovie tech once again, but I did a hidden zealot rush and barged into his expansion with 12 zealots at a time when he thought I should have had only 4, and those in my main. His defenses of 6 lings and a morphing sunken were totally inadequate. Naturally my opponent called me a lucky rushing newbie, but the reality is that I had been planning to do that exact thing against whatever zerg I faced for a long time in advance.
Section 4 (Military Disposition) verse 2:
"One who cannot be victorious assumes a defensive posture; one who can be victorious attacks. By assuming a defensive posture, the strength will be adequate, whereas in offensive actions it would be inadequate."
Sun-Tzu refers again to the mass/power equation. When the enemy has more mass, play defensively and allow your power to play out. When the enemy has less mass, attack right away, before his power plays out, or use your mass to shield a big rush for power. This verse also leads right into an important strategic concept in the next verse, which I will discuss below.
verse 4:
"Thus one who excels at warfare first establishes himself in a position where he cannot be defeated while not losing any opportunity to defeat the enemy. For this reason, the victorious army first realizes the conditions for victory, and then seeks to engage in battle. The vanquished army fights first, and seeks victory after."
This verse is a little more difficult to understand but it captures one of the more subtle aspects of Starcraft. I went over this in the basic strategy section. Basically, you start out conservatively, always scouting the enemy first without committing to anything. Once you have sufficient information, you make your move, switching to an aggressive strategy and going for victory. If you play aggressively without first scouting, you vanquish yourself.
Section 5 (Strategic Military Power) verse 1:
"In general, commanding a large number is like commanding a few. It is a question of dividing up the numbers. Fighting with a large number is like fighting with a few. It is a question of configuration and designation."
This verse pertains a lot more to real life warfare then starcraft, but it does provide a little insight on the use of both good hotkeying and good organisation. Why is it easier to control 12 units then 60? The biggest reason is because it's easy to keep track of the role of each individual unit in the group of 12. When you are forming your hotkeys, remember that. Form your hotkeys so that each number represents a group of units with a specific role. Suppose you want to do the lure and encircling maneuvre discussed above. One set of hotkeys should consist of the luring forces, and another the encircling forces. That is how you can control large numbers as if they were few. Also, your base organisation can benefit from this. There should be a logic to how your buildings are placed. Most players haphazardly put their tech buildings anywhere, or else follow some rote for no other reason than that that's how they've always done it. However, it's rare, but building placement can occasionally be critical to the outcome of the game. Not only should your gateways obviously be clumped, but bear in mind how your units will come out of them. Can your forces easily move to the front lines? If they have to walk around and mill about before escaping from your main, it's almost as if you've needlessly added to their build time. Furthermore, remember which tech buildings are important, and keep in mind their hitpoints. Against zerg, your most important buildings to defend during a drop are actually your forges, so don't use them as a wall, keep them protected. Your least important building is the citadel of adun of course, but you can also usually afford to go without your observatory or support bay, or even an archives or core for a little while, so long as you don't lose them both at once. For that reason, you should keep them separate if possible. Meanwhile, against terran, your most critical building is your cybernetics core, while your forge is relatively unimportant. Also, of course, make sure you aren't walling off your own troops if you should need to defend against a drop.
Section 6 (Vacuity and Substance)
"Thus if I determine the enemy's disposition of forces while I have no perceptible form, I can concentrate my forces while the enemy is fragmented. If we are concentrated into a single force while he is fragmented into 10, then we attack him with ten times the strength. Thus we are many and the enemy is few, If we can attack his few with our many, those who we engage in battle will be severely constrained."
That verse represents perhaps the pinnacle of tactical and strategic thought in the Art of War. It describes in detail how to successfully attack and defeat a defensively postured force of equal strength, the strategic holy grail.
In order to defeat a defensive enemy of equal strength, there are two phases to carry out. The first is probing, in which you scout the enemy's disposition. Locate the areas where the enemy will be vulnerable to attack. These include his tech centers, his mining bases, and his production. Those areas will be your objectives. Next, determine where his defenses are located. Ultimately you want to be able to lauch a main assault on one target. So in your probing assaults, you will manipulate the enemy into leaving one area of his defenses lacking. If your intention is to launch a doom drop on his main, then you will continually attack his front lines or outlying expansions in order to draw away defenses. If your intention is to attack one expansion, you will continually attack his main base or other expansions. In these little attacks, you will never risk losing many troops. You will simply pretend that you are intending to attack but did not have adequate scouting information, and decided to pull back at the last second. You will try this more then once, and with each successful defense, your enemy will be more confident that he is winning. Finally you will continually move your force around, and never allow him to see it. You will make sure that he has no observers, burrowed troops, or overlords watching what you are doing. If he scans you, you will move to a new spot and wait until he can no longer trust that information. Only when you are ready to make your true assault should your army form up into a single battle ready force. The second phase is the execution. No major battle should be undertaken without all proper preparation. You should know what you are attacking, you should know how your enemy can respond, and you should be ready for it. Hotkey your attacking troops appropriately and keep them as tightly compact as possible. Concentrate your maximum amount of force on one area, and try to make that area an area the enemy will try to defend. When you attack the area, you know his forces will respond immediately, so assume a defensive posture once you have moved to the objective. If you are attacking an expansion, once you have moved your forces to the expansion and defeated the defenses there, align them into a defense posture. Protect your templars and have them ready to psi storm all approaching forces. Spread your dragoons into an arc and ready your zealots to flank. You will force the enemy to come to you, and there you will have the advantage. That is the essence of Vacuity and Substance. Create a vacuity in the enemy, then replace it with your substance. That is how to secure an objective against an equally powerful force.
In an actual game, the circumstances are constantly changing. Starcraft is an extremely dynamic game, so make sure that you don't lock yourself into this paradigm. This isn't a 'strategy' per se, something that you should necessarily focus all of your energies on. Rather, this is a general way to conduct an entire game, a sort of guideline for how to use your troops during the game. Identify the enemy's weakness, then make moves to exploit it. If you should notice that the enemy's weakness has changed, act on the new information, don't insist on trying to force the situation back to your original conception. If, in a probing assualt, you realise that you have stumbled upon a true vulnerability, don't pull back merely to prepare for the chance of a possible later vulnerability: take opportunities as they come. A bird in the hand is worth 2 in the bush. However, as you grow more and more experienced, and play better and better players, you will soon realise that weaknesses are made more often then found. It is in those cases that following these guidelines will be helpful. If you watch the overall flow of games between pro-gamers, you will be able to see that, usually, both players operate the entire game according to this paradigm.
Section 7 (Military Combat) verse 8:
"In order await the disordered; in tranquility await the clamorous. This is the way to control the mind."
verse 9:
"With the near await the distant; with the rested await the fatigued; with the sated await the hungry. This is the way to control strength"
Both verses speak to different aspects of real life warfare, but to the same aspect of Starcraft. Whenever possible, engage in combat only in terms favourable to yourself. If the enemy has set up an ambush of your forces, retreat instantly before major damage is done. If you are setting up an ambush on the enemy, prevent retreat if possible or set up the ambush in a way that major damage will be done before retreat can be effected. Setting up an ambush requires two things: you must know first of all what is to your advantage, and you must know second of all how to accomplish this.
Whenever attempting to ambush the zerg, remember that your forces gain advantage by awaiting the zerg and forcing them to advance into you. In this way your templar will have an easy time storming the enemy. It is easy to storm an enemy approaching a templar standing still. It is hard to move a templar forward to storm an enemy. It is easy for enemy lurkers to defend a position. It is hard for them to move forward, then burrow, to attack one. Therefore your forces will always be better off if they await an attack from the zerg. So if you wish to engage the zerg in favourable combat, set up your strategy in a way that forces the zerg to attack you, and not the opposite. Wherever possible, attack a lightly defended zerg expansion and ambush the reinforcements. Wherever possible, lure the zerg forces into your templar. Do not allow your zealots to be lured into his lurkers and hydras. Be proactive in a standoff. Do not allow his hydras to move forward and pick your troops off; instead await his hydras with your templar and storm them before they can fire on you. Do not attack move your forces into a harassing hydra force backed up by lurkers unless forced to by the mass power equation. Instead use the longer range of psi storm to entice him into attack moving into you. Usually the zerg will simply give ground rather then be pressed into combat. In that way you can safely pressure a zerg containment force. In that way you can gain cost efficiency in a situation that you should not.
Whenever attempting to ambush terran, remember the opposite is true. Terran can always force you to attack him when he is in a strong position. This is because his tanks outrange you when in siege mode, and because his mines are stationary and less effective when laid in the middle of a fight, it will be advantageous to him if he can do it. Terran's goal will always be to use his range to force you to engage him on his own terms. If your force is standing still awaiting the terran, this is exactly what will happen; since tanks outrange psi storm, this will always be easy for him in normal circumstances. So your goal is to create circumstances in which you can engage the enemy in favourable positions. A position favourable to protoss is one that is wide open and flat, so that nothing impedes your zealots and dragoons from advancing into firing range of the longer ranged siege tank. Also, your forces should be as spread out as possible while still reaching the enemy positions more or less at the same time, which means attacking from multiple directions at once. In that way splash damage from tanks and mines will be mitigated as much as possible. Finally, and most favourably of all, you should always attack the enemy when he is moving wherever possible. Abandon all other circumstances and advantages immediately if you see the opportunity to attack a large terran army on the move with a large army of your own. Tanks unsieged and moving forward are weaker then dragoons. Vultures fighting you without the benefit of their mines do not have the capacity to deal significant damage. Whenever terran is moving, his tanks necessarily have less range, therefore it is far easier to effectively employ psi storm. And, dt's almost always get in close and get several free shots before they are scanned because in all the commotion and confusion the terran almost always fails to notice the dts and forgets to scan. Usually by the time he does scan, his forces are already dead. Therefore it is to your advantage to attack the terran--only in these circumstances of course--rather then allowing the terran to attack you, thus forcing upon you negative circumstances. The difference in attacking a well prepared, stationary terran force in terrain of his choosing, and attacking an unprepared and unsuspecting terran in the open is like night and day. The exact same forces in these different circumstances will have the exact opposite results: the protoss will be completely vanquished in the former, and the terran will be completely vanquished in the latter, and the victorious forces will suffer almost no losses in both cases. That is the art of tactics in starcraft.
Here is an example of both this section, and the previous one, Vacuity and Substance:
In keeping with the principles of Military Combat, my early mid-game army is roving the map unpredictably when I scout a new expansion attempt by terran to a second natural close to his main. I quickly move down to dispatch it. However, it seems as though terran has caught me in his trap, as he is now sieging tanks, building turrets, and laying mines to trap me in that canyon. With my mostly dragoon force, there is no chance of my breaking through; any frontal attack to escape would be decimated. I was in a bad place.
  |
Or was I? I had placed observers, as you can see, and knew perfectly well that terran would move to trap me in that canyon once I attacked the second expansion. I had brought along a shuttle, and I used it to cross over the small cliff seperating his natural from the base I was attacking, and unloaded 4 zealots into the now undefended mineral line. He was forced to move his army back to defend his natural from my zealots, so I picked the zealots up again. While that was happening, I used my dragoons and observer to clear out the mine field containing me. He quickly tried to move his tanks back and siege up again, but by now I had another small force coming from my main flanking his containing force, and to add insult injury, I unloaded the same zealots which were in my shuttle onto his sieging tanks. I destroyed the containing army with ease.
   |
In order to force these circumstances upon the terran, you must remember several key rules.
Rule 1: unless terran uses mass dropships, he must move overland throughout the game, often through terrain that is unfavourable. The opportunities do exist.
Rule 2: in order to locate these opportunities, use observers to scout them out. Place observers as I have said previously. That is, observers patrolling the empty mains and their naturals, and at least 2 and as many as 5 observers in the middle of the map.
Rule 3: In order to be ready to capitalise on these opportunites, keep forces in places that terran does not know. Move forces around frequently, and seemingly to no real purpose. A smart terran will always scan your army before moving his own in order to establish what is safe for him. Therefore, make his scanner information obsolete. Anytime you notice that your army has been scanned, move it somewhere else, break it up or form it up differently then before. Do this occasionally even if you don't see that you have been scanned, for the enemy is always scanning you and you will not know every time it happens. Also you may be moving into a position that he previously scanned and now assumes is empty and safe, when really your forces are now awaiting him there. By changing strategic conditions constantly, you take away terran's greatest advantage, namely his ability to easily force you to engage him on his own terms. By moving around often and making his scanner information obsolete, you force him to waste mana locating your army far more then he would like, and take mana away from scanning that could be used to detect your tech or expansions.
Rule 4: proactively create favourable circumstances whenever possible. How? If he has forces moving across vulnerable terrain into a safe position, intercept those forces before they arrive. If he is launching an attack on your mineral only with a well emplaced siege, find a way to move an army out of the threatened area and into the middle of the map where you can safely cut off reinforcements. This can be done by retreating towards another main when he first approaches, rather then your own main. It can also be done by using shuttles to ferry troops. You can then use these 'sprung' troops to occupy the center of the map (unexpectedly if at all possible). He will be unable to reinforce his position since your troops in the center of the map have every tactical advantage. You can then flank the terran containment force with superiour numbers, creating advantage where before there was an untenable position. Another way to force circumstances on the terran is to attack a target that he must defend. Whenever the terran attempts to sieze another main, you must attack that attempt right away. This will create a disadvantageous situation for the terran, for he now must fight you in several places at once. That means that his forces cannot remain stationary but must move around to defend seperate areas. Therefore his forces will be moving around in the middle of the map in a predictable path, which makes it very easy to ambush him. That is the art of strategy in Starcraft. Finally, by staying ahead in expansions at all times, you force the terran to attack you, and any time he attacks you, he must first open himself to vulnerability that you can exploit with a good ambush.
That is all I have to write about the Art of Starcraft. The remaining verses and chapters in the Art of War refer only to real life warfare, covering such subjects as terrain and psychology of manipulating real men, not game sprites.
