God of the Battlefield: Part 1

Text byVer

May 24th, 2011 20:36 GMT

Introduction

During 2006-2007, Ma Jae Yoon (aka Savior) was the best Brood War player in the world. Not only did he win multiple championships, but he dominated his opposition in a way that had never been seen for the Zerg race. His play captivated the hearts of countless fans, and even the staunchest critics were forced to admit that he was one of the greatest of all time.

His career began on a sharp decline in March 2007, after a loss to Kim Taek Yong (Bisu) in the MSL Final. Though he would retain the loyalty and support of many fans - and occasionally show them flashes of his past brilliance - he would never challenge for a title again.

In 2010, Savior was implicated in one of the most devastating events in e-Sports history, the pro-gamer match fixing scandal. Nearly a dozen progamers had received bribes to intentionally lose their matches, which were bet upon by a handful of shady parties on illegal e-Sports gambling websites. Korean courts found Savior guilty of being a broker, a middle-man between the gamblers and the bribed progamers, though it is suspected that he may have been more directly involved as well.

Savior was sentenced to one year in prison (this sentence has been suspended for 2 years; sentencing occurred in May 2010). Furthermore, all of the progamers implicated were given life-time competition bans by KeSPA (the governing body for BW e-Sports in Korea). GomTV employees have mentioned that the banned BW players would not be allowed to compete in the GSL as well.

Though Savior seriously undermined the institution of e-Sports through his despicable crimes, we must be able to separate his faults as a human being from his accomplishments as a player. Though he will rightfully never play another game of professional Brood War, he left a significant, lasting mark on the way the game is played today.

Over a year ago, Ver wrote an excellent analysis of Savior's ZvT play at its apex, but the advent of the match-fixing scandal made us wary of posting an article praising him. Now that a year has passed since the match fixing scandal, and wounds have begun to heal (though they never will, completely), we decided we were ready to publish the following article.

The God Of the Battlefield: Part One

Once among the most loved of all champions, Savior now is loathed as a true villain. The tainting of his formerly great legacy happened almost overnight. Gone is the once overwhelming desire to see his return, being replaced by scorn and mocking.

But the match fixing cannot override or disqualify the beauty and quality of Savior's play. Whatever actions he may have taken outside the game do not change what he did within it, for it. Savior is probably the most studied player in history, yet to this day almost everyone, even pros, do not realize the full extent of his discoveries. To demonstrate why his innovations made such a large impact, it is first necessary to understand the unmatched obstacles that Savior surpassed. Following that, the gritty details of his innovations will be analyzed to show exactly what and how Savior did was so unique.

TvZ is something of a unique matchup in Broodwar. On the whole it looks innocent enough, but there has been one pattern firmly etched into Broodwar's history: top Zergs lose to top Terrans. In Starleague finals Terrans have emerged victorious 16 times to 7 against Zergs. The difference is most striking in comparing the very best themselves.

First examine Boxer (54-16, 77.14% over 28 months), Nada (112-47, 70.44% over 34 months), iloveoov (37-0, 100% over 13 months), and Flash (50-12, 80.65% over 10 months) for TvZ. Sit down, gaze in awe, and bask in their greatness. Not only are their records incredible or even unbelievable, the quality of their opponents and the lengths of their domination are also peerless.

Now take a look at their counterparts: Yellow (70-32, 68.63% over 23 months) July (40-29, 57.97% over 11 months), Gorush (27-11, 71.05% over 5 months), and Jaedong (42-19, 68.85% over 12 months). Barely even comparable. The best of the Zergs is not at the level of the worst of the Terrans with respect to TvZ. Only Gorush is remotely near the Terran four in his success against the highest quality of opposition, but his was also by far the shortest duration of dominance at 5 months. If you drop down to merely excellent players at the matchup, it would take a full paragraph to list the two dozen or so TvZ all-stars while you can count on at most two hands the number of Zerg players who have been truly good at ZvT. That is the reality of ZvT history.

Savior is the lone exception to the fate of his brethren, the only one to ever flip ZvT upside down. It was his ZvP that was most talked about because it was statistically better, but his ZvT was by far his defining attribute. Not only was he the only Zerg to truly succeed against Terran, he flat out dominated in by far the hardest conditions any player has ever faced. His play was so astonishing and of bottomless depth that it is with complete confidence that I can make the following claim:

The peak level of ZvT was reached by Savior in February and March 2007. It is not to Jaedong, Effort, or Zero that we should be looking, but Ma Bonjwa himself. Savior's play was so beyond his time that for years everyone studied his games yet never understood half of what he did.

Who comes close in series dominance? Boxer? Jaedong? Bisu? Flash? Nope, none are remotely comparable except Flash in TvT against pathetic opposition (11-2). Only the two greatest players ever, iloveoov and Nada, can come near Savior with 11-0 series win periods in TvT (oov) and TvZ (Nada) against a succession of strong opponents. But Savior's series dominance is not just measured simply by the number of his victories, but by the quality of his opponents and difficulty of circumstances under which they were achieved. Savior was the most dominating series player in the history of SC and his ZvT showed that better than anything else.

The only players on that list above who didn't fit in tip-top S-class TvZ levels were Yooi and Justin. Everyone else, including the statistically out-of-place Iris, made the 6 dragons look like pushovers. If you had to gather the best lineup of players at one matchup of any period, nobody would come close to these 7.

Savior's performance is even more out of place next to the fact that only 4 out of the other top 10 Zergs held a >50% ZvT win rate from January 2006 to March 2007! And only one out of those 4 was actually winning games against Starleague-caliber opponents. Considering that a Terran player won every single league that Savior didn't participate in during the period, it's pretty safe to say that Savior was the only reason why that era was not the most dominating for Terran and worst for Zerg in history.

The King of ZvT

Because Savior reached six Starleague finals in a row, he played solely versus the absolute toughest competition possible. As the score sheet shows, he repeatedly beat down the hottest Terran players of the day as well as the other three bonjwas at their best matchup. How many Zergs besides Savior does Nada have a losing record against (with a minimum of 5 games)? Three. Boxer? Zero. iloveoov? Zero. Yet Savior decisively defeated them all by unheard of margins when they were still dominating every other Zerg.

To further magnify his achievements, Savior played seven of those series within two months! The worst moment was when he played two Bo5's against Hwasin and Iris on consecutive days. Suffice to say, no player has ever went through as intense of a struggle as Savior did from January to March 2007.

This was the worst semifinals I've had go to through. The maps were tough, and my schedule was very tiring as well. I don't even have time to be happy that I reached 5 consecutive finals because I am so busy and can't think about other things. Today was the semifinals and tomorrow is the semifinals too.

-Savior after defeating Hwasin

Because he was stuck in the same matchup on the same maps, his opponents had unparalleled mounds of data to develop counterstrategies with while Savior alone had to incessantly innovate to avoid being overtaken by every other player working to unseat him. S-class pros are terrifyingly good at dissecting builds and preparing appropriate counters, yet Savior never allowed them that chance. To avoid becoming predictable and to prevent getting gunned down by precise timing attacks, Savior literally did something new every single game and continuously stayed ahead of the competition, something no other player has ever managed to do even in much easier circumstances.

Even with his star-studded opponents and overwhelming scheduling still not enough to bring him down, Savior's enemies had one final trump card: the mapmakers.

I think even I can beat a zerg user using terran on that map...it is difficult to beat even our team’s practice partners on that map…

-Savior, on Longinus 2

It is bad for Zerg. Not only do I feel bad after training for that map, but during the training I lose smooth(ly) [inevitably]

-Savior, on Reverse Temple

Hwasin predicted that if the game extended beyond mid-game, then NaDa would win. Reverse Temple was just that good of a map for Terran users.

Iris also came to the same conclusion. No matter what starting position the Terran was at, Terran would be favored and Savior would have a hard time on Reverse Temple.

Contrary to the rest of his colleagues, Savior showed his very best in the worst conditions of them all. Reverse Temple and Longinus are the icons of Savior's legacy, for they best represent the trials Savior had to overcome. It was here that Ma Jae Yoon did what was thought to be impossible over a dozen times. Not only did he win on these maps, but he won in decisive fashion and made them look advantageous for Zerg while doing so. Not to mention nearly all of his games were against some of the best Terrans ever.

You can't emphasize enough how much you need to be a paradigm shifter.

-Savior

One can at first interpret Savior's ZvT success to his revolutionary new strategies, and this conclusion is partially valid. When Savior first emerged with his new strategy of 3 Hatch Muta, pumping drones, and transitioning to lurkers while double expanding into 4 gas defiler, he managed to outsupply Nada while having defilers out before even fighting a battle! That's how far ahead Savior was of everyone else in terms of strategy, and he owed much, if not all of his early success – particularly the mind blowing 5-0 against iloveoov – to the strategic advantages of his many ZvT revolutions.

Essentially every SC star has emerged sporting something new and shiny, whether it be revolutionary mechanics or a completely new concept/strategy. The more radical the idea or execution, the greater the subsequent dominance. Boxer, Nada, iloveoov, and Flash, in TvZ, Bisu in PvZ, Jaedong in ZvZ, and Savior in ZvP are just a few among many other examples. While all of these players initially won consistently because their opponents had no clue how to play against them, they experienced a slow, or sometimes rapid decline when their rivals figured out how to counter or equalize against the previously dominant strategy. This has been, and will always be the case. Innovations only have a limited amount of time to be made use of before they become another draw-even strategy or obsolete altogether. And once that period of time is over, that player always declines in performance. Only Savior is the exception to this rule.

Savior's ZvT dominance began because of his new strategy of 3 Hatch Muta into lurkers covering a double expansion and massing drones, all with a fast hive. While it sounds similar to the more familiar variation everyone should know, it is actually conceptually different in many ways as it was optimized against the slower expansion builds of 2 Rax Academy CC and 2 Rax CC. He unveiled a mostly finished product on April of 2005, but in January of 2006 he was already running into trouble. Thanks to iloveoov and Midas in June of 2006 his original strategy was completely outdated after only 21 games.

Everyone in the spotlight eventually gets gunned down. Well, almost everyone...

But Terrans are not the kind of race to stop at just keeping up with Savior's brand new concept with a faster expansion. No, iloveoov, Casy, Midas, Nada, Hwasin, and Iris tag teamed Zergs to oblivion by correcting their sense of timing, reactions to mutalisks, tech transitions, and totally transformed their anti-hive play. All of these changes were so far-reaching that despite originating in 2006, they are still mostly the standard today. It looked like Savior would fall victim to what had happened to every other upstart player once their strategies were figured out. And yet with his brilliant strategy in shambles and his play found out and outdated, Savior actually started winning more despite becoming increasingly handicapped by maps and scheduling!

How did he do it? Well quite simply, he improved faster in other, unseen ways. The collective mind of every Terran, by far the most innovative race in SC history, failed to surpass him despite having an unprecedented amount of opportunities to test Savior. With every other Zerg lagging so far behind, it was literally Savior against the World, and Savior won. Not only did he create the strategic equality against Terran in both economics and decision making that has survived to this day, but he advanced a much lagging area of ZvT: tactics. Much to their downfall, Zergs have rigorously copied Savior's strategic innovations but have mostly overlooked his tactical ones. Tactics separated Savior from every other Zerg, and they were the primary reason why he reached the highest level of ZvT three years ago.

What are Savior's tactics? To keep matters as simple as possible, "tactics" is simply a broad term I use to explain Savior's battlefield actions because they do not fit under any previous concept. Tactics are more of a concept dependent on creativity and psychology than a rigid set of actions. Different situations call for different tactics, and the examples picked out in these two articles are simply a small portion of the many that Savior used. Hopefully by examining various examples of Savior using certain tactics in certain circumstances, one can get a glimpse at his terrifying and comprehensive understanding of the matchup.

Still confused at what tactics are? Here's a helpful demonstration by the Maestro himself:

Bonjwa Training Step 1: Hidden Lurker Ambush

Not good enough?

*Note that a video link fast-forwarded to each tactical scene will be provided as seeing them in action is essential.

The central principle that all of the maneuvering tactics are based around is simply having more, properly placed units at the right place at the right time. Despite being almost always outnumbered, Savior generally had the upper hand when the two sides came to a clash. By keeping this principle in mind at all times, it will be much easier to understand just how his tactics work.

The point here is not to focus on the individual tactics themselves. No one tactic is as important as the overall mindset behind it, for different situations call for different tactics and the number of possibilities are only limited by creativity. Preparing for every conceivable plan and movement his opponents could throw at him was impractical. Rather, Savior could instantaneously choose the best suited tactic to any unforeseen situations.

The first, and perhaps most basic principle of ZvT maneuvering is to form a mobile, multi-purpose army that is free to roam around the map. The worst outcome in ZvT is to have the Zerg army trapped in its base as the Terran slowly batters down the castle. Savior would often use a line of sunkens and a few lurkers to defend his bases from any immediate blitz, while scattering his army far away from his bases. By doing so, Savior magnified his army's strength by several times. He ended up using more sunkens than other Zergs, but the benefits of his maneuvering always proved to be well worth the investment.

Savior vs Iris on Reverse Temple - VOD

Iris goes to attack Savior's base while Savior runs his army to a key location.

Savior's army here is doing 5 things:

1) Threatening a backstab on Iris's main (Savior's winning technique in this game)
2) Threatening an attack on Iris's army at any sign of weakness
3) Threatening to cut off reinforcements at a moment's notice
4) Defending Savior's natural and 3rd base
5) Hiding in the fog and staying completely out of reach of Iris's attack capabilities

All five of these points are serious threats that his opponent must devote time and concentration to defending. Iris cannot have constant information of Savior's army location and movements, and thus has no way to know his intentions. Ergo, simply by running his army around instead of keeping it trapped inside his base, Savior forced Iris to consider every possibility and therefore greatly improved his situation without killing a single unit.

Small details like the precise positioning of his army were absolutely key to Savior's success.

Savior vs iloveoov on Ride of Valkyries - VOD

Savior's game here is so remarkable because it so aptly illustrates how Savior used his tactics to allow him to cut many corners on timings, which was very important against the cheater terran (named for his ability to make a seemingly unfair amount of units) himself. After helping himself to a comfortable early lead through a new 8 rax -> CC strategy, oov begins a threatening attack with 5 tanks, 40 infantry, and 2 vessels against a paltry 12 mutas and 7 lurkers. Such a unit imbalance all but assured oov an easy victory, if only he could actually fight a battle. Swooping in and out with his mutalisks and reburrowing his lurkers, Savior forced oov to continually regroup and advance at a snail's pace. By the time a battle was actually fought, Savior's army had more than doubled in size, and despite oov's reinforcements the Terran menace was halted.

Savior's dismantling of oov showed the growth of Savior's tactics. oov, always the clumsy and mechanically handicapped gorilla, was incapable of multitasking and his attack was unnecessarily delayed by Savior's shenanigans. As for Savior, he succeeded in delaying oov's attack sufficiently but he did not stop oov from reinforcing his army. Thus he halted the monster, but could not crush him. This meeting between Savior and oov occured in January of 2006. Now let's fast forward and see how much more refined Savior was 13 months later.

Savior vs Iris on Longinus II - VOD

Along with having a mobile army and delaying the Terran timing attack, the other basic step for all levels of ZvT is to prevent the Terran from reinforcing their main army. In other words, once that Terran ball moves into the middle, it becomes isolated from civilization—a suicide squad, so to speak. These three steps are all that are needed in order to achieve relative superiority at the decisive battle, the entire goal of midgame ZvT.

This game is a fantastic example of how effective maneuvering can be. Iris began his attack at 10:39, and given how close the mains are on Longinus he could have arrived at Savior's natural entrance, the critical area, in 25 seconds. But Iris never made it to Savior's natural; his army was annihilated as soon as dark swarm was finished at 14:04! In other words, Savior with a much inferior army delayed a 25 second trip to over 3 minutes and 25 seconds! By making many different threats and forcing Iris to respond to them all, Iris ended up marching around in circles and never reached his destination.

How did Savior turn a simple expedition into the trip from hell? Let's go through step by step:

+ Show Spoiler [Explanation] +

The first key point to remember is how Savior won their previous game in the series on Reverse Temple, featured above. There, a seemingly inevitable Iris win thanks to impossible positions and no muta damage was thwarted by a perfectly-timed backstab. On the equally Terran-favored Longinus 2, Iris would be sure to not make the same mistake twice and Savior knew that. His backstabs were many times a game winner, but they were just as strong as a threat alone.

Savior threatens a fake backstab as soon as Iris runs out, forcing Iris to turn southward to block it before resuming his northern march.

Savior deploys two high ground lurkers to delay Iris's initial advance and to punish any sloppiness.

Savior threatens a backstab as soon as Iris clears the way and starts advancing northward.

Savior moves his army to block Iris's southern advance and snipes two careless Vessels.

Iris gives up marching back and forth and just storms through and advances through the middle to the northwest ramp, preparing to assault Savior's natural. Savior continues to threaten another backstab and kills some stragglers.

Savior cuts off all reinforcement routes and keeps his army just out of reach in the shadows. At any point it can attack Iris's army out of the darkness from every angle.

These Lurkers prevent Iris from reinforcing, killing some units when Iris isn’t looking.

As soon as Iris's reinforcements clash with the retreat blocking lurkers Savior launches a pincer on Iris's now isolated army.

Iris, instantly recognizing Savior's movements, barely escapes the trap intact, retreating and reorganizing his army. Yet even though he saved his army, his position is no better than a minute earlier and the defiler clock is ticking down.

Iris is just outside of his destination but with no way to get reinforcements, moving quickly from the high ground is a grave risk when he doesn't know the location of the rest of Savior's army. Therefore he slowly pushes forward while keeping most tanks sieged and infantry in position.

Iris pulls back his units to chase back the imaginary threat of a group of lings on his rear. Another movement slowing down any advance possibility.

Savior plants his lurkers in order to prevent any retreats. They may seem useless at the moment but very shortly they will prove decisive.

The delaying is over, consume is researched! Savior begins the swarm assault on Iris's trapped army.

Note how his second swarm traps Iris in a tiny corner instead of pushing forward up the hill with 2 swarms and letting Iris potentially retreat.

The last stragglers of a once mighty army...

...get torn to pieces.

The main imbalance on Longinus, and the only map where this is possible, is that the Terran can actually make two timing attacks on favorable terrain before consume is out due to the extra income and rush distance. Preventing Iris from even reaching the sunkens given these conditions would be completely unthinkable if it were not Savior that was playing him.

The key to remember is that from Iris's point of view, Savior's units, particularly his lurkers, could be anywhere. At any moment they could burst from the darkness and ram into his natural, or if he moved from the high ground he could be flanked by a wave of units from that high ground. It is only natural that Iris became increasingly paranoid, moving all his infantry to chase a dozen lings or redirecting his whole army several times to run after some lurker/ling. After all, he never knew what lay just out of sight in the shadows or underground. Yet at the same time that paranoia was necessary in order to save his army from the potentially devastating ambush that Savior sprung. In order for Iris to even make use of his army he needed to deduce Savior's army movements from small shreds of information, so that he might guess whether Savior was truly threatening his troops or merely making a feint. Isn't that asking a little too much from poor Iris, or anyone for that matter?

A noteworthy contemporary performance was demonstrated by Hogil in this VOD. Hogil's maneuvers were not as refined or purposeful as Savior's but they were more than adequate to dispatch Mind's potentially devastating attack after Hogil had failed to do any damage with his mutalisks.

But what about an elite level modern Terran? Surely some fancy running around can't touch the most skilled Terran ever who has benefited from years of evolution? Savior's successor, Effort, showed that Flash was just as susceptible as Iris, Nada, or any of the others Savior repeatedly took apart when he decisively won on Heartbreak Ridge in this game. With neither side at an advantage, Effort succeeded in isolating and slowing Flash's attack purely with maneuvers. Flash moved out with a much superior army, but after a crisp maneuver half of Flash's total troop count was sitting at home, unable to link with the main force. After further delaying Flash's isolated army, consume was finished and Flash lost his army much like Iris. On such a Zerg map as Heartbreak Ridge, unlike Longinus II, losing your main army for nothing is akin to losing the game, and Effort finished Flash off within minutes.

While blocking reinforcements to later crush an attack is all well and good, there's an even better way to use lurkers.

Savior vs Nada on Tau Cross - VOD

The strength and viability of these listed tactics vary greatly. Some tactics are merely a useful stepping stone in the overall path to victory. On the other hand, some are so potent they can win an otherwise even game in one stroke. This particular tactic is the latter, where Savior seized victory in an instant despite having no visible advantage. Normally the Zerg wants to delay the Terran army and fight at the last minute possible, but here Savior had an opportunity for an instant win because he had the vital information of how much defense Nada left behind during his push (not enough!). Savior and Nada's positions looked equal, but that parity was thrown out the window when Savior pulled off this brilliant maneuver.

Savior pulls a win from thin air.

As soon as Nada walked a few seconds away from his base, Savior took control and ran his army in between Nada's army and his base. Savior burrowed half of his lurkers behind Nada's army, blocking his path home. Meanwhile the rest of his army crashed into Nada's natural, eviscerating his main and fully shutting down all means of production. By the time Nada cleared a path back home, his war machine was at a fraction of its old capacity and not surprisingly he was overrun a few minutes later.

Nada's base is overrun and his army cannot save it in time.

Savior vs Midas on Hitchhiker - VOD

Savior sacks Midas's expansion while Midas watches helplessly.

As soon as Midas moves out to attack Savior's expansion at 1 to the east, Savior raids Midas's 3rd base from the south entrance. Despite the fact that Midas was seconds away, he could only stand by and watch helplessly. This attack was not a game winner because Savior did not deal properly with the Science Vessels, but it inflicted a disproportionate amount of damage considering the disparity in army strength. Midas could have easily crushed Savior's army in a straight up fight, but Savior's maneuvering and position gave him much better results than simply attacking head on.

Note that Savior gets several free marine kills on Midas's rescue army because Midas is too busy microing his men at his base that he doesn't change his previous attack order quickly enough. Troop losses like these are not an error; they are simply a result of Savior overloading his opponent and then putting himself in a position to take advantage of it with no downside.

At first glance this tactic doesn't seem particularly amazing. After all, it appears to be just a fancier version of backstabbing. However, it's actually quite different. Normally when the Zerg army backstabs they have to wait for the Terran army to move a fair distance away, far enough that it cannot easily return to defend against a backstab. However, by the time the backstab arrives, the Terran will have already made several rounds of reinforcements to defend. On the other hand, if the Zerg attacks immediately as the Terran leaves home, they might find a couple marines drunk on the job instead of a well manned barricade waiting to greet them.

In addition, it finely manipulates the psychology of his opponents. Because the army blockade backstab can strike when the Terran army is close to home, the Terran feels compelled to save their base but cannot due to the lurkers. Even if he did want to counterattack, there is always indecision and delay in making a decision because of the instinct to defend given the army's proximity to home. Even if they do finally decide to counter, their army is quite a distance away. Normal backstabs carry the risk that the backstabbing force can eventually be mopped up while the Terran counterattacking army razes the Zerg main. However, the psychological upsides of the army blockade variation almost entirely removes this risk, making it a very potent maneuver. It was partially because of this maneuver that Savior's threat of backstabs had such an enormous effect.

Some keen observers might remember that Luxury used this exact same tactic to mount an incredible comeback against Leta on Carthage in the Lost Saga MSL here.

Simultaneous attacks are no foreign concept to high level Brood War. The Protoss army that engages to pull attention away from a dark templar drop in the Zergs main, the Terran army that fights at the Zerg's fourth gas while simultaneously dropping the Nydus at the third; both of these situations and similar ones occur at least semi-frequently and are well incorporated into the text book play of other races. Zergs, however, have restricted their usage of this principle to ZvP only and use simultaneous attacks far too infrequently in ZvT.

Think about all of Flash's late game TvZs and how one-dimensionally all of his opponents try to approach it. They repeatedly gather their masses of ultraling with defilers and just try to punch through his lines a few times, then follow it up by getting punched out themselves after their ultra/ling evaporates to 20 tanks. Sure once in awhile they might try a drop far too late, but they never think about trying to create multiple side attacks in order to increase the success of their big blow.

Behind this seemingly simple tactic lies its very counterintuitive nature: Savior accomplishes more in the big battle despite using less of his army! Diverting a few ultras and maybe a defiler with some lings here and there ends up magnifying his real army's strength despite fighting with less units simply because of the strain it causes on his opponent.

Savior vs Nada on Longinus II - VOD

+ Show Spoiler [Explanation] +

Savior drops a handful in the main to start off.

Hidden units spring into action in Nada's fourth.

Nada's army and vessels are all out of position.

While Savior would likely have won the fight anyway, it is well worth watching the sequence of events to see how little he lost for all his gains. Nada's tanks were left unsupported by his infantry and his vessels were all out of position because they were initially running over to go help defend the first two threatened areas when the main army hit Nada's exposed third.

Savior vs iloveoov on Ride of Valkyries - VOD

While most of his other matches where Savior used simultaneous attacks were against opponents who were playing aggressively due to the constraints of the maps and strategies, he also proved more than capable of winning against the original creator of Flash's now immortalized TvZ defensive setup. In this game from the very beginning of 2006! Savior illustrates exactly how every Zerg should be playing the late game against Flash and company.

+ Show Spoiler [Explanation] +

Savior prepares for his center attack with a powerful drop that does almost nothing against the well prepared oov.

The most epic attack you will ever see...and a complete flop.

In preparing for round 2, oov must still keep a force solely for the threat of another drop. Note how the drop was a total disaster but the mere threat of it gives Savior a significant advantage in the next wave. oov's sorely needed men were busy stuck at home defending the mere possibility of a drop.

Savior sneaks two ultras around the top before the attack, forcing oov to spend precious time running scvs, vessels, and units around. An important result of these two ultras is that oov saw them and instinctively sent his vessels and other reinforcements over to irradiate. However, that meant his vessels and units were not in position when Savior attacked a few seconds later, heavily contributing to his defeat.

With oov's forces and attention divided in three places, Savior finally breaks through in spectacular fashion.

While this game occured earlier in Savior's career and before he reached his peak level, it illustrates well how to attack a modern lategame tank-based defense. Multiple attacks spread across the map with liberal usage of drops help keep the Terran uncertain of the Zerg intentions and plans. Zergs must start fast and early with drops and simultaneous attacks instead of sitting back and eagerly massing up a big army to be used in one giant thrust. Above all the Zerg must not let the Terran have enough time to fully solidify their position. Had Savior waited any longer, he would have lost all chances once the middle became the Maginot line and alternative attacking options were blocked.

Not surprisingly, Effort was the first to dissect Flash in complete Savior fashion and gave him his first real lategame loss since he perfected the Terran defensive system. Admittedly Effort did have a solid advantage from his early game trickery but his approach is still correct; the advantage just made victory faster.

While similar enough to the previous tactic, counterattacks are really a whole different beast. It is one thing to plan and execute multiple attacks simultaneously, and quite another to react to an attack prepared at the enemy's discretion while planning and executing a counter at the very same moment. The very fact that Savior had the gall to send out sizable armies to counter when he was in a perilous position is commendable as it is, but the timing at which he coordinates both his and his opponent's actions is surreal.

Savior vs Nada on Longinus II - VOD

Savior's only expansion defense is shortly going to die.

As Nada is preparing to attack the 12 expansion, he irradiates all the lurkers and one defiler atop the ramp but does not irradiate the other despite having the energy. Even when the path is cleared, his army and vessels sit around needlessly for several crucial seconds, giving Savior precious time to ready his defense. However, this otherwise inexcusable error has a very good reason.

Rather than using his units to defend his key expansion, Savior goes on the attack!

Instead of concentrating his attention on making a potential game-winning attack, Nada is forced to retreat and defend his 4th base. While Nada is being stalled by moving SCVs, lifting his command center, and rerouting reinforcements to deal with the threat, he is not moving his units into the vulnerable expansion. On the other hand, Savior is doing damage and gaining the time he needs to mobilize his army and save the expansion.

As Nada was slowed too much to effectively raze the expansion, he is forced to defend the ramp and wear down Savior's army. For the second time he misses a critical irradiate on a reinforcing defiler and some potential others on ultras as the rampaging ultras invade his mineral only and force him to pull more units and SCVs. Furthermore, the other half of his army is sitting inactive in the middle despite having the ability to raze Savior's natural as Savior's army is out of position (as Nada does do 30 seconds later). Nada is simply overloaded at the wrong moment!

The result of little things adding up. Yes, those marines are plagued.

Nada deserves praise for the entire sequence as well, as nobody but Casy would have managed to pull off half of what he did. Defending and evacuating two bases while keeping up non-stop production and yet still planning and executing an otherwise game-winning attack is asking quite a lot. Nada did the best almost anyone could hope for in such a situation, but Savior's tactics were simply too much.

Savior vs Nada on Reverse Temple - VOD

How could this be possible?

As it appears at first glance, this situation is one of the most inexplicable mysteries of any finals. How could Nada let three lurkers devastate his only army and not react for 9 seconds!? Lack of scan energy is also not valid explanation as Nada would have stimmed and organized his units first. Pressure is not valid as Nada is the definition of clutch. No, Nada had not reacted at all. Why?

The cause.

The important point to notice is the exact timing of Savior's actions. He moves the 3 lurkers in attack range half a second before his muta/ling backstab comes into vision at Nada's natural.

The only plausible interpretation of the above sequence is that Nada was so surprised at Savior's backstab that it did not even come to mind that his army could be attacked during this time. Nada clearly reacted immediately to the backstab, pulling SCVs to surround the bunker. Yet normally Nada could have defended and macroed and we would never have noticed he wasn't moving his main army. Nada must have been so distracted that he did not even move his army forward to siege Savior's natural, much less do anything else. It wasn't that Nada was overrun by necessary tasks for 9 seconds; he had completely forgotten about his main army in the face of immediate and unexpected danger. Nada quit multitasking and focused all of his efforts on defending his base for those key seconds. A horrific mistake to be sure, but then again, no opponent ever played Savior and did not make glaring forced errors.

Why else would Savior potentially suicide 3 crucial lurkers at that exact timing if not in anticipation of the events that followed?

Savior vs Iris on Neo Arkanoid - VOD

Nothing can save the expansion.

Near the end of perhaps the greatest game of all time, Iris destroys Savior's critical Nydus and runs thirty infantry and two Vessels into Savior's undefended mineral line where he can quickly snipe his remaining drones and hatchery. Savior's only defense is a panicking defiler consuming drones left and right and two horribly out of place lurkers and hydras who cannot save it in time. The game is over once Iris moves down and snipes the hatchery, and there is nothing left to stop him. Yet right before he is about to do so, he suddenly turns back for no apparent reason!

Moments later, Iris retraces his steps and attempts the above plan but now two plagues are thrown out and Savior's lurkers have the time needed to be positioned properly. The once proud force that could have won the game fell apart and did absolutely no damage. In the entire game Iris made more than fifteen dangerous attacks of that nature and in no other case did he ever make such a glaring and unthinkable decision to pull back at the very moment of triumph. However, in no other point during the game did Savior make a sudden strike while Iris was also attacking.

Why did Iris not level the expansion?

Right at that heartbreaking moment for every Iris fan and Savior hater alike was a drop on Iris's 4th base at 6. Given that Iris couldn't really defend his expansion, he did not need to spend more than a glance on the drop: lifting the command center and running SCVs was all that was necessary. But either Iris panicked upon seeing the drop or his concentration was momentarily broken, it is impossible to know all the details. All that is for certain is if not for Savior's drop at that timing, Iris would have won the game.

To a certain degree the way that these examples turned out is simply due to luck. All of Savior's offensive actions were justified in their own right; it is the particularly perfect timing in a simultaneous defense/offense situation that is astounding. It is highly unlikely that Savior masterminded each and every tiny outcome of the event, and much more probable that he was simply setting up a scenario where his opponent would likely make a mistake of some kind that he could act on. A pure luck explanation, however, is not valid simply because Savior performed this kind of simultaneous timing successfully in so many different tactics in a wide variety of situations that it is quite clear there was a great deal of thought behind it.

Furthermore, it always turned out that Savior had allocated enough units to each respective task in order to assure success at all of them. Just like with the simultaneous attacks tactic, Savior was accomplishing more casualties for less cost at each point of conflict.

At first this tactic makes it seem like Savior was just out-multitasking Nada and Iris, which is quite impossible as Savior was one of the slowest champions ever. Savior was using two important factors to his advantage. First, Savior was the one initiating the counter actions, meaning that he was mentally prepared for what would happen. Nada and Iris were both caught completely off guard and these momentary lapses in their concentrations from unexpected sources created forced errors. Secondly, Savior was making use of racial advantages. In any battle a Terran army that does not stim and engage properly is going to suffer an enormous drop in output, while unmicroed ultralisks and already burrowed lurkers are essentially as good as microed ones. Therefore, Savior could spend all of his attention after the initial movement on the area which would demand more focus while his opponent had to go into triage, meaning that one army or another was going to fight very suboptimally. Oftentimes his opponent would do everything poorly simply because he wasn't expecting such a sudden overload of tasks.

His ever-present ability and incredibly varied means to forcibly make his opponents play worse, is without question, Savior's most unique and unmatched skill. You do not just play against Savior, Savior plays you.

Savior vs Light on Longinus 2 - VOD

A key factor in winning ZvT battles is fighting when the Terran is not in proper formation. Rather than straight up charging into the fray, it is often better to let the Terran weaken themselves before engaging in battle. Normally the timing is very difficult to ascertain and follow through on, but Savior used a simple yet effective idea to ensure the right timing: deliberately weakening his own position as bait.

Light's army is hopelessly out of position.

While Savior would normally prevent Light from taking the expansion's high ground, in this situation he created the opportunity to lead Light into a trap by not blocking the ramp. As soon as Light saw an open ramp, he seized the opportunity and bolted up. However, in his haste to seize the 'opportunity' Savior provided him, his formation disintegrated. For when Savior attacked scarcely a moment later, he fought against perhaps half the firepower Light had. Marines and tanks were cramped next to each other with no room to run. Tanks were unsieged, medics were out of place, many units were not able to attack all at once, and worse yet, the marines had no room to split or retreat. Once the lurkers burrowed, Light's army vanished.

The unsurprising result.

At first glance one might criticize Light's play here as walking willingly into a trap. But it was certainly not a poor play by Light nor a mistake against any other player. Had Light managed to safely get to the high ground, Savior's crucial expansion would have been toast and Light likely would have won. But because Savior intentionally lured him up by retreating his high ground units, Light lowered his guard and focused solely on maximizing his advantage, giving Savior all the opening he needed. This tactic is a good example of how Savior utilized normally correct decisions against his opponents. In virtually every one of these examples Savior's opponents were not making normally bad moves. It was simply Savior's genius that he could create and turn circumstances where the correct decision, one that any pro would instantly seize, was in actuality a seed of disaster. And when Terrans saw these chances and tried to take advantage, they opened a moment of weakness where Savior would crush them.

Savior vs Casy on Reverse Temple - VOD

In this much understudied match, Savior wrote the textbook of how to deal with an overpowering Terran midgame army. By maneuvering in a very calculated fashion, Savior ensured that Casy would have no choice but to enter a battle he could only lose. While some of the previous examples showed how to maneuver and attain relative superiority, this game featured the best maneuvering against a timing attack you will ever see against a highly skilled player who is making nearly all the correct decisions, without a single wasted movement from either side.

The pre-defiler timing attack is the most important moment of a typical 3 Hatch Muta vs 1 Rax FE game. At the highest levels of play, the outcome of the battle depends chiefly on how much damage the Zerg did with mutalisks. But Savior could not reliably do any damage with mutalisks against modern defenses, a normally fatal flaw on such a map as Reverse Temple. Yet despite his predicament, Savior annihilated Casy's army before Casy ever reached Savior's base, an otherwise impossible feat.

Savior makes a feint attack towards Casy's army.

In order to delay the Terran advance, Savior makes a feint towards the advancing army in the middle of the map to make them cautious and move forward slower. When Savior moves in, Casy is forced to siege up and pull his vessels back in response, for he has no way of knowing if it is a feint or real attack. Once Casy halts and prepares for battle, Savior then retreats and runs up and around Casy's army, preparing for phase two. The timing and placement of Savior's feint was carefully planned, for it was neither too close to Casy's target of Savior's natural nor too close to Casy's own base. As we shall see, fulfilling these two criteria was extremely important.

Let's think about the scenario from Casy's point of view: Savior charged but backed off without fighting. 'Clearly he feels that he can't win in the open field and is trying to buy time for more units and defilers. Since my army is momentarily paused I should use that period to reinforce my army to keep up my relative superiority while advancing towards Savior's natural, and not give him the time he needs to get defilers. Because it will take a little time to wear down the defenses of his natural my reinforcements will arrive just before he must attack, thus giving me the maximum possible advantage.' All of this is completely logical and the correct choice. And exactly what Savior predicted.

After Savior retreats from the feint, Casy naturally unsieges his army and moves towards Savior's natural while bringing forward his reinforcements. Meanwhile, Savior's army retreats up and around the temple walls and descends upon Casy's rear lines, running smack into Casy's reinforcements while Casy's army is too far away to help! All of a sudden Savior's intentions are clear.

Savior ambushes the reinforcements.

Losing some of his reinforcements for nothing puts Casy in a pickle. Not only does this lower his absolute troop count, but it forces him to cease sending more reinforcements in order to protect his under-defended natural from the possible threat of a backstab. If he had sent any more reinforcements, he might even have to retreat his army in order to protect the natural, buying Savior even more time. Or else, Savior can merrily charge into the Terran base, winning the game right there with a backstab. Either way, the Terran will be confronted with a big choice and likely will pause their movements in a moment of indecision, giving even more time and possibly a chance to attack. But Savior is not done yet...

Savior lays a trap for Casy's army but Casy is too quick.

Savior positioned two lurkers at a spot so that when Savior started to attack Casy's reinforcements, Casy's army would run into these two lurkers. Savior correctly reasoned that Casy's attention would be first drawn to the army attacking his reinforcements and the simultaneous movement of Casy's main army would recklessly run into the trap lurkers before Casy would realize and start microing. Unfortunately for him, Casy's inhuman awareness and speed let him micro and retreat his reinforcements, retreat his main army after only one marine was lost, and irradiate Savior's trap lurkers, all at the same time! While this trap ended in failure, it was a good example of threats Terrans always had to be prepared for at any time.

Savior's tactics are too much for Casy's micro.

Even though Casy avoided the lurker trap, his position was still grim. If he retreated, he would lose any opportunity to put pressure on the Zerg and give Savior freedom to make drones and expand faster than he should. If he didn't retreat, he would be forced to fight a battle while greatly outnumbered. Casy chose the best option, to press on, yet despite his stellar micro the overwhelming hordes overran his once mighty force and Casy's position quickly collapsed from there.

This series of maneuvers is the perfect demonstration of the great advantage that tactical mastery can give. Casy's relative army strength dwarfed Savior's when he first pushed out. Had Savior fought on Casy's terms right then and there, he would have been clobbered. He would still have been clobbered had he simply waited for Casy to arrive at his doorstep like every other Zerg who didn't do muta damage on such a tough map. But because Savior danced around the Terran army, disrupting Casy's attack while buying critical time to amass his own army at the pre-determined battle location, Savior's relative battle strength far exceeded Casy's at the key fight. And Casy was given no choice but to fight and lose.

Such a series of maneuvers forces the Terran into several no-win situations. If the Terran refuses to bring his reinforcements and simply pushes on, all Savior has to do is delay him a little bit more and prevent any reinforcements. In this scenario the Terran army will be outnumbered just the same as normally and will certainly lose in the same fashion. If Casy retreats to secure his reinforcements, Savior can simply repeat the process and delay Casy's army again, forcing the same decision and same situation. And note that if Savior succeeds in delaying the Terran army too long, defilers will be out and the Terran attack will have failed. Savior has other possibilities as well: a backstab is always an option, as well as a surprise flank on an unready Terran army. In short, Savior gains every advantage simply by running around with a purpose. In the worst case, he might lose a few units from poor control, not kill anything, and only delay the Terran army for so long. This scenario is still good enough to win the key battle with ease anyways. In the best case, he gains a decisive advantage when modern Zergs would simply pray for equality.

The pre-defiler timing push timeframe is a period of the game that the Zerg normally dreads, for there a mistake will lose outright, while good play will merely make the game even or give them a small advantage. Savior took this perilous phase and flipped it around: with best play the Terran ends up well behind, with a mistake or even just solid play they end up dead lost. But it was not that he had one perfect answer, rather, he redefined the system that Terran players had previously created to their advantage. In Savior's system, both players had many options, but so long as Savior chose correctly at every turn he would always end up better. And because he created the new paradigm, he understood it far better than anyone else and thus always made better decisions than his adversaries.

The average Starcraft viewer and player is naturally attracted to the openly flashy, micro intensive player. It is hard for anyone to claim that when he was new that he did not enjoy watching someone like Boxer, Nada, July, Jaedong, or Bisu. In fact, even for veterans this still holds true. Savior is not that kind of player where everything, or rather almost anything, is visible to the untrained eye. Looking back it is apparent that his best plays went completely unnoticed. In fact, he earned many detractors for supposedly doing the same thing every game!

Savior was a unique type of player whose play becomes more and more beautiful as it is examined at a deeper level and it becomes astonishing just how deep his play and thought processes were. The degree to which Savior placed and timed every little detail can only be described as inhuman. At some points it seems like he would just get lucky, sometimes multiple times every single game, with how perfectly everything worked out. However, given the sheer number of instances where everything seemed to go exactly as if Savior was controlling both players himself, it's quite clear that luck is not the answer: Savior is.

So yes, keep on loving the flashy, dynamic players that you have always loved. But take a little more time and effort to really look at the best of Savior's play and you will see something that nobody else has consistently displayed and probably ever will. For Savior is one of a kind, someone who is so far beyond anyone else in the area of battlefield tactics that it doesn't seem odd at all that he possibly reached the peak of ZvT almost three years ago.

And the best part? The Maneuvering tactics illustrated above are only the beginning. The more intricate Ambushing tactics take Starcraft to yet another level. The unparalleled obstacles that Savior faced demanded more than just the maneuvering illustrated above, and when confronted, he always rose to the occasion.

Waxangel

United States32927 Posts

May 24 2011 20:43 GMT

We will allow some discussion about the match-fixing scandal in this thread, but please refrain from going too far off topic.

JIJIyO

Canada1957 Posts

May 24 2011 20:44 GMT

Ma Jae Yoooooooooooooonnnnnnnnnnnnnn =( I miss you.

Hirnfrost

Germany938 Posts

May 24 2011 20:45 GMT

love the pictures

brings tears to my eyes

Count9

China10928 Posts

May 24 2011 20:45 GMT

Holy. Shit. I've only read half and have to leave, but from the half I've read this article is amazing, especially for people who weren't there in the Savior era or for people who think a SC2 player is "dominant". They don't understand that true feeling of dominant when watching the game the opposition just has 0 chance and the result a forgone conclusion.

StimiLant

United States534 Posts

May 24 2011 20:46 GMT

Epic long read, thank you, you are awesome ver.

neobowman

Canada3324 Posts

May 24 2011 20:46 GMT

Excellent, excellent article.

Ryps

Romania2740 Posts

May 24 2011 20:46 GMT

Forever Savior fan !
IPXZerg, Ma Bonjwa <3
When I saw the pictures I knew it was disciple's work

Very good article

Archas

United States6531 Posts

May 24 2011 20:46 GMT

Holy shit, this looks absolutely amazing. I've only reached the Maps section thus far, but glancing at the rest, this is a masterfully crafted article. Kudos to Ver, HawaiianPig, and Disciple!

CursOr

United States6335 Posts

May 24 2011 20:46 GMT

#10

Oh my god, I was just watching some of the old VoD/Tribute videos of sAviOr! Always gets me excited to play.

Watching sAviOr @ WWI in Paris is what got me into playing BroodWar "properly"... because it was so impressive.

Hopefully a lot of the new Starcraft 2 players who never followed BroodWar will check this out.

edit: Goddamn that is a gorgeous OP. Work of love right there. Imma link this to a good few of my SC2 buddies.

AskJoshy

United States1625 Posts

May 24 2011 20:47 GMT

#11

This article could be its own book. Well done, sir. =)

Necosarius

Sweden4042 Posts

May 24 2011 20:47 GMT

#12

SO SEXY!!! Will have to read later @___@

Mirl

Turkey173 Posts

May 24 2011 20:48 GMT

#13

this is so... AMAZINGGGGGGGGGGGGGGGGGGGGGG

what a great article...

Ma Jae Yoon,you always will be in my heart...

kOre

Canada3642 Posts

May 24 2011 20:48 GMT

#14

And to think that this is only Part 1 ... :D Amazing work!

Royalcommand

Korea (South)189 Posts

May 24 2011 20:48 GMT

#15

Great and well written article! i miss MJY :[

Carnivorous Sheep

Baa?21242 Posts

May 24 2011 20:49 GMT

#16

Savior forever. No matter what he did, no one can deny that he taught us how to play Zerg. Ma Jae Yoon - forever bonjwa :3

6d.Leek

United States76 Posts

May 24 2011 20:49 GMT

#17

Wow, great write up. I think his success is overlooked to some degree because of the match fixing...but this article does kind of put into perspective just how awesome he was. I'll make sure to watch those VODs sometime in the future.

Heyoka

Katowice25012 Posts

May 24 2011 20:49 GMT

#18

This is a really great look at Savior I'm glad its finally out :D

I came into Brood War in early 2008 so I missed his actual period of dominance, its very interesting to read about and go back to those games. That series vs NaDa in OSL finals is crazy.