EPT=>How to make a game plan
=>What to do when unexpected things happen
The focus is on how to think instead of what to do. It presupposes only shallow understanding of the game, yet I write it such that it can also benefit high-ranked players. Tips are accompanied by progamer-class examples.
Edit: The article has been somewhat improved to address misunderstandings. Also, a couple inaccuracies were fixed; thanks to those who pointed them out.
Macro does matter
This is so important that I shall mention it first.
The most commonly given advice on the Strategy forum by far is: "Improve your macro." While not always the cause of defeat, it is always pertinent, therefore I state it here, surely to the satisfaction of many that are tired of repeating it. Really, don't waste too much time improving your strategy if your macro is deficient; improve your macro first. Strategy rarely makes the difference if one cannot cope. Sad but true.
"No you don't!!!"
Suppose we're roleplaying. You're a fighter facing a mage about to cast a fireball. What do you do?
a) Charge and strike the mage down
b) Run for cover and brace yourself
The good answer is obviously a), yet in Starcraft we tend to instead choose b), especially facing fearsome builds like 3-hatch mutas. Here's a notable example:
2008/02/14 GomTV MSL S4 Ro8 Group C Jaedong vs Flash Set 2 on Blue Storm
+ Show Spoiler +
Flash braces himself preparing for the inevitable while Jaedong takes the whole map. Result: one-sided massacre.
So what is the best counter to that build? When these two meet again, Flash has the answer: the firebat push. Some will object that firebats don't fare well against mutalisks, but that's irrelevant, as the following game shows:
2008/02/15 Bacchus OSL 2008 Ro8 Week 1 Group A Jaedong vs Flash Set 1 on Troy
+ Show Spoiler +
Flash has every reason to expect that Jaedong would once again pick his favourite build, and that he would have become complacent enough to neglect his defense. He just rushes with marines and medics, then pumps out several firebats. Some mutalisks come out but too little too late. gg.
This is why nowadays Zerg players build up to 7-8 sunken colonies at their natural, whenever they feel some pressure, instead of just 2-3.
This is why nowadays Zerg players build up to 7-8 sunken colonies at their natural, whenever they feel some pressure, instead of just 2-3.
Similarly, what is the best counter to the Bisu build? You guessed it, the hydralisk push:
2008/01/23 Bacchus OSL 2008 Ro16 Week 2-1 Group C Bisu vs July on Blue Storm
+ Show Spoiler +
This is indeed the infamous game that July wins by decision because Bisu's connection dropped. Still, July had clearly won.
The weakness of any tech build, especially two-bases ones, is that they are very vulnerable to early game pushes. If you expect your opponent to go that way, consider charging instead of defending!
Unfortunately, this doesn't always work; the map, for example, may not allow it. Here's a well-known example ending in spectacular failure:
2008/06/01 Razer TSL 2008 Finals Draco vs JianFei Set 5
+ Show Spoiler +
Don't try forcing your way up a ramp that way. It doesn't work.
so you have to use your judgement in those matters. Yet bracing yourself for enemy harrass remains only second-best; offense is the best defense.
That won't work? Switch tech
Few would argue nowadays that mechanics don't make the difference in Starcraft. To rise among the best, players have to train themselves for thousands of hours until they become macro machines. They can become so strong that even if one outsmarts them, they outmacro their opponent and win nonetheless.
Yet this comes at an expense: the more one practices mechanics, the more difficult it becomes to deviate from practiced scenarios. That may lead even progamers to persist in situations when it is clear that they should change their game plan, as these two games illustrate:
2008/10/17 Incruit OSL 2008 Ro4 Group A GGPlay vs Fantasy Set 1 on Plasma
+ Show Spoiler +
GGPlay opens 2-hatch mutas, planning to go for the kill quickly, but Fantasy wisely makes plenty of goliaths and turrets to counter these. He should then save his mutalisks for harrass and scouting, and instead defend, expand, tech to lurkers and cracklings. You don't send mutalisks to attack goliaths, right? Wait, he does, and obviously that fails; the mutalisks melt. gg.
2008/10/17 Incruit OSL 2008 Ro4 Group A GGPlay vs Fantasy Set 2 on Medusa
+ Show Spoiler +
While this game is famous for the praise that Fantasy received for his play [!?], I shall instead focus on GGPlay's. He opens with hydralisks, then proceeds to add mutalisks, planning once again to end the game quickly. But then he sees the valkyries. And the goliaths, again. He sure knows that sending mutalisks against the best anti-air that the Terran arsenal has to offer isn't effective; the outcome will be the same as in the previous game. He also knows that Fantasy's valkyries were hunting down his overlords, that he won't break his opponent with hydralisks alone, that his strategy will fall short. The situation commands to relent, switch to lurkers, cracklings, defilers. But he stubbornly chooses simplification anyway, against the odds, and loses. Again.
The above underlines the virtues of patience and adapting to the opponent's play. In those situation, one has to take a deep breath and change plans. You might think that it's no longer possible, but that's irrelevant; you have to try, otherwise you're dead anyway. Don't suicide your units by going all out against the odds.
Build orders: keep it simple
As I stated before, this guide isn't about providing build orders. Yet we shall briefly discuss what makes a build order viable.
Let's enumerate a few successful build orders:
=>2/3-hatch mutas
=>3-gate +1 speed zealots
=>speed zealots and archons
=>1-base carriers
=>2-fact vulture push
=>1-base reaver+shuttle
These builds have something in common: they name the very thing that they aim to produce. The key to devising at least half-decent builds is to focus on something simple but worthwhile and make it happen as soon as possible. In other words, build only:
=>The absolutely required tech buildings
=>The bare minimum defense to hold until the goal is achieved
Everything else is pork! Low-ranked players often lose trying to produce everything at once without clear focus, and lose to even a late push while hardly having any units. Really, forget about researching three techs and four upgrades at once, forget about making a third base (or even a second, if you don't need it) thinking it would otherwise hurt your economy later on, forget about pumping out units you don't need or static defenses beyond the strict minimum. Do whatever you can to get your shit out ASAP!!!!
For example, some think that one-base tech is usually not viable. But in many cases delaying the expansion to get the tech sooner makes all the difference:
2008/11/06 ClubDay Online MSL 2008 Ro8 Group D Bisu vs FireBatHero Set 1 on Medusa
+ Show Spoiler +
Bisu goes 1-base reavers+shuttle and races to drops before his opponent builds enough turrets. FBH gets 0wN3d. gg. If Bisu had delayed his tech to instead expand, worrying about his economy, it wouldn't have worked nearly as well.
Should you want more examples, just watch games from help threads [H] in the Strategy forum. Haphazard, inefficient builds are arguably the prime cause of defeat for D-rank players.
Doomed to failure? Try something new
Remember Katrina? I'm sure veteran Terran players remember that map all too well, because it was considered imbalanced in favor of Protoss in TvP; the Protoss would just go carriers 100% of the time and exploit the map's cliffs to decimate goliaths with gosu micro.
Just as against 2/3-hatch mutas or the Bisu build (see above), players long chose the conservative, lose-slowly-but-surely approach of making tons of goliaths and lasting as long as they could to save face. Then one progamer notably attempted something different:
2008/02/12 GomTV MSL S4 Ro8 Group B JangBi vs UpMagic Set 2 on Katrina
[spoiler]UpMagic makes cloaked wraiths instead of going pure goliaths. Jangbi gets caught his pants down moving his carriers around the map with no observers!
In the end, UpMagic loses; it is difficult to capitalize on your gains with wraiths. But at least give him credit for having tried something instead of opting for the sure loss.[/spoiler]
Whenever a build is too successful for too long, players executing it become complacent and neglect crucial things. Try to exploit these weaknesses.
Your attempt may fail, but that doesn't diminish your merit. Some might find inspiration into Grand Admiral Thrawn's word from the Star Wars novel The Last Command by Timothy Zahn:
[spoiler]Luke Skywalker has just escaped Imperial capture, eluding Star Destroyer's Chimaera tractor beam by flying his X-Wing out of a detonating freighter. Grand Admiral Thrawn, who had the previous tractor beam operator summarily executed for his ineptitude, once again walks to the starboard seeking an explanation for his successor's failure.
Thrawn led the way to the aft stairway and descended to the starboard crew pit. He walked past the crewers at their consoles, past the officers standing stiffly behind them, and came to a halt at the control station for the starboard tractor beam. "Your name," he said quietly to the young man standing at rigid attention there.
"Ensign Mithel," the other said, his face pale but composed. The expression of a man facing his death.
"Tell me what happened, Ensign."
Mithel swallowed. "Sir, I had just established a positive lock on the freighter when it broke up into a cluster of trac-reflecting particles. The targeting system tried to lock on all of them at once and went into a loop snarl."
"And what did you do?"
"I--sir, I knew that if I waited for the particles to dissipate normally, the target starfighter would be out of range. So I tried to dissipate them myself by shifting the tractor beam into sheer-plane mode."
"It didn't work."
A quiet sigh slipped through Mithel's lips. "No, sir. The target-lock system couldn't handle it. It froze up completely."
"Yes." Thrawn cocked his head slightly. "You've had a few moments now to consider your actions, Ensign. Can you think of anything you should have done instead?"
The young man's lips twitched. "No, sir. I'm sorry, but I can't. I don't remember anything in the manual that covers this kind of situation."
Thrawn nodded. "Correct," he agreed. "There isn't anything. Several methods have been suggested over the past few decades for counteracting the covert shroud gambit, none of which has ever been made practical. Yours was one of the most innovative attempts, particularly given how little time you had to come up with it. The fact that it failed does not in any way diminish that."
A look of cautious disbelief was starting to edge into Mithel's face. "Sir?"
"The Empire needs quick and creative minds," Thrawn said. "You're hereby promoted to Lieutenant... and your first assignment is to find a way to break a covert shroud. After their success here, the Rebellion may try the gambit again."
"Yes, sir," Mithel breathed, the color starting to come back into his face. "I--thank you sir."
"Congratulations, Lieutenant Mithel." Thrawn nodded to him, then turned to Pellaeon. "The bridge is yours, Captain. Resume your scheduled flight. I'll be in the command room if you need me."[/spoiler]
Don't give up just yet!
To word it differently: Don't leave the game while you still have units. Really. At the very least, consider making one last push; you never know, it might work.
Another option is to defend and cope with the opponent's attacks. Buy time. You might just be able to turn the game around.
Here's a short list of famous comebacks:
2007/06/24 GomTV MSL S2 Ro8 Group B Savior vs FirebatHero Set 1 on Python
[spoiler]After a protracted fight, Savior succeeds in breaking into FBH's main. You'd think this is over, but not FBH; he instead relocates his buildings to an expansion and entrenches there, microing like a god. He doesn't stand a chance, does he? He's just making the game last longer... Wait, is that possible? Savior runs out of mineral!!! FBH secures the very last mineral spot on the map!!! FBH WINS!!!!!
Tell me with a straight face that you saw it coming the first time you watched this game.[/spoiler]
2003/08/15 MyCube OSL 2003 Ro16 Group C Game 3 Boxer vs Joyo on Paradoxxx
[spoiler]Boxer drops into Joyo's base but loses all his goliaths to psionic storms, then carriers with dragoons and arbiters wreak havok to his base. Expect him to resign any moment now... Wait, it seems he wants to play to the end, after all. He persists at fighting a losing game until... Joyo runs out of mineral!!! Boxer denies him the last mineral spot that he could secure!!! Unbelieveable, BOXER WINS!!!!![/spoiler]
2008.08.17 WCG Korea 2008 3rd Place Match Much vs Luxury Set 1 on Andromeda
[spoiler]By midgame Luxury had secured map dominance, which in PvZ spells doom on the Protoss player. But Much doesn't concede. Watch him sweat like a pig desperately struggling to turn the game around, withstanding endless waves of Zerg units, until...
... he loses. Admit it, I had you big time.
Still, who would have thought that Much could even last that long? It's the kind of games that makes us push back our limits even when we lose them, so they're worth playing to the end.[/spoiler]
Conceding too early is the lamest way to lose. Even if you think this is over, try something desperate. Whether it works or not, it's worth it.
Conclusion
My goal was to convince you that, between players with comparable mechanics, strategy and innovation can make a huge difference; in fact, it may even overcome a moderate deficit in mechanics. If I succeeded in that, I didn't waste my time.
If you're looking for more games and advice, I recommend watching the games from the Spirit Tournament, commented by Artosis and ex-coach Dan. To repeat the latter's recurrent advice:
=>"Master the Basics!"
=>"Read Sun-Tzu's Art of War!"
=>"Know Yourself and Know your Enemy!"
=>"Be Like a Ninja!"
) corsair/ht mixed with the usual protoss ground force to combat the terran push. The reasoning behind that was the stargate and fleet beacon would tip off the terran that you were going carriers, so he would mass goliaths. Then you build high templars to counter the inevitable mass goliaths, and corsairs with disruption web to help against the tanks. The point is to get him to waste money massing goliaths instead of more tanks. Also, if he doesn't scout it or doesn't switch to goliaths, you can easily tech switch to carriers since you already have the tech required for them. Arbiters are just a tribunal away as well.
